<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Agile True North - Medium]]></title>
        <description><![CDATA[Finding authentic direction in Agile leadership and practice. Real-world insights on servant leadership, team transformation, and Agile methodologies from someone walking the path. Join the journey toward building adaptive, people-centered organizations through honest experiences - Medium]]></description>
        <link>https://medium.com/agile-true-north?source=rss----e69a8c7e233d---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Agile True North - Medium</title>
            <link>https://medium.com/agile-true-north?source=rss----e69a8c7e233d---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 07:48:54 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/agile-true-north" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[The Release Was Successful. So Why Was Everyone Still Working?]]></title>
            <link>https://medium.com/agile-true-north/the-release-was-successful-so-why-was-everyone-still-working-7b496e088b12?source=rss----e69a8c7e233d---4</link>
            <guid isPermaLink="false">https://medium.com/p/7b496e088b12</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[software-engineering]]></category>
            <dc:creator><![CDATA[Soumya]]></dc:creator>
            <pubDate>Mon, 16 Nov 2026 17:16:01 GMT</pubDate>
            <atom:updated>2026-11-16T17:16:01.516Z</atom:updated>
            <content:encoded><![CDATA[<h4><strong>A successful deployment is a milestone. A reliable system is a responsibility. And the difference often becomes visible only after everyone celebrates.</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*3qs2rIB83Q6jzVkG" /><figcaption>Photo by <a href="https://unsplash.com/@markusspiske?utm_source=medium&amp;utm_medium=referral">Markus Spiske</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>The release went well. <br>At least, that’s what the status update said.</p><p>The deployment was completed within the planned window. Initial checks passed. <br>The teams confirmed their activities were done. The release communication went out. Green across the board. <br>After weeks of planning, testing, coordination, and last-minute conversations, everyone had good reason to feel relieved.</p><p>And then came Monday morning.<br>A few users reported unexpected behavior. A downstream process wasn’t working quite as expected.<br>Support needed clarification on a recent change. Someone from engineering had to jump back in.<br>Another meeting was scheduled. A few more follow-ups appeared. Nothing necessarily catastrophic. <br>But enough to make you wonder — <strong>If the release was successful, why were we still working so hard to make it work?</strong></p><p>That question has stayed with me because it reveals something we don’t always talk about during delivery planning.<br>Sometimes we define success around the moment something goes live. But the people using and supporting that solution experience success very differently.</p><h4>We Planned for the Release. Did We Plan for What Came After?</h4><p>Think about how much attention a major release receives.</p><p>There are readiness reviews. Testing checkpoints. Dependency discussions. Go/no-go decisions. Deployment plans. Rollback procedures. Stakeholder communications.And usually, plenty of meetings.<br>All necessary.</p><p>But here’s something I’ve come to appreciate.</p><p><strong>We can spend weeks preparing to deploy a solution and surprisingly little time preparing the organization to operate it.</strong></p><p>The difference may not be obvious during planning. It becomes visible afterward.</p><p>When a support team receives an issue but isn’t sure which group owns it. When monitoring shows that a service is available, but a critical business transaction isn’t completing.<br>When an operational workaround exists, but only a few people know about it.<br>When an unexpected problem requires someone to reconstruct conversations from weeks earlier.<br>These aren’t always failures of technical execution. Sometimes they’re failures of coordination, readiness, or knowledge transfer. <br>And that’s what makes them interesting from a delivery leadership perspective.</p><h4>The First Thing I’d Question Is Our Definition of Done</h4><p>In Agile, we spend considerable time discussing the Definition of Done. What conditions must be satisfied before work can be considered complete?<br>Testing? Acceptance? Documentation? Deployment? <br>But when we’re delivering something that will be used and supported by others, I think there’s another dimension worth considering.</p><p><strong>Done for whom? <br></strong>Done for the development team? Done for the project? Done for the business user? Done for the support team?</p><p>Those answers may not be identical.<br>A feature can be technically complete but difficult to support. A migration can finish on schedule while operational teams are still learning how to respond to unexpected behavior.<br>A new capability can be available without users being fully prepared to adopt it.</p><p>This doesn’t mean we should keep every project open indefinitely. It means our completion criteria should reflect the responsibilities that come after delivery.</p><p>I’ve found it useful to think about three separate questions:</p><ul><li><strong>Can we release it?</strong></li><li><strong>Can people use it successfully?</strong></li><li><strong>Can the organization support it reliably?</strong></li></ul><p>A good release plan should help us answer all three.</p><h4>The Handoff Is Where Small Gaps Become Big Frustrations</h4><p>Consider a hypothetical situation. <br>A team delivers an enhancement that connects several systems. Each participating team completes its assigned work. Testing is signed off. <br>The release proceeds.<br>A few days later, a business process fails intermittently. The application team checks its component. Everything looks fine. The integration team checks its service. No obvious failure. <br>The support team starts asking questions. <br>Who owns the issue? Which team should investigate first?<br>Where can they see the full transaction? Who makes the decision if a workaround is needed?<br>Suddenly, the biggest challenge isn’t necessarily fixing the defect. <strong>It’s figuring out how the organization should respond.</strong></p><p>And this is where I think delivery leaders can make a meaningful difference. Not by trying to become the technical expert for every component. But by asking questions that expose the gaps between them.</p><p><strong>Who owns the end-to-end experience? What happens when an issue crosses team boundaries?<br>Does support know where to begin? Have we tested the handoff, not just the functionality?</strong></p><p>Because sometimes every individual team can complete its responsibilities and the overall experience can still fall short.</p><blockquote><strong><em>A solution can be ready within each team and still not be ready across the organization.</em></strong></blockquote><h4>Hypercare Should Be More Than a Safety Net</h4><p>Many major releases include a hypercare period. <br>Extra monitoring. Faster issue response. Closer coordination between teams. A defined window to stabilize the solution.</p><p>That makes sense.</p><p>But I think we sometimes treat hypercare as an extension of deployment support rather than an opportunity to improve how the organization operates.</p><p>Imagine two approaches.</p><p>In the first, the team tracks issues, resolves them, and closes hypercare once the number of open tickets becomes manageable.</p><p>In the second, the team does all of that but also looks for patterns.<br>Which issues were detected by monitoring? Which were reported by users? Where did ownership become unclear? Which problems took longer to resolve than expected? What information was missing when support first investigated? Were similar issues appearing repeatedly?</p><p>The second approach gives us something beyond a stable release. It gives us evidence about where the delivery and support process needs to improve. And that’s an opportunity we shouldn’t waste.</p><h4>The Numbers Matter. But the Questions Behind Them Matter More.</h4><p>This is where reliability metrics become useful. Not because we need another dashboard. But because certain measurements can help us understand whether our releases and support processes are becoming more dependable.</p><p>Take <strong>change failure rate</strong>.<br>It tells us what proportion of changes result in a failure requiring remediation. If that number increases, it may be a signal to examine testing, release practices, change size, or deployment readiness.<br>But the number alone doesn’t tell us why.</p><p>Or consider <strong>mean time to restore (MTTR)</strong>. How long does it take to restore service after an incident? Suppose a team consistently restores service within 30 minutes. That sounds encouraging.<br>But what if the same incident occurs every week? Are we really improving reliability?<br>Or have we simply become very good at recovering from a problem we haven’t eliminated?</p><p>And then there are <strong>service level indicators (SLIs)</strong> — measurements of how a service actually performs, such as availability, latency, or successful completion of a critical transaction.</p><p>A service might technically be available while users are still unable to complete the task that matters to them.<br><strong>That distinction is important.<br></strong>Because a healthy system isn’t necessarily the same thing as a healthy user experience.</p><blockquote><strong><em>Reliability metrics become valuable when they help us ask better questions, not when they simply give us better-looking reports.</em></strong></blockquote><h4>What I Would Do Differently</h4><p>If I were preparing for a significant release, I would want the conversation to extend beyond the usual readiness checklist. <br>Not by adding more meetings or introducing unnecessary governance. But by making a few important things visible earlier.</p><p><strong>First, I would bring operational teams into the conversation before the release.</strong></p><p>Not simply to inform them of what’s coming. But to understand what they need in order to support it. <br>Sometimes a 20-minute conversation with the people who will operate the solution reveals gaps that several planning meetings have missed.</p><p><strong>Second, I would make cross-team ownership explicit.</strong></p><p>Especially for workflows that span multiple applications or services. When something goes wrong, people shouldn’t have to discover the escalation path while the issue is already affecting users.</p><p><strong>Third, I would agree on what stability looks like.</strong></p><p>Not just successful deployment. What signals should we monitor? What level of unexpected behavior is acceptable during hypercare? What conditions would indicate that we aren’t ready to transition to normal support?</p><p><strong>Fourth, I would treat recurring issues as improvement opportunities.</strong></p><p>Resolving an issue is necessary. Understanding why it happened — and whether it is likely to happen again — is where the longer-term value begins.</p><p>And finally, <strong>I would make improvement actions visible and accountable.</strong></p><p>A finding without an owner is easy to forget. An action without a timeline is easy to postpone. And a completed task without evidence of improvement may not have changed much at all.</p><h4>A Retrospective Can Identify a Problem. It Can’t Fix It by Itself.</h4><p>This is one of the patterns I think many teams will recognize. <br>A release retrospective happens. People share thoughtful observations. Someone mentions that support wasn’t involved early enough. Another person raises a testing gap. <br>Someone else points out that the escalation process was unclear. Everyone agrees. Action items are documented. The meeting ends.</p><p>Then the next release begins. Priorities shift. Calendars fill up. And those improvement actions slowly lose visibility.</p><p>Three months later, another release encounters a similar issue. Another retrospective. Another action item. Another agreement to improve. <br>At some point, I think we need to ask — <strong>Are we learning from our releases, or are we simply documenting what happened?</strong></p><p>The difference is follow-through.</p><p>For a meaningful improvement, I would want to know: Who owns it? When should it be completed? What specifically needs to change? How will we validate that it worked? And will that change become part of the normal release process?<br>Because closing an action item isn’t the same as changing how work happens.</p><h4>Reliability Is a Shared Outcome, Not a Department</h4><p>It’s tempting to think of reliability as something owned entirely by engineering or operations. And certainly, technical expertise is essential. But many of the conditions that influence reliability are created much earlier.<br>Unclear requirements. Unmanaged dependencies. Compressed testing windows. Late scope changes. Poorly coordinated handoffs. Missing operational documentation. Unresolved ownership questions.<br>These are also delivery concerns. And they often require collaboration across multiple teams.<br>That doesn’t mean everyone needs to become an SRE or an incident response specialist.</p><p>It means everyone involved in delivering change should understand how their decisions affect what happens after that change reaches users.</p><p><strong>Reliability doesn’t begin when an incident occurs. It begins with the decisions we make while planning and delivering the work.</strong></p><h4>A Small Experiment to Try</h4><p>For your next major release, try adding one short conversation before declaring readiness. Bring together representatives from delivery, engineering, testing, and support.</p><p>Instead of walking through the entire release plan again, ask just five questions:</p><ol><li><strong>What could go wrong after deployment that our current testing might not reveal?</strong></li><li><strong>How would we know something was wrong before a user reported it?</strong></li><li><strong>Who would own the response if the issue crossed multiple teams?</strong></li><li><strong>What information would the support team need immediately?</strong></li><li><strong>What would make us confident enough to end hypercare?</strong></li></ol><p>You don’t need an elaborate framework. You don’t need another large dashboard. You need enough clarity to identify the gaps while there’s still time to address them.<br>And after the release, revisit those questions. See which assumptions held. Which ones didn’t. And what should change before the next release.</p><p>That is how a readiness conversation becomes a continuous improvement practice.</p><h4>A Green Release Is Only the Beginning</h4><p>I still believe teams should celebrate successful releases. The effort behind them deserves recognition. But I’ve become more interested in what happens after the celebration. <br>Does the solution behave as expected? Can users accomplish what they need? Can support teams respond confidently? Are we detecting problems early? Are we learning from incidents?<br>Are the improvements we identify actually becoming part of how we work?</p><p>Those questions tell us something a deployment status alone cannot. Because delivery success is a moment. Reliability is something we demonstrate over time. And perhaps that’s the mindset shift worth making.</p><p>Instead of asking only — <strong>“Did we release successfully?”<br></strong>We should also ask — <strong>“Did we leave the organization better prepared to operate what we delivered?”</strong></p><p>That’s a much more meaningful definition of success. And one that deserves just as much attention as the release itself.</p><p><a href="https://medium.com/agile-true-north"><em>Agile True North</em></a><em> is about the lessons that rarely fit neatly into frameworks — the ones we learn while working with real teams, real constraints, and real people.</em></p><p>Something to think about — Think of the last release your team celebrated. What happened in the days that followed? <br>Did the work truly become easier for the people using and supporting it — or did the real work begin after the dashboard turned green?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7b496e088b12" width="1" height="1" alt=""><hr><p><a href="https://medium.com/agile-true-north/the-release-was-successful-so-why-was-everyone-still-working-7b496e088b12">The Release Was Successful. So Why Was Everyone Still Working?</a> was originally published in <a href="https://medium.com/agile-true-north">Agile True North</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Most Valuable Skill in the Age of AI Might Be Knowing What to Ask]]></title>
            <link>https://medium.com/agile-true-north/the-most-valuable-skill-in-the-age-of-ai-might-be-knowing-what-to-ask-e1e57f36605d?source=rss----e69a8c7e233d---4</link>
            <guid isPermaLink="false">https://medium.com/p/e1e57f36605d</guid>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[future-of-work]]></category>
            <category><![CDATA[critical-thinking]]></category>
            <category><![CDATA[prompt-engineering]]></category>
            <dc:creator><![CDATA[Soumya]]></dc:creator>
            <pubDate>Fri, 06 Nov 2026 17:16:01 GMT</pubDate>
            <atom:updated>2026-11-06T17:16:01.479Z</atom:updated>
            <content:encoded><![CDATA[<h4><em>When answers are becoming easier to generate, perhaps the real advantage lies in knowing which questions deserve to be asked.</em></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*cq2FQ0Y0ftpJOkVL" /><figcaption>Photo by <a href="https://unsplash.com/@brookecagle?utm_source=medium&amp;utm_medium=referral">Brooke Cagle</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>I remember a time when being the person with the answers carried a certain kind of credibility. <br>Someone had a problem. You knew the process. You understood the system. You had the experience. You could tell them what needed to happen next. <br>And that was valuable.It still is.</p><p>But lately, I’ve been thinking about how much that equation is changing. Today, I can ask an AI tool to explain a concept, structure a document, analyze information, suggest a workflow, or help me build something I wouldn’t have known how to build on my own.</p><p>Within seconds, I have an answer. Sometimes a very impressive one. But I’ve also learned something through these experiences.</p><p><strong>Getting an answer quickly doesn’t necessarily mean getting the answer you need.</strong></p><p>And perhaps that’s where a different kind of skill is becoming increasingly important. Not simply knowing the answers. But knowing what to ask.</p><h4>The First Answer Is Rarely the Most Interesting One</h4><p>Let me start with something familiar. Imagine asking an AI assistant: <em>“Help me automate our weekly project status report.”</em></p><p>You’ll probably get a reasonable response. It may suggest pulling data from a project management tool, summarizing completed work, highlighting blockers, and distributing the report automatically.<br>Nothing wrong with that.</p><p>In fact, it might be a useful starting point. But what if the real problem isn’t the time spent preparing the report? <br>What if the real problem is that leadership keeps discovering delivery risks too late?<br>Suddenly, the conversation changes.<br>We aren’t just trying to produce a report faster. We’re trying to understand why important signals aren’t reaching the right people early enough.</p><p>Now the questions become different.</p><p>What information tells us a project may be drifting? Which dependencies have been waiting too long? <br>What work repeatedly carries over? Where are decisions pending? <br>Which risks are already visible in the workflow but aren’t being discussed? Who needs to know, and when?</p><p>Those questions may eventually lead us to a completely different automation. And potentially a much more valuable one. <br><strong>The quality of the solution changed because the quality of the questions changed.<br></strong>That’s something I’ve come to appreciate more and more.</p><h4>I’ve Never Believed You Need to Know Everything to Contribute Something Valuable</h4><p>My own experience with AI has reinforced this belief. I don’t come from a traditional software development background. <br>I understand technology through the work it enables, the systems it connects, the people involved, and the problems we’re trying to solve.</p><p>When I first started exploring AI-assisted automations, I wasn’t approaching them as someone who could independently write every line of code. I approached them as someone who understood the workflow.<br>I knew where the frustrations were. I knew what information was missing. I knew what a useful result should look like. <br>And perhaps most importantly, I knew when an answer didn’t make sense.<br>That became my starting point. <br>I could explain the problem to an AI tool. Ask it to suggest an approach. Understand the steps. Challenge the assumptions. Ask it to break down the implementation.<br>Test the result. Identify what wasn’t working. And keep refining. <br>Sometimes the first attempt worked. Often it didn’t. But every iteration taught me something. <br>Not just about the technology. About how to think more clearly. <br>And that experience has changed how I look at technical confidence.</p><blockquote><strong><em>You don’t always need to know how to build something from scratch. But you do need to understand what you’re trying to build, why it matters, and how to recognize whether it works.</em></strong></blockquote><p>That doesn’t eliminate the need for technical expertise. <br>There are still situations where architecture, security, reliability, integration design, and engineering judgment are essential. <br>But it opens the door for people with different kinds of expertise to participate meaningfully in creating solutions.</p><p>And I think that’s important.</p><h4>A Vague Question Usually Produces a Vague Solution</h4><p>One thing I’ve noticed when working with AI is how easily we can confuse a polished response with a useful one.</p><p>Ask — <em>“How can I improve sprint delivery?” </em>You’ll probably receive a familiar list. Improve planning. Remove blockers. Strengthen communication. Clarify priorities. Track velocity. All reasonable suggestions. <br>But how much of that actually helps? Now imagine explaining the situation differently.</p><p><em>“We have a team that consistently completes most of its planned work, but a few stories carry over every sprint. Those stories usually depend on another team, and the dependency isn’t identified until development has already started. I don’t want to introduce more meetings or create another reporting burden. What changes could help us identify these risks earlier?”</em></p><p>That’s a different question. It provides context. It explains the recurring pattern. It identifies a constraint. It describes what we don’t want. And it makes the desired outcome clearer.</p><p>The response now has a better chance of addressing the actual problem. This isn’t just a lesson in prompting AI. It’s a lesson in communication. Because the same principle applies when we speak with our teams. <br>When we explain problems vaguely, we often receive generic solutions. When we explain them thoughtfully, we create room for better conversations.</p><h4>Good Questions Require Good Observation</h4><p>Here’s something I don’t think gets discussed enough. Before you can ask a useful question, you often need to notice something worth questioning. And noticing requires attention. Consider a team that repeatedly carries work from one sprint to the next. <br>It’s easy to ask — <strong>“Why aren’t we completing our commitments?” <br></strong>But what if we look more closely? Are the same types of stories carrying over? Are dependencies being discovered too late? <br>Are priorities changing during the sprint? <br>Are stories entering development before important questions are answered? Is the team taking on more work than its actual capacity allows? <br>Or is the definition of completion different across the people involved?</p><p>Each question opens a different path. And each path could lead to a different solution. The challenge isn’t necessarily finding more data. Most teams already have plenty of it. The challenge is recognizing which patterns deserve attention.</p><blockquote><strong><em>The most useful question often comes from noticing what everyone else has started accepting as normal.</em></strong></blockquote><p>That is true whether we’re coaching a team, improving a process, or exploring an AI solution.</p><h4>AI Can Be an Excellent Thinking Partner. But Only If We Let It Challenge Us.</h4><p>One of the things I enjoy most about working with AI is that I don’t have to stop at the first answer.</p><p>I can ask — <em>“What am I missing?” “What assumptions are you making?” <br>“What could go wrong with this approach?” “Is there a simpler solution?” <br>“What would someone from engineering challenge here?” <br>“What would make this recommendation unsuitable for our situation?”</em></p><p>These questions change the interaction. <br>Instead of treating AI as a machine that produces answers, I start using it as a way to explore possibilities. But there is an important caution here. <br>AI can sound confident even when its reasoning is incomplete or its assumptions are wrong. <br>It may agree with a weak idea. It may suggest a technically plausible solution that doesn’t fit the environment. It may overlook a dependency, invent a detail, or misunderstand the intent.</p><p>So I don’t think the goal should be to become better at getting AI to agree with us. Quite the opposite. <br><strong>We should become better at asking it to challenge us. <br></strong>And then we need to evaluate those challenges ourselves. That’s where critical thinking remains essential.</p><h4>Sometimes the Best Prompt Is a Better Conversation With Yourself</h4><p>This is a realization I didn’t expect. When I try to explain a problem clearly enough for an AI tool to help me, I often discover gaps in my own thinking.</p><p>I start writing — <em>“I want to automate this process because…”</em></p><p>And then I pause. Because I haven’t fully explained why the process exists. Or I realize I don’t know who owns a particular decision. Or I haven’t defined what a successful outcome would look like. Or I’ve assumed something is a technical problem when it may actually be an ownership problem.</p><p>The act of explaining forces me to think. And sometimes that thinking is more valuable than the generated answer. I’ve started seeing prompting as more than writing instructions for a tool.<br>At its best, it’s an exercise in making our reasoning visible. <br>What do I know? What am I assuming? What information is missing? What constraints matter? What would success look like? <br>How will I recognize a bad result?</p><p>Those are useful questions even when AI isn’t involved.</p><h4>This Reminds Me of Something I’ve Always Valued in Agile Coaching</h4><p>A good coach doesn’t always walk into a room with the answer. Sometimes the most valuable contribution is a question that helps people see the situation differently.</p><p>A team says — <em>“We need more capacity.”</em></p><p>You ask — <strong>“What would you stop doing if additional capacity wasn’t available?”</strong></p><p>A stakeholder says — <em>“Everything is high priority.”</em></p><p>You ask — <strong>“If only two things could be delivered this quarter, which two would you protect?”</strong></p><p>A team says — <em>“We need another dashboard.”</em></p><p>You ask — <strong>“What decision are we unable to make with the information we already have?”</strong></p><p>A leader says — <em>“We need better predictability.”</em></p><p>You ask — <strong>“Where does our current plan usually stop reflecting reality?”</strong></p><p>These aren’t complicated questions. But they can change the direction of a conversation. <br>They move people away from the first proposed solution and toward a deeper understanding of the problem. <br>And that’s why I see a strong connection between coaching and working effectively with AI.</p><p>Both benefit from curiosity. Both require context. Both involve challenging assumptions. Both become more valuable when we resist the urge to settle for the first answer. <br>Of course, coaching is fundamentally human. It involves trust, empathy, relationships, and understanding what people may not feel comfortable saying.</p><p>AI doesn’t replace that.</p><p>But the discipline of asking thoughtful questions carries across both worlds.</p><h4>Knowing What to Ask Also Means Knowing When to Stop Asking</h4><p>There is another side to this. We can become so focused on exploring possibilities that we never make a decision.<br>Another prompt. Another alternative. Another analysis. Another comparison. Another refinement. <br>AI makes it remarkably easy to keep generating options. And sometimes that’s useful.<br>But at some point, we have to decide. Is this good enough to test? Do we have enough information? What risk remains? Who needs to review this? What should happen next? <br>The ability to ask questions is valuable. <br>The ability to recognize when we have enough information to act is equally important. <br>Otherwise, we risk replacing analysis paralysis with AI-assisted analysis paralysis. And that’s not much of an improvement.</p><h4>The Skill Isn’t Just Asking Better Questions. It’s Recognizing Better Answers.</h4><p>Imagine two people using the same AI tool. Both ask it to design an automation for identifying delivery risks. Both receive a detailed implementation plan. <br>One accepts the output because it looks complete. The other starts examining it. <br>Where is the data coming from? How current is that information? What does the system consider a blocker? Will it distinguish a genuinely critical dependency from a routine delay? Could it create unnecessary alerts? What happens when the underlying data is incomplete? Who reviews the recommendations before they’re shared?</p><p>The second person may not know how to implement every technical component independently. But they understand the operating environment well enough to identify important gaps. <br>That understanding matters. <br>Because a technically functional automation can still be operationally ineffective. It can generate noise. It can create false confidence. It can make the wrong behavior easier. <br>Or it can solve a problem nobody actually has.</p><p><strong>The ability to evaluate an answer may become just as important as the ability to generate one.</strong></p><h4>What Does This Mean for People Who Aren’t Technical?</h4><p>I think this is where the conversation becomes especially interesting. There are many experienced professionals who understand their organizations deeply. <br>They understand customers. Processes. Dependencies. Business constraints. Risk. Communication. Decision-making. <br>They know where work gets stuck. They know which activities consume time. They know what teams complain about but rarely formally raise.</p><p>And yet, when conversations turn toward AI, some may assume they don’t have enough technical knowledge to contribute.</p><p>I would challenge that assumption. You don’t have to begin by learning how to build a complex system. You can begin by understanding one workflow. <br>One recurring frustration. One decision that takes too long. One repetitive activity that consumes valuable attention.</p><p>Then ask — Could this be simpler? Could information be easier to access? Could we identify the issue earlier? Could a tool help? <br>What would we need to validate? Who should be involved?</p><p>That’s a meaningful starting point. And if the solution requires deeper technical expertise, bring in the people who have it. <br>The goal isn’t to pretend everyone has the same skills. It’s to recognize that meaningful innovation often requires different kinds of expertise working together.</p><h4>A Small Experiment to Try</h4><p>The next time you use AI to solve a work problem, resist the temptation to begin with — <em>“Give me a solution for…”</em></p><p>Instead, try a different conversation. Start by describing the situation. Explain what is happening today. Describe what isn’t working. Share the constraints. Explain what a better outcome would look like.</p><p>Then ask — <strong>“Before suggesting a solution, what questions would you ask to understand this problem better?”</strong></p><p>Let the tool challenge your description. Answer the questions. Notice where you struggle to provide clarity. Those gaps may tell you something important. Then ask for possible approaches. But don’t stop there.</p><p>Ask — <strong>“What assumptions does each approach depend on?” “What would make this approach fail?” “What is the simplest version we could test?”</strong></p><p>And finally — <strong>“What should a human validate before we trust the result?”</strong></p><p>You may end up with a better solution. But more importantly, you may end up with a better understanding of the problem. <br>And that’s a skill worth developing far beyond AI.</p><h4>Perhaps Expertise Is Changing, Not Disappearing</h4><p>I don’t believe AI makes experience irrelevant. If anything, I think experience becomes valuable in a different way. <br>When answers were difficult to obtain, knowing where to find them mattered enormously. <br>When producing something required specialized skills, the ability to create it carried obvious value.</p><p>Those skills still matter.</p><p>But as AI makes more information and capabilities accessible, the challenge shifts. We need to know what deserves attention. What information can be trusted. What assumptions should be challenged. <br>Which trade-offs matter. When a solution is unnecessarily complex. When something sounds right but doesn’t fit reality. <br>And when the most responsible answer is:</p><p><strong>“We don’t know enough yet.”</strong></p><p>That’s not a weakness. That’s judgment. And judgment is built through experience, curiosity, reflection, and the willingness to question our own thinking.</p><h4>The Question Behind the Question</h4><p>I’ve started noticing that many workplace requests contain another question underneath them.<em>“<br>Can we automate this report?” </em>may really mean — <strong>“Why does getting reliable information take so much effort?”</strong></p><p><em>“Can we improve sprint velocity?” </em>may really mean — <strong>“Why does our work keep getting interrupted?”</strong></p><p><em>“Can AI write our requirements?” </em>may really mean — <strong>“Why do we keep discovering misunderstandings after development begins?”</strong></p><p><em>“Can we build a better dashboard?” </em>may really mean — <strong>“Why aren’t we recognizing problems early enough to do something about them?”</strong></p><p>The first question often points toward a solution. The second points toward the problem. And understanding that difference can change everything.</p><p>So perhaps the most valuable skill in the age of AI isn’t simply knowing how to use the latest tools. It isn’t having the perfect prompt. And it certainly isn’t accepting every confident answer we receive. <br>It’s having the curiosity to look beyond the obvious. The patience to understand the problem. The confidence to challenge an answer. And the judgment to know what to do with what we learn.</p><p>Because AI may continue getting better at answering our questions. But the responsibility for deciding <strong>which questions matter</strong> still belongs to us.</p><blockquote><strong><em>The future may reward people who know how to work with AI. But it will also need people who know when to question it.</em></strong></blockquote><p>And maybe that’s where meaningful leadership begins. Not with having all the answers. But with the courage to ask a better question.</p><p><a href="https://medium.com/agile-true-north"><em>Agile True North</em></a><em> is about the lessons that rarely fit neatly into frameworks — the ones we learn while working with real teams, real constraints, and real people.</em></p><p><em>Something to think about — Think of the last time you asked AI, a colleague, or your team for a solution. <br>Did you already understand the problem well enough to recognize whether the answer was right?</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e1e57f36605d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/agile-true-north/the-most-valuable-skill-in-the-age-of-ai-might-be-knowing-what-to-ask-e1e57f36605d">The Most Valuable Skill in the Age of AI Might Be Knowing What to Ask</a> was originally published in <a href="https://medium.com/agile-true-north">Agile True North</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[We Asked for AI. But What Problem Were We Trying to Solve?]]></title>
            <link>https://medium.com/agile-true-north/we-asked-for-ai-but-what-problem-were-we-trying-to-solve-438817ab9b94?source=rss----e69a8c7e233d---4</link>
            <guid isPermaLink="false">https://medium.com/p/438817ab9b94</guid>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[future-of-work]]></category>
            <category><![CDATA[digital-transformation]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Soumya]]></dc:creator>
            <pubDate>Tue, 03 Nov 2026 17:21:01 GMT</pubDate>
            <atom:updated>2026-11-03T17:21:01.513Z</atom:updated>
            <content:encoded><![CDATA[<h4><em>When adopting new technology becomes the goal, it’s surprisingly easy to forget what we were trying to improve.</em></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*R6oYKUXYInry4ifn" /><figcaption>Photo by <a href="https://unsplash.com/@tinkerman?utm_source=medium&amp;utm_medium=referral">Immo Wegmann</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>It started with a question that seems to be appearing in almost every workplace conversation lately.</p><p><strong>“Where can we use AI?” </strong>A perfectly reasonable question.</p><p>After all, AI is changing how we work. The possibilities are exciting, and nobody wants to be the team that gets left behind.</p><p>So the brainstorming begins. Can we automate status reports? Can AI summarize meetings? Can we generate requirements faster? <br>Can we create dashboards automatically? Can we build an assistant to answer questions?</p><p>Ideas start flowing. People get excited. Someone volunteers to explore a proof of concept. <br>Before long, there’s a list of AI opportunities, a few demonstrations, and perhaps even a roadmap.</p><p>Everything looks promising.</p><p>But somewhere in the middle of all that excitement, I think we need to ask a different question.<br><strong>What problem were we actually trying to solve?<br></strong>Because there is a subtle difference between finding a problem worth solving and finding something for AI to do.<br>And that difference matters more than we sometimes realize.</p><h4>The Pressure to Do Something With AI Is Real</h4><p>I understand the excitement. I share it.</p><p>I’ve spent considerable time exploring how AI can improve everyday delivery work, from requirements preparation and workflow automation to identifying risks and reducing repetitive reporting.</p><p>And some of those experiments have genuinely changed how I approach my work. But I’ve also noticed something interesting about the broader conversation around AI.</p><p>Sometimes the discussion begins with the technology rather than the need. <em>“We should have an AI use case.” <br>“Can we demonstrate some AI adoption this quarter?”<br>“What processes can we automate?”<br>“Can we build an agent for this?”</em></p><p>None of these questions is necessarily wrong.<br>Experimentation matters. Learning matters. Sometimes you discover possibilities simply by trying something new.<br>But there is a risk when adoption itself becomes the measure of progress.</p><p>We can end up celebrating the number of AI use cases without asking whether anyone’s work actually improved.<br><strong>A successful demonstration is not necessarily a successful improvement.<br></strong>And I think that’s where the conversation needs to mature.</p><h4>Let Me Start With Something Familiar</h4><p>Think about the weekly status report. Almost every team has some version of it.<br>Someone collects updates. Checks the work management tool. Follows up on missing information. Reviews blockers. Updates a dashboard. Writes a summary. Sends it to stakeholders. <br>And repeats the process the following week.</p><p>Now imagine someone suggests — <em>“Let’s use AI to generate the status report.”</em></p><p>Sounds like a great idea. And honestly, it might be.<br>But before building anything, I would want to understand what is actually painful about the current process.<br>Is collecting information taking too long? Are people entering the same information in multiple places?<br>Are updates inconsistent? Are important risks getting buried?<br>Does leadership receive information too late to act?<br>Or is the report being generated every week even though nobody really uses it?</p><p>Those are very different problems. And they don’t all require the same solution.<br>If the problem is repetitive information gathering, automation might help. If the problem is inconsistent data, we may need better working agreements. <br>If the problem is unclear ownership, AI won’t magically create accountability.<br>And if nobody uses the report?<br>Perhaps the most valuable improvement is to stop producing it.</p><p><strong>Not every inefficient process needs to be automated. Some need to be questioned first.</strong></p><h4>I’ve Learned to Look for Friction Before Looking for Tools</h4><p>One thing that has shaped my approach to AI is paying attention to the small frustrations people have stopped talking about.</p><p>The tasks everyone accepts as part of the job. The manual follow-ups. The repeated copying of information. The time spent reconciling updates from different systems. The same questions answered every week.</p><p>The same risks discovered only after someone manually reviews dozens of work items. These things rarely appear as major organizational problems. They’re just how work gets done. And that’s what makes them interesting.</p><p>Because sometimes the best opportunities aren’t the ones announced in a strategy meeting. They’re hiding inside ordinary work.</p><p>Someone says — <em>“I spend every Monday putting this together.”<br></em>Or — <em>“I have to check five different places to get the complete picture.”<br></em>Or — <em>“We usually find out something is delayed when someone asks about it.”</em></p><p>Those statements catch my attention. Not because every frustration deserves an AI solution. But because they reveal where human attention is being consumed. <br>And <strong>human attention is one of the most valuable resources we have.</strong></p><h4>A Small Automation Can Reveal a Much Bigger Problem</h4><p>Let me share an example of the kind of thinking that has influenced my work.</p><p>When exploring ways to improve requirements preparation, the obvious opportunity was speed. Writing a detailed epic manually could take considerable time, especially when information was scattered across discussions, notes and business requirements. <br>AI could help generate a structured first draft. <br>Useful.<br>But once you look beyond the drafting time, another problem becomes visible.<br>How often do requirements move back and forth because important details weren’t clear at the beginning? How often does engineering need another clarification?<br>How often do acceptance criteria need to be revisited? How much time is spent aligning everyone on what was actually intended?</p><p>Now the opportunity becomes more interesting. It isn’t simply about writing an epic faster. <br>It’s about helping people think through requirements more completely before the work reaches engineering.<br>A well-designed assistant can prompt for missing context, identify ambiguity, suggest acceptance criteria and help structure the discussion.</p><p>But the product owner still owns the requirement. Engineering still validates feasibility. The team still needs the conversation.</p><p><strong>The goal isn’t to replace the thinking. It’s to make the thinking better prepared.</strong></p><p>That distinction changes how I evaluate automation.</p><h4>Faster Isn’t Always Better</h4><p>Here’s a question I think we should ask more often. <br><strong>What happens after we make something faster?</strong></p><p>Suppose AI reduces a task from 40 minutes to 10. That’s a meaningful improvement. <br>But what happens to those 30 minutes? Do people have more time to solve problems? Do they improve the quality of their work? <br>Do they spend more time understanding customer needs? <br>Or does the organization simply expect them to complete four times as many tasks?</p><p>This is where efficiency and effectiveness can quietly part ways. We can become incredibly efficient at producing things that don’t create much value.</p><p>More reports. More documentation. More requirements. More meeting summaries. More dashboards.<br>All generated faster than ever before. <br>But if those outputs aren’t improving decisions or outcomes, what have we really gained?</p><blockquote><strong><em>AI can help us produce more. But producing more isn’t automatically the same as creating more value.</em></strong></blockquote><p><strong><em>And perhaps the harder leadership question is not how much time AI saves. It’s what we choose to do with the time it gives back.</em></strong></p><h4>Sometimes AI Makes a Broken Process Look Better</h4><p>This is the part I think deserves more attention.</p><p>Imagine a team has a recurring problem with unclear ownership. Work gets stuck because nobody knows who is responsible for the next step. Someone proposes an AI assistant to identify and follow up on blocked items.</p><p>The assistant works beautifully. It scans the workflow. Identifies aging tasks. Sends reminders. Generates summaries. Escalates unresolved items. Wonderful. <br>Except the underlying ownership problem still exists. We’ve made the confusion easier to monitor. We haven’t necessarily made it disappear.</p><p>In fact, we may have created a very sophisticated system for repeatedly reminding people about a problem nobody has addressed.<br>I’ve come to believe that before automating a workflow, we should understand whether the workflow itself makes sense.</p><p>Who owns the work? What triggers the next step? Where does the information come from? Who makes the decision? What should happen when something goes wrong? <br>If those answers aren’t clear, automation may simply move confusion faster.</p><p><strong>Technology can accelerate a process. It cannot guarantee that the process deserves to be accelerated.</strong></p><h4>But We Shouldn’t Overthink Every Experiment Either</h4><p>There is another side to this conversation. And I think it’s important. <br>If we demand a perfect business case before trying anything with AI, we risk discouraging the very experimentation that helps people discover what’s possible. <br>Not every experiment needs to begin with a large transformation initiative. Sometimes you try something because you’re curious. <br>You build a small assistant. You automate a repetitive task. You experiment with prompts. You discover a capability you didn’t know existed. And that experiment opens up an entirely new possibility.</p><p>I’ve experienced this myself.</p><p>You don’t always need to be the most technical person in the room to explore what AI can do. You do need curiosity, a willingness to learn, and enough understanding of the workflow to recognize whether the result makes sense.<br>The key is knowing the difference between <strong>experimentation and operational adoption</strong>.</p><p>Experimentation asks — <em>“What is possible?”<br></em>Adoption asks — <em>“Is this useful, reliable, safe, and worth integrating into how we work?”</em></p><p>Both matter.</p><p>But they shouldn’t be confused. A clever prototype may deserve celebration. It doesn’t automatically deserve a permanent place in the workflow.</p><h4>The Human Judgment We Should Be Protecting</h4><p>One concern I keep returning to is the temptation to automate not only repetitive tasks, but also the thinking around them.</p><p>Take delivery risk.</p><p>AI can scan work items, identify aging dependencies, recognize patterns and generate risk summaries.</p><p>That is valuable. But an aging dependency isn’t automatically a critical risk. A delayed story isn’t automatically a delivery failure. <br>A team with lower velocity isn’t automatically underperforming. <br>Context matters. People matter. Trade-offs matter.</p><p>A system can identify something unusual. Someone still needs to understand what it means. And that requires judgment. I don’t want AI making every decision for me. I want it helping me notice what I might otherwise miss. I want it reducing the time spent collecting information so I can spend more time understanding it. I want it helping me ask better questions.<br>Not removing the need to ask them.</p><blockquote><strong><em>The real opportunity isn’t to automate human judgment. It’s to create more room for it.</em></strong></blockquote><h4>There Is Also a Trust Question</h4><p>Let’s say we build an AI assistant that produces a weekly summary. It looks polished. The language is confident. The formatting is excellent. Everyone loves how professional it looks. <br>But one important dependency is missing. A decision is described as finalized when it is still under discussion. <br>A risk is understated because the underlying work item wasn’t updated.</p><p>Now what?</p><p>The report may look better than the manually prepared version. But is it more trustworthy? <br>This is why I think AI adoption requires more than measuring time saved.</p><p>We also need to understand the reliability of the output. Can people trace important statements back to their sources? Can they identify what the system inferred versus what was explicitly recorded? <br>Is there a review process for high-impact information? Who owns the final communication? And what happens when the system is wrong?</p><p>Because in complex work, <strong>a polished wrong answer can be more dangerous than an unfinished honest one. </strong>That is something worth remembering as AI-generated information becomes more common in our everyday workflows.</p><h4>The Question I Would Ask Before Approving an AI Use Case</h4><p>If someone brought me an idea for a new AI workflow tomorrow, I wouldn’t begin by asking which model or platform they planned to use.</p><p>I would begin with a few simpler questions. <br><strong>What is frustrating people today? </strong>Not theoretically. What is actually happening?<br><strong>How are we solving it right now? </strong>What steps do people take? Where does time get consumed?<br><strong>Why does this problem exist? </strong>Is it repetitive effort, unclear ownership, missing information, poor process design or a genuine need for better intelligence?<br><strong>What should become better? </strong>Less manual effort? Faster decisions? Fewer errors? Better visibility?<br><strong>How will we know? </strong>What evidence would convince us that the solution helped?</p><p>And finally:<br><strong>What still needs a human? </strong>That last question matters more than it might initially appear.</p><p>Because good automation should make responsibilities clearer, not more ambiguous.</p><h4>Maybe We’re Asking the Wrong First Question</h4><p>I don’t think organizations should stop asking where AI can help. That’s an important conversation. But perhaps it shouldn’t always be the first one.</p><p>Instead of beginning with — <strong>“Where can we use AI?” <br></strong>What if we started with — <strong>“Where is work unnecessarily difficult?”<br></strong>Where are people spending time without creating proportional value? Where are decisions delayed because information is fragmented? Where do the same problems keep appearing? Where are people doing work that a better process could eliminate? Where would earlier insight change the outcome?</p><p>And only then — <strong>“Could AI help us solve this?” </strong>Sometimes the answer will be yes. Sometimes a simple automation will do. Sometimes a process change will be enough. Sometimes better ownership will solve the problem.<br>And sometimes the smartest decision will be to leave the process alone.</p><p>That’s not resistance to innovation. That’s being intentional about it.</p><h4>A Small Experiment to Try</h4><p>Here’s something I’d encourage you to try with your team.<br>In your next retrospective, improvement discussion or planning conversation, don’t begin by asking for AI ideas. <br>Ask everyone to complete this sentence — <strong>“I wish I didn’t have to spend so much time on ______.”</strong></p><p>Let people talk. You might hear about reports. Follow-ups. Data reconciliation. Meeting preparation. Searching for information. Repeated clarifications. Manual validation. Now choose one frustration. Not necessarily the biggest. <br>Just one that is frequent, understandable and reasonably safe to experiment with.</p><p>Before proposing a solution, ask — <strong>Why are we doing this work in the first place? What part of it genuinely requires human thinking? What part is repetitive? What would improve if we removed that repetition?</strong></p><p>Then explore whether AI is the right tool. You may discover a useful automation. You may discover a broken process. Or you may discover that the work never needed to exist.</p><p>All three are valuable outcomes.</p><h4>The Real Transformation May Be in How We Think</h4><p>AI is moving quickly. The tools will change. The models will improve. Capabilities that feel impressive today may become ordinary tomorrow.</p><p>And that’s exciting.</p><p>But I suspect the organizations that benefit most won’t necessarily be the ones with the largest number of AI experiments. <br>They’ll be the ones that become better at recognizing meaningful problems. The ones that understand their workflows. The ones that question unnecessary complexity. The ones that know where human judgment matters. The ones that measure improvement rather than simply adoption.</p><p>Because at the end of the day, nobody benefits from an AI solution simply because it contains AI.</p><p>They benefit because something becomes easier. Clearer. Faster. More reliable. More useful. Or simply less frustrating.<br>And perhaps that’s the question we should keep coming back to: <strong><em>If we removed the words “AI-powered” from this solution, would we still be excited about the problem it solves?</em></strong></p><p>If the answer is yes, we may be onto something meaningful. Because the goal was never to find more work for AI.</p><p><strong>The goal was to make work better for people.</strong></p><p><a href="https://medium.com/agile-true-north"><em>Agile True North</em></a><em> is about the lessons that rarely fit neatly into frameworks — the ones we learn while working with real teams, real constraints, and real people.</em></p><p><em>Something to think about: If your team could automate one frustrating part of its day tomorrow, what would you choose? <br>And would automating it actually solve the problem — or just make the problem easier to live with?</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=438817ab9b94" width="1" height="1" alt=""><hr><p><a href="https://medium.com/agile-true-north/we-asked-for-ai-but-what-problem-were-we-trying-to-solve-438817ab9b94">We Asked for AI. But What Problem Were We Trying to Solve?</a> was originally published in <a href="https://medium.com/agile-true-north">Agile True North</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Work Is Moving. The Outcome Isn’t.]]></title>
            <link>https://medium.com/agile-true-north/the-work-is-moving-the-outcome-isnt-dcb8562b8a80?source=rss----e69a8c7e233d---4</link>
            <guid isPermaLink="false">https://medium.com/p/dcb8562b8a80</guid>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[performance]]></category>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[agile]]></category>
            <dc:creator><![CDATA[Soumya]]></dc:creator>
            <pubDate>Thu, 29 Oct 2026 15:26:01 GMT</pubDate>
            <atom:updated>2026-10-29T15:26:01.593Z</atom:updated>
            <content:encoded><![CDATA[<h4>A full backlog, healthy velocity, and a busy team can still produce surprisingly little change.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*YA6IuVk1BDhhUCHP" /><figcaption>Photo by <a href="https://unsplash.com/@getswello?utm_source=medium&amp;utm_medium=referral">Swello</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>The dashboard looked good. Stories were closing. Velocity was steady. The sprint board was moving. <br>People were busy. Very busy. <br>Stand-ups were happening. Reviews were happening. <br>Status updates were full of accomplishments. Nobody could reasonably say the team wasn’t working hard.</p><p>And yet, somewhere in the middle of one of those conversations, a much simpler question surfaced: <strong>What actually became better because of all this work?</strong></p><p>The room got quieter. Not because nothing had been delivered. A lot had been delivered. But suddenly we weren’t talking about stories, epics, milestones or percentages anymore. <br>We were talking about <strong>outcomes</strong>. And that’s a very different conversation.</p><h4>Activity Is Easy to See</h4><p>One reason we gravitate toward activity is that it’s wonderfully visible. <br>We can count it. Stories completed. Features released. Tickets closed. Milestones achieved. Sprint velocity. Hours spent.<br>Projects launched. Dashboards love these things.</p><p>They move. They change colors. They give us something concrete to report. Outcomes are messier.</p><p>Did customers actually use what we built? Did the process become faster?Did errors decrease? Did the business make better decisions? Did we remove the problem we originally set out to solve? <br>Did anyone’s life actually become easier?</p><p>Those questions take longer to answer. Sometimes the answer isn’t immediately available. Sometimes the answer isn’t flattering.</p><p>So without realizing it, organizations can become very good at measuring <strong>the work</strong> and surprisingly weak at measuring <strong>what the work changed</strong>.</p><h4>We Delivered It. Did It Work?</h4><p>This is one question I wish we asked more often. <br>A feature goes live. Celebration. <br>An initiative reaches 100%. Green.<br>The project closes. Success.</p><p>Then everyone moves to the next priority. But what happened afterward? Did adoption happen? Did the expected efficiency appear? Did the manual process actually disappear? Did users behave differently?<br>Did the business metric move? Did the team solve the original problem?</p><p>Sometimes we don’t know. And that should make us uncomfortable.</p><p>Because: <strong><em>Delivery completion tells us that we finished building something.<br>It doesn’t tell us whether building it was worthwhile.</em></strong></p><h4>The Original Problem Can Disappear Behind the Solution</h4><p>I’ve seen this happen more than once. <br>A problem is identified. Let’s say a process takes too long.<br>A solution is proposed. Maybe it’s a dashboard.<br>An automation. A new workflow. A platform enhancement. A new data product.<br>Everyone gets busy designing and delivering it. Requirements appear. Stories get written. Dependencies emerge.<br>Plans get created. Meetings multiply. Months pass.<br>Eventually the solution itself becomes the center of the conversation.</p><p><strong>Are we on schedule? How many stories are left? Is testing complete? When can we launch?</strong></p><p>All reasonable questions. But somewhere along the way, the original question can quietly disappear: <strong><em>“Did we actually make that slow process faster?”</em></strong></p><p>That’s the question the work existed to answer. Everything else was supposed to be in service of it.</p><h4>A Perfectly Delivered Solution Can Still Solve the Wrong Problem</h4><p>This may be one of the more uncomfortable lessons in delivery. We can execute beautifully against the wrong assumption.</p><p>On time. On budget. Within scope. Great documentation. Excellent governance. <br>And still create something that doesn’t meaningfully improve the outcome. That doesn’t necessarily mean anyone failed.</p><p>Sometimes we learned something. Sometimes customer behavior was different from what we expected. Sometimes the environment changed. Sometimes the original hypothesis was simply wrong.</p><p>The danger isn’t being wrong. The danger is <strong>not checking</strong>.</p><p>If success is defined entirely by delivery, we can declare victory before reality has had a chance to respond.</p><h4>The Dashboard Can Accidentally Reinforce This</h4><p>Imagine a leadership dashboard showing:<br><strong>Initiative A — 90% complete<br>Initiative B — Green<br>Initiative C — 34 of 40 stories delivered</strong></p><p>It looks reassuring.</p><p>Now imagine another dashboard asking:<br><strong>Initiative A — Expected to reduce processing time by 30%. Current change: 4%.<br>Initiative B — Target adoption: 70%. Current adoption: 18%.<br>Initiative C — Expected to eliminate 500 manual hours/month. Current reduction: 120 hours.</strong></p><p>Suddenly we are having a very different conversation.</p><p>The first dashboard asks: <strong>Are we doing the work?</strong></p><p>The second asks: <strong>Is the work doing what we expected?</strong></p><p>We probably need both.</p><h4>“We Need a Dashboard” Is Not an Outcome</h4><p>This is one I’ve become increasingly sensitive to.</p><p>Someone says: <strong>“We need a dashboard.”</strong></p><p>Okay. Why?</p><p><em>“Leadership needs visibility.” </em>What visibility?</p><p><em>“They need to understand performance.” </em>What will they do differently once they understand it?</p><p>Now we’re getting somewhere. <br>Maybe the real need is: <strong>Leaders need to identify underperforming regions early enough to intervene before the quarter closes.</strong></p><p>That’s an outcome. The dashboard is simply one possible solution.<br>This distinction matters because once we fall in love with the solution too early, we stop exploring the problem.</p><p>The same applies to:<br><strong>“We need AI.”<br>“We need automation.”<br>“We need a new platform.” <br>“We need another report.”</strong></p><p>Maybe. But first — <strong><em>What should become different if we do this successfully?</em></strong></p><p>That question has saved me from many solution-first conversations.</p><h4>AI Makes This Question Even More Important</h4><p>Right now, organizations are understandably excited about AI and automation. And there is enormous opportunity there. <br>I am excited about it too.<br>But I’ve also started asking a question whenever we talk about automating something: <strong>What are we giving people back?</strong></p><p>Time? Better information? Faster decisions? Fewer errors?</p><p>Less repetitive work? Earlier visibility into risk?</p><p>If we automate a report that took someone three hours every week, that’s useful. But the deeper value isn’t: <strong>“The report is automated.”</strong></p><p>The value is: <strong>“We returned three hours of human attention every week.”</strong></p><p>Then comes the next question: <strong>What will we do with that attention?</strong></p><p>Better analysis? More time with customers? Earlier problem-solving? More thoughtful decisions?</p><p>Automation is most powerful when we don’t just automate the task.</p><p>We understand <strong>what human capacity the task was consuming and where that capacity can create more value instead.</strong></p><h4>Busy Teams Can Hide Weak Prioritization</h4><p>There is another uncomfortable possibility. Sometimes everyone is busy precisely because we haven’t made enough choices.</p><p>Ten initiatives are active. All ten have meetings. All ten have plans. All ten have stakeholders.<br>All ten need status updates. Everyone is moving between them.</p><p>The organization looks incredibly active. But very little finishes. This connects to something I’ve learned repeatedly: <strong><em>Starting work creates activity. Finishing meaningful work creates value.</em></strong></p><p>Those aren’t the same thing.</p><p>Sometimes the fastest way to create more progress isn’t to work faster. It’s to work on fewer things.</p><h4>Output Still Matters</h4><p>I don’t want to swing too far in the other direction. <br>Stories matter. Milestones matter. Delivery dates matter. Teams need operational measures.</p><p>You cannot manage complex work by asking everyone to think only about some distant business outcome.</p><p>Outputs are how outcomes get created.</p><p>The problem begins when <strong>output becomes the definition of success rather than evidence that we’re progressing toward it.</strong></p><p>A team can control whether it builds something well. It may not fully control whether customers adopt it. <br>That’s why outcomes should not become another weapon used to judge teams.</p><p>They should become a way to learn.</p><p><strong>We expected X. We delivered Y.</strong></p><p><strong>Reality produced Z. What did we learn?</strong></p><p>That’s a much healthier conversation.</p><h4>Maybe Success Has Three Questions</h4><p>I’ve started thinking about initiatives through three simple questions.</p><p><strong>1. Did we deliver what we said we would? — </strong>That’s execution.</p><p><strong>2. Did people actually use or adopt what we delivered? — </strong>That’s behavior.</p><p><strong>3. Did that behavior create the change we expected? — </strong>That’s outcome.</p><p>Consider an automation initiative.</p><p><strong>Execution: </strong>The workflow was built and deployed.<br><strong>Behavior: </strong>Teams actually started using it.<br><strong>Outcome: </strong>Manual effort dropped and decisions happened faster.</p><p>Or a new reporting capability.</p><p><strong>Execution: </strong>The dashboard launched.<br><strong>Behavior: </strong>Leaders began using it regularly.<br><strong>Outcome: </strong>Problems were identified earlier and intervention happened sooner.</p><p>Those three questions tell a much richer story than: <br><strong>Project status — Complete.</strong></p><h4>There Is Another Question We Rarely Ask</h4><p>Sometimes an initiative delivers exactly the outcome we wanted.</p><p>Fantastic.!!!</p><p>But then I think there is one more question worth asking: <strong><em>“Was the outcome worth what it cost us?”</em></strong></p><p>Not only money. Time. Attention. Opportunity cost. Complexity. Maintenance. Technical debt. Change fatigue.</p><p>Every initiative consumes organizational energy.</p><p>If something produces a 2% improvement but required six teams for nine months, perhaps we should understand that too.</p><p>Success isn’t simply: <strong>Did the metric move? <br></strong>It’s also: <strong>Was this the best use of our limited capacity?</strong></p><p>That question moves us beyond delivery management and into actual decision-making.</p><h4>What Would We Stop If the Outcome Didn’t Move?</h4><p>Here’s a question that can be surprisingly revealing. <br>Imagine we’ve delivered 70% of an initiative. We’ve learned enough to see that the expected outcome isn’t appearing. Would we stop?<br>Would we change direction? Would we reduce the remaining investment?</p><p>Or would we finish because: <strong>“We’ve already come this far.”</strong></p><p>Organizations can become emotionally attached to completion. Finishing feels responsible.Stopping feels like failure.</p><p>But sometimes continuing to invest in something simply because we’ve already invested heavily is the least Agile decision available.</p><p>Learning should be allowed to change the plan. Otherwise we’re not adapting. We’re just executing.</p><h3>The Most Valuable Question May Come After “Done”</h3><p>Agile teams spend a lot of time defining Done. That’s important. But maybe there is another question worth adding:<strong><em>“What will tell us this actually worked?”</em></strong></p><p>Not: When is it deployed?</p><p>Not: When does the epic close?</p><p>Not: When do we reach 100%?</p><p>But: <strong>What evidence would convince us that the problem is actually better?</strong></p><p>If we can’t answer that before we start, we may have a harder time recognizing success when we reach it.</p><p>Or recognizing failure when we don’t.</p><h4>A Small Experiment to Try</h4><p>Pick one major initiative you’re working on right now. <br>Ignore its status for a moment. Ignore the percentage complete. Ignore how many stories are closed.</p><p>Ask: <strong>What problem were we originally trying to solve?</strong></p><p>Then: <strong>What did we expect to become different?</strong></p><p>Then: <strong>What evidence would tell us that change actually happened?</strong></p><p>And finally: <strong><em>If we delivered everything currently planned and that outcome didn’t change, would we still call this successful?</em></strong></p><p>Sit with that question for a moment. Because sometimes the answer may be uncomfortable. And that’s useful.</p><h4>Motion Is Not the Enemy</h4><p>Teams need motion. Work has to move. Things need to get built. Plans need execution.<br>But movement should eventually take us somewhere.</p><p>The goal isn’t fewer stories. Fewer meetings. Fewer dashboards. Or fewer projects. <br>The goal is to maintain a clear line between: <strong>the problem we wanted to solve, the work we chose to do, </strong>and <strong>the change we expected that work to create.</strong></p><p>When that line disappears, activity can become incredibly convincing.<br>Everyone is working. Everything is moving. The dashboard looks healthy.</p><p>And yet the thing that mattered most hasn’t changed. That’s when I think we need the courage to stop asking: <strong>“How much have we completed?”</strong></p><p>and ask the harder question: <strong><em>“What is different because we did?”</em></strong></p><p><a href="https://medium.com/agile-true-north"><em>Agile True North</em></a><em> is about the lessons that rarely fit neatly into frameworks — the ones we learn while working with real teams, real constraints, and real people.</em></p><p><em>Something to think about: Pick one initiative your team recently marked “Done.” If you removed the project plan, the Jira board, and the status dashboard from the conversation — what evidence would you use to prove that it actually worked?</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dcb8562b8a80" width="1" height="1" alt=""><hr><p><a href="https://medium.com/agile-true-north/the-work-is-moving-the-outcome-isnt-dcb8562b8a80">The Work Is Moving. The Outcome Isn’t.</a> was originally published in <a href="https://medium.com/agile-true-north">Agile True North</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Dashboard Is Green. I Don’t Believe It.]]></title>
            <link>https://medium.com/agile-true-north/the-dashboard-is-green-i-dont-believe-it-db96630fb126?source=rss----e69a8c7e233d---4</link>
            <guid isPermaLink="false">https://medium.com/p/db96630fb126</guid>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[project-management]]></category>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[analytics]]></category>
            <dc:creator><![CDATA[Soumya]]></dc:creator>
            <pubDate>Mon, 26 Oct 2026 16:36:01 GMT</pubDate>
            <atom:updated>2026-10-26T16:36:01.474Z</atom:updated>
            <content:encoded><![CDATA[<h4>Sometimes the first sign that something is wrong isn’t a red status. It’s a pattern nobody has called out yet.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*OY0bFEy_V-uQKpQc" /><figcaption>Photo by <a href="https://unsplash.com/@plutonicmedia?utm_source=medium&amp;utm_medium=referral">Loui Kiær</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Everything was green. Milestones — green. Sprint health — green. <br>Dependencies — apparently under control. <br>No major blockers. No escalation required.</p><p>If you looked at the dashboard, there wasn’t much to worry about.</p><p>But something didn’t feel right.</p><p>The same work had appeared in multiple updates. A dependency had been described as <em>“almost resolved”</em> for longer than I was comfortable with. <br>A decision kept moving from one meeting to the next.</p><p>The team was delivering, but more work seemed to be carrying over. Nobody had declared a major risk.</p><p>Yet all the small signals were beginning to tell a different story. The dashboard said green. <strong>I wasn’t convinced.</strong></p><h4>Green Is a Status. Health Is a Story.</h4><p>I’ve always valued dashboards. They bring structure to messy programs. They help us see progress, milestones, capacity, risks and dependencies without sitting through hours of status conversations.</p><p>But over time, I’ve learned something important —</p><blockquote><strong><em>A dashboard can tell you what the status is.<br>It can’t always tell you what the status is becoming.</em></strong></blockquote><p>That’s a very different question.</p><p>A milestone can still be green today while the conditions required to deliver it are slowly deteriorating.</p><p>A dependency can technically be <strong><em>“on track”</em> </strong>while every conversation about it becomes less certain.</p><p>A team can hit its sprint commitment while quietly accumulating work that keeps getting pushed forward.</p><p>A decision can remain open without affecting today’s status — until suddenly it affects everything.</p><p>The dashboard isn’t necessarily wrong. It just hasn’t caught up yet.</p><h4>Risk Usually Whispers Before It Shouts</h4><p>We tend to notice risk when something becomes obvious.</p><p>A milestone slips. A dependency misses its date.</p><p>A blocker becomes critical. Capacity disappears.</p><p>A release moves. Now the status turns amber or red.</p><p>But by that point, the risk may have been developing for weeks.</p><p>I’ve learned to pay attention to the whispers.</p><p><em>“We should know by next week.”</em></p><p><em>“We’re still waiting for confirmation.”</em></p><p><em>“It should be okay.”</em></p><p><em>“We carried it over, but we’re expecting to finish this sprint.”</em></p><p><em>“The other team hasn’t committed yet, but they know it’s important.”</em></p><p><em>“We haven’t made the decision, but there’s still time.”</em></p><p>None of these statements automatically means something is wrong. But when the same sentences keep appearing week after week, they stop being updates.</p><p><strong>They become signals.</strong></p><h4>One Delay Is an Event. Repetition Is a Pattern.</h4><p>This distinction has become increasingly important to me. <br>One story carrying over? It happens.<br>One dependency slipping? Programs are messy.<br>One decision taking longer than expected? Normal.</p><p>But when several things repeatedly behave the same way, I start paying attention. Because sometimes the most useful question isn’t: <strong>“What is blocked?”</strong></p><p>It’s — <strong><em>“What keeps almost becoming blocked?”</em></strong></p><p>That’s where interesting things begin to surface.</p><p>Maybe one team is consistently waiting on another. <br>Maybe work repeatedly enters a sprint before dependencies are ready.<br>Maybe the same type of requirement keeps getting clarified late.<br>Maybe decisions consistently take two weeks longer than the plan assumes.<br>Maybe the team has enough capacity on paper, but interruptions are consuming it.</p><p>None of these may trigger a red status individually. Together, they tell a story.</p><h4>I Pay Attention to Aging</h4><p>One of the simplest signals I have found useful is <strong>age</strong>.</p><p>How long has this dependency been open? How long has this decision been waiting? <br>How long has this work item remained in the same state? How many updates has this risk appeared in?</p><p>How many times have we said -<em>“We’re following up.”</em></p><p>Aging matters because unresolved work doesn’t always look dramatic. Sometimes it just sits there. Quietly.<br>Until the remaining schedule becomes smaller than the time required to solve it.</p><p>And suddenly everyone asks — <strong>“How did this become urgent?”</strong></p><p>It didn’t suddenly become urgent. It was slowly becoming urgent while we were still calling it green.</p><h4>Listen to the Language Around the Work</h4><p>There are signals dashboards don’t capture very well. <br>Language is one of them. <br>Compare — <strong>“Team X confirmed delivery for October 12.”</strong></p><p>with — <strong>“Team X is targeting October 12.”</strong></p><p>with — <strong>“We’re hoping Team X can still make October 12.”</strong></p><p>The date hasn’t changed. The status might not have changed. But the confidence certainly has. <br>I’ve learned to listen for words like: <em>hopefully. probably. should. expecting. waiting. trying.</em></p><p>Again, none of these words is inherently a problem. <br>But when certainty keeps decreasing while the status stays the same, it’s worth asking another question.</p><h4>Sometimes the Team Tells You Without Telling You</h4><p>Another signal I’ve learned not to ignore is the conversation itself.</p><p>When a team that normally discusses problems openly suddenly becomes very quiet, I notice.<br>When the same blocker gets a very short update every week, I notice.<br>When someone says — <em>“We’ll figure it out.” </em>but can’t explain how, I notice.</p><p>When people start working around a dependency rather than resolving it, I notice.<br>When conversations increasingly happen after the meeting instead of during it, I notice.<br>Not because any of these proves something is wrong.<br>But because <strong>delivery health lives in conversations as much as it lives in data.<br></strong>Sometimes people sense risk before the reporting system does.</p><h4>The Goal Isn’t to Turn Everything Amber</h4><p>There is a danger on the other side too. If we treat every uncertainty as a risk, dashboards become useless.</p><p>Everything turns amber. People stop paying attention. Teams feel like they’re constantly being escalated. That’s not the answer. <br>The goal isn’t to become suspicious of every green status.<br>The goal is to ask whether the evidence underneath the status still supports it.</p><p>That distinction matters.</p><p>Instead of saying — <strong>“I don’t believe this is green.”</strong></p><p>I might ask — <strong>“What needs to remain true for this to stay green?”</strong></p><p>That’s a much better conversation.</p><p>Maybe the answer is — Team X must deliver the API by Friday. The architecture decision must happen this week. <br>Two open stories cannot carry over again. We need the additional engineer starting Monday. <br>Now we have something useful. We haven’t changed the status. <br>We’ve identified the <strong>conditions supporting it</strong>.</p><h4>That Changed How I Think About Dashboards</h4><p>For a long time, I thought a good dashboard should answer: <strong>Where are we?</strong>Now I think the better dashboards also help us ask: <strong>Where are we heading?</strong>That means looking beyond static status.</p><p>Not just: How many items are blocked? <br>But: <strong>How long have they been blocked?</strong></p><p>Not just: How many dependencies are open?<br>But: <strong>Which ones are getting older without resolution?</strong></p><p>Not just: What did we deliver this sprint?<br>But: <strong>What keeps carrying over?</strong></p><p>Not just: Do we have capacity?<br>But: <strong>Is demand consistently exceeding it?</strong></p><p>Not just: Is the milestone green?<br>But: <strong>What assumptions need to hold for it to remain green?</strong></p><p>That is when reporting starts becoming something more useful. It becomes <strong>delivery intelligence.</strong></p><h4>Data Can Surface the Signal. Judgment Still Matters.</h4><p>Today, we have more ways than ever to surface these patterns. Dashboards can highlight aging work. Automation can flag dependencies that haven’t moved.</p><p>Workflow data can reveal repeated carryover.</p><p>AI can summarize risks across dozens of work items and identify patterns that would take someone hours to find manually. <br>I find all of that incredibly useful. <br>But I don’t think the goal is to automate judgment. The goal is to give judgment <strong>better signals</strong>.</p><p>A system might tell me — <strong>This dependency has been open for 19 days.</strong></p><p>It still takes a person to ask: <strong>Why? Does it matter? What does it affect?<br>Who needs to act? And what happens if nothing changes?</strong></p><p>That’s the part I don’t want automation to replace. That’s the part I want automation to give us more time to do.</p><h4>A Small Experiment to Try</h4><p>The next time you look at a perfectly green dashboard, don’t immediately ask: <strong>“Are we on track?”</strong></p><p>Try these instead:</p><ul><li><strong>What has been open longer than expected?</strong></li><li><strong>What keeps carrying over?</strong></li><li><strong>Which decision are we assuming will happen in time?</strong></li><li><strong>Which dependency has no firm commitment behind it?</strong></li><li><strong>Where has confidence changed even though status hasn’t?</strong></li></ul><p>And finally — <strong><em>What needs to remain true for this program to stay green?”</em></strong></p><p>That last question can change the entire conversation. Because good delivery leadership isn’t about predicting every problem. <br>And it certainly isn’t about turning everything red before something goes wrong.</p><p>It’s about noticing when small signals begin forming a pattern — and creating the right conversation while there is still time to do something about it.</p><p>The dashboard may still be green. Maybe it should be.</p><p>But sometimes the most valuable thing we can bring to the room is the curiosity to ask:</p><p><strong>“What is this green status not telling us yet?”</strong></p><p><a href="https://medium.com/agile-true-north"><em>Agile True North</em></a><em> is about the lessons that rarely fit neatly into frameworks — the ones we learn while working with real teams, real constraints, and real people.</em></p><p><em>What is your early warning signal — the thing that makes you look twice even when everything still appears to be on track?</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=db96630fb126" width="1" height="1" alt=""><hr><p><a href="https://medium.com/agile-true-north/the-dashboard-is-green-i-dont-believe-it-db96630fb126">The Dashboard Is Green. I Don’t Believe It.</a> was originally published in <a href="https://medium.com/agile-true-north">Agile True North</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The API Was Done. The Program Wasn’t.]]></title>
            <link>https://medium.com/agile-true-north/the-api-was-done-the-program-wasnt-772edc79b3eb?source=rss----e69a8c7e233d---4</link>
            <guid isPermaLink="false">https://medium.com/p/772edc79b3eb</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[program-management]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[leadership]]></category>
            <dc:creator><![CDATA[Soumya]]></dc:creator>
            <pubDate>Fri, 23 Oct 2026 15:56:01 GMT</pubDate>
            <atom:updated>2026-10-23T15:56:01.649Z</atom:updated>
            <content:encoded><![CDATA[<h4>Ownership doesn’t live inside a team’s boundary — it lives in the gaps between them</h4><p>The API shipped on schedule.<br>Code reviewed, tests passing, deployed to production. The engineering team closed the ticket and moved on.</p><p>Three weeks later, a downstream team’s dashboard started showing broken numbers. <br>Nobody had told them the response format had changed.<br>Nothing was technically wrong. <br>Nobody had lied about status. <br><strong>And the program still broke.</strong></p><h4>The Question Nobody Asked</h4><p>At one of my previous companies, I watched this happen during a platform integration effort — new API, clean implementation, green checkmark in every tracker.</p><p>What never got asked in any review: <strong><em>who else is calling this today, and do they know it’s changing?</em></strong></p><p>The team that owned the API didn’t own that answer. <br>Neither did I, at first. It fell into the space between teams — the space nobody’s ticket covers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*RdrEnFyIw6xF-QsD" /><figcaption>Photo by <a href="https://unsplash.com/@kaleidico?utm_source=medium&amp;utm_medium=referral">Kaleidico</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>Why This Keeps Happening</h4><p>An API contract is a promise. When you change it, you’re not just shipping code — you’re quietly asking every consumer of that API to trust you kept the promise, or to notice if you didn’t.</p><p>Most teams test that their own service works. Fewer teams trace who’s depending on it downstream. Almost nobody owns the conversation of <em>telling</em> those consumers before the change lands, not after it breaks something.</p><p>Part of this is structural. The team that built the API is measured on shipping the API. The team consuming it is measured on their own roadmap, not on watching someone else’s changelog. Nobody’s job description says “notice the gap between these two teams” — so the gap just sits there, unowned, until it breaks something visible enough that someone has to.</p><p>That gap isn’t a technical failure. It’s a visibility failure. And visibility across systems is exactly the job of whoever is holding the program together — whatever their title says.</p><h4>Getting the Room in One Place</h4><p>When the broken dashboard surfaced, the instinct in most orgs is to treat it as a bug — assign it to whoever’s on call, patch it, move on. That would have fixed the symptom and left the actual gap untouched, ready to open up again with the next change.</p><p>Instead, I pulled both teams into the same conversation — the ones who’d shipped the API change and the ones who’d been quietly consuming it. Not to assign blame. To rebuild the picture nobody had: who actually depends on this, and how would any of us know if that changes again.</p><p>That conversation surfaced things neither team had said out loud. The API team hadn’t known this particular dashboard existed downstream. The consuming team hadn’t known there was a lighter-weight way to flag “we’re touching this” before a change went out. Both gaps were fixable. Neither would have surfaced without putting the two teams in the same room.</p><p>We closed it out with a short, targeted retro — not the standard sprint retro, just this incident. Three questions drove it: how did this ship without anyone flagging the dependency, what would have caught it earlier, and what’s the one thing we change so it doesn’t happen the same way twice. We walked out with a lightweight consumer-notification step added to the release checklist — nothing heavy, just enough that “who’s calling this” became a question someone had to answer before shipping, not after something broke.</p><h4>What I Do Differently Now</h4><p>Before I call any integration “done,” I ask one blunt question: <em>who consumes this, and have they said yes?</em></p><p>Not “have they been notified.” Have they said yes. Silence isn’t sign-off.</p><p>I also stopped trusting a green status from the team that built the change to tell me anything about the teams that depend on it. Those are two different questions, and only one of them gets asked by default.</p><h4>The Real Lesson</h4><p>A component can be perfect and a program can still fail — because delivery doesn’t happen inside one team’s boundary. It happens in the handoffs between them.</p><p>Nobody owns those handoffs by default. Someone has to decide to own them — to notice the gap, pull the right people into a room, and turn a broken dashboard into a fixed process instead of a one-time patch. That’s not an engineering task. It’s a leadership one.</p><p>You don’t need to write the API to ask who’s calling it. You just need to ask before it breaks, not after — and when it does break anyway, treat it as a chance to close the gap for good, not just for now.</p><p>Who’s depending on the thing you just called done?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=772edc79b3eb" width="1" height="1" alt=""><hr><p><a href="https://medium.com/agile-true-north/the-api-was-done-the-program-wasnt-772edc79b3eb">The API Was Done. The Program Wasn’t.</a> was originally published in <a href="https://medium.com/agile-true-north">Agile True North</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[When Everything Is a Priority, the Team Quietly Chooses for You]]></title>
            <link>https://medium.com/agile-true-north/when-everything-is-a-priority-the-team-quietly-chooses-for-you-8ac0e2e86db1?source=rss----e69a8c7e233d---4</link>
            <guid isPermaLink="false">https://medium.com/p/8ac0e2e86db1</guid>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[project-management]]></category>
            <category><![CDATA[decision-making]]></category>
            <dc:creator><![CDATA[Soumya]]></dc:creator>
            <pubDate>Mon, 19 Oct 2026 16:46:01 GMT</pubDate>
            <atom:updated>2026-10-19T16:46:01.849Z</atom:updated>
            <content:encoded><![CDATA[<h4>Prioritization doesn’t disappear when leaders avoid making trade-offs. It simply moves somewhere less visible.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*9fr8Q2AGE1eNtij-" /><figcaption>Photo by <a href="https://unsplash.com/@airfocus?utm_source=medium&amp;utm_medium=referral">airfocus</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>We’ve probably all heard some version of this conversation.</p><p><strong>“Which one is the priority?” </strong><em>“They’re all important.”</em></p><p><strong>“If we have to choose?” </strong><em>“We really need all of them.”</em></p><p><strong>“Do any of the dates have flexibility?” </strong><em>“Not really.”</em></p><p>And then comes the sentence that somehow ends the discussion: <strong>“Let’s see how much we can get done.”</strong></p><p>The meeting ends. Nothing has been deprioritized. No date has moved. No scope has changed. No additional capacity has appeared.</p><p>On paper, everything remains a priority. But something interesting happens next. <strong>The team prioritizes anyway. </strong>They just do it quietly.</p><h4>Work Cannot Execute a Strategy That Refuses to Choose</h4><p>This is something I’ve come to believe strongly. You can put five initiatives at the top of a roadmap.<br>You can label all five P0. You can give all five executive sponsorship. You can tell everyone they’re equally important.</p><p>But if the same people are required to deliver them at the same time, reality eventually makes the decision for you.</p><p>Someone works on Initiative A before Initiative B. A review for C waits because D has a deadline tomorrow. Someone postpones E because the stakeholder for A is asking for updates every day.</p><p>Prioritization still happened. It just wasn’t deliberate. And that’s the part that worries me.</p><blockquote><strong><em>When we refuse to make trade-offs explicitly, we don’t eliminate the trade-off.<br>We delegate it to the people closest to the work.</em></strong></blockquote><p>Often without telling them we’ve done it.</p><h4>The Team Creates Its Own Prioritization System</h4><p>When everything is important, people need some way to decide what to do first. So an unofficial system emerges.</p><p>Maybe the loudest stakeholder wins. Maybe the closest deadline wins. Maybe the person who follows up most often wins. Maybe the executive-sponsored initiative wins. Maybe the easiest work gets completed first because at least something can move to Done. Maybe the most uncomfortable conversation gets avoided until tomorrow.</p><p>None of this necessarily happens intentionally.</p><p>People are simply trying to navigate conflicting signals. But eventually the actual priority order may look nothing like the official one.</p><p>The roadmap says:</p><p><strong>A = Critical<br>B = Critical<br>C = Critical<br>D = Critical</strong></p><p>Reality says:</p><p><strong>A = whoever called this morning.<br>B = whoever has the nearest deadline.<br>C = whoever escalated last week.<br>D = we’ll get to it when we can.</strong></p><p>That’s not prioritization. That’s survival.</p><h4>“Just Do Your Best” Can Be an Expensive Instruction</h4><p>There are moments when asking people to stretch is completely reasonable.<br>A critical incident. An important launch. A temporary surge.</p><p>Teams can accomplish extraordinary things when there is a clear reason and a defined period of intensity. <br>But temporary stretch and permanent overload are very different things. <br>If demand consistently exceeds capacity, telling the team to <em>“do your best”</em> doesn’t solve the mismatch. It simply makes the mismatch personal.</p><p>Now individuals start asking themselves:</p><ul><li><strong>Am I moving fast enough?</strong></li><li><strong>Should I work later?</strong></li><li><strong>Should I skip this meeting and finish that task?</strong></li><li><strong>Which stakeholder can I afford to disappoint?</strong></li><li><strong>Why am I still behind when I’ve been working all day?</strong></li></ul><p>A structural problem slowly starts feeling like an individual performance problem. That’s dangerous. Because the team may not actually have an execution problem.<br>They may have a <strong>choice problem that nobody upstream has been willing to resolve.</strong></p><h4>Context Switching Has a Cost We Rarely Put on the Roadmap</h4><p>One of the easiest ways to make everything look achievable is to divide capacity mathematically.</p><p>We have ten people.<br>Project A gets 30%. Project B gets 25%. Project C gets 25%. Operational work gets 20%. Perfect.!!<br>The spreadsheet balances. Reality doesn’t work quite so neatly.</p><p>People don’t switch from one complex problem to another with zero friction. There is context to reload. Conversations to remember. Documents to reopen. Decisions to reconstruct. Code, data, requirements or workflows to understand again.</p><p>And there is another cost that is harder to quantify: <strong>attention.</strong></p><p>You can allocate 25% of someone’s time. It’s much harder to allocate 25% of their brain. <br>When people are constantly moving between priorities, sometimes everything progresses. But nothing really moves.</p><h4>Capacity Planning Cannot Fix a Prioritization Problem</h4><p>I’ve spent a lot of time around capacity conversations. <br>How much work can the team take? Who is available? Where are the gaps? What can fit into this quarter?</p><p>Those are useful questions. But capacity planning has a limit.</p><p>If the organization wants 140 units of work from 100 units of capacity, there are only a few real choices.</p><p>Reduce demand. Increase capacity. Move dates. Reduce scope. Accept risk. Improve the way the work is done. Usually some combination of those.</p><p>What we cannot do is keep 140 units of demand, keep 100 units of capacity, keep every date, keep every scope commitment…</p><p>…and solve the remaining 40 units with optimism.</p><p>Yet organizations do versions of this all the time.</p><p>The spreadsheet gets adjusted. Percentages move. Names move between rows. And eventually the numbers fit. But the work doesn’t.</p><h4>Sometimes “Priority” Actually Means “Important”</h4><p>This distinction has helped me in many conversations. A lot of things are important.<br>Customer commitments are important. Technical debt is important. Compliance is important. Reliability is important. New capabilities are important. Operational improvements are important.</p><p>That doesn’t mean they can all happen first.</p><blockquote><strong>Importance describes value. Priority describes order.</strong></blockquote><p>Those aren’t the same thing. <br>If everything is genuinely important, that makes prioritization more necessary — not less.<br>Because prioritization isn’t deciding which work matters and which doesn’t.</p><p>It’s deciding — <strong><em>Given our constraints, what matters first?</em></strong></p><p>That is a much harder conversation. Which is probably why we sometimes avoid it.</p><h4>“But We Can’t Deprioritize Any of Them”</h4><p>I’ve heard this many times. And sometimes it is true.<br>Maybe all five commitments really do need to happen. Then the conversation shouldn’t end there.<br>It should become: <strong>What has to change to make that possible?</strong></p><p>Do we add capacity? Change sequencing? Reduce scope? Move a date? Accept more risk? Stop something else?</p><p>Because if none of those things can change, then saying <em>“we need everything”</em> isn’t a plan.</p><p>It’s a statement of desire. And teams cannot execute desire. They execute choices.</p><h4>Watch What People Stop Doing</h4><p>One of the most interesting signals of overload isn’t what the team is working on. It’s what quietly disappears. <br>Documentation gets postponed. Technical debt doesn’t get addressed. <br>Retrospective actions remain open. <br>Improvements are deferred. People stop experimenting. <br>Knowledge sharing becomes optional. Planning becomes rushed. People prepare less for meetings.<br>Small quality issues get accepted because there isn’t time to investigate them properly. Nothing dramatic happens.</p><p>The team is still delivering. Maybe even delivering a lot.</p><p>But gradually, all the work that helps the system remain healthy gets sacrificed to the work that looks urgent. This is why overloaded teams can appear productive for quite a long time.</p><p>Until the accumulated cost becomes visible.</p><h4>The Quietest People Often Absorb the Trade-Off</h4><p>There is a human side to this that dashboards rarely show. When priorities conflict, some people will raise the issue. Others won’t.</p><p>They’ll simply absorb it. They work later. They switch constantly. They try to satisfy every stakeholder. They don’t want to be seen as someone who “can’t handle it.”<br>And because they keep delivering, the system receives a dangerous signal: <strong>Apparently this workload is sustainable.</strong></p><p>It isn’t.</p><p>Someone is subsidizing the plan with their personal capacity. That is not capacity planning. And it isn’t something we should celebrate as commitment.</p><h4>Prioritization Is Not a Ranking Exercise</h4><p>We often treat prioritization as:</p><ol><li>Initiative A</li><li>Initiative B</li><li>Initiative C</li><li>Initiative D</li></ol><p>That’s useful.<br>But I think real prioritization requires another question — <strong><em>What are we willing not to do yet?</em></strong></p><p>That’s harder. Because putting A at #1 feels positive. Saying D will not start until next quarter feels like a commitment. <br>But that’s exactly why it matters. <br>A priority becomes meaningful only when it changes what happens to something else.</p><p>Otherwise, it’s just a label.</p><h4>A Question I’ve Started Asking Differently</h4><p>Instead of: <strong>“What is the highest priority?”<br></strong>I’ve found another question often creates a better conversation: <strong><em>“If we could only finish two of these this quarter, which two would we protect?”</em></strong></p><p>Suddenly the discussion changes. People start talking about outcomes. Customer impact. Revenue. Risk. Dependencies. Strategic commitments. Cost of delay.</p><p>What actually happens if something moves. We stop debating whether everything is important. Of course everything is important. <br>We start discussing <strong>relative importance under constraint. </strong>That’s where prioritization becomes real.</p><h4>And Sometimes the Answer Is Still “All of Them”</h4><p>Fine.</p><p>Then there is one more question:<strong><em>“What are we willing to change to make that true?”</em></strong></p><p>More people? Less scope? Different sequencing? Later dates? Higher risk tolerance?<br>Something has to give. And if the answer is: <strong>Nothing. <br></strong>Then reality will eventually choose for us.</p><h4>A Small Experiment to Try</h4><p>Look at the major priorities your team is carrying right now.</p><p>Don’t ask whether they’re important. Assume they are.</p><p>Instead ask: <strong>If two of these collide tomorrow, which one wins? Does everyone give the same answer?</strong></p><p>Then ask: <strong>What work are people quietly postponing to keep these commitments moving?</strong></p><p>And one more: <strong>Are we asking the team to make a trade-off that should actually be made somewhere else?</strong></p><p>That last question may be uncomfortable. But it is worth asking. <br>Because when leadership doesn’t choose, teams don’t suddenly gain infinite capacity.</p><p>They choose for us. Through delays. Through context switching. Through shortcuts. Through overtime. <br>Through whichever stakeholder calls first. Through the work that quietly never gets done. <br>The choice always happens. The only question is whether we make it <strong>deliberately and visibly</strong>…</p><p>or allow the people with the least authority to make it silently.</p><p><a href="https://medium.com/agile-true-north"><em>Agile True North</em></a><em> is about the lessons that rarely fit neatly into frameworks — the ones we learn while working with real teams, real constraints, and real people.</em></p><p><em>Something to think about: Look at your current priorities. If everything cannot happen at once, do you know what will move? More importantly — does the team know too?</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8ac0e2e86db1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/agile-true-north/when-everything-is-a-priority-the-team-quietly-chooses-for-you-8ac0e2e86db1">When Everything Is a Priority, the Team Quietly Chooses for You</a> was originally published in <a href="https://medium.com/agile-true-north">Agile True North</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Don’t Escalate the Problem. Escalate the Decision.]]></title>
            <link>https://medium.com/agile-true-north/dont-escalate-the-problem-escalate-the-decision-df168f64ca47?source=rss----e69a8c7e233d---4</link>
            <guid isPermaLink="false">https://medium.com/p/df168f64ca47</guid>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[decision-making]]></category>
            <category><![CDATA[project-management]]></category>
            <dc:creator><![CDATA[Soumya]]></dc:creator>
            <pubDate>Fri, 16 Oct 2026 15:21:01 GMT</pubDate>
            <atom:updated>2026-10-16T15:21:01.629Z</atom:updated>
            <content:encoded><![CDATA[<h4>The most useful escalation isn’t “We have a problem.” It’s “Here is the decision we can no longer make alone.”</h4><p>For a long time, I thought escalation meant something had gone wrong.</p><p>A team couldn’t resolve an issue. A dependency wasn’t moving. A stakeholder wasn’t responding. A deadline was at risk. <br>So the problem went <em>up</em>.</p><p>And if you were good at what you did, ideally you solved things before they ever reached that point. At least, that’s how escalation often felt.</p><p>Over time, I have come to see it very differently. Because sometimes the team has done everything right. <br>They’ve investigated. They’ve collaborated. They’ve followed up. <br>They’ve explored alternatives. They understand the problem perfectly well. And they still can’t move forward.</p><p>Not because they need more problem-solving. But because the next step requires a decision they don’t have the authority to make.</p><p>That’s when I started thinking:</p><blockquote><strong><em>Maybe we shouldn’t escalate problems at all.<br>Maybe we should escalate decisions.</em></strong></blockquote><h4>“We Have a Problem” Isn’t Enough</h4><p>Imagine receiving this update — <strong>“Team A is delayed and this may impact our release.”</strong></p><p>Important? Absolutely. <br>But what should the person receiving it actually do? Ask why? Call Team A?Schedule another meeting? Tell everyone to work faster?</p><p>Now imagine the same situation framed differently: <br><strong>“Team A’s dependency is expected two weeks later than planned. That moves integration testing beyond our current release window. We see three options — move the release, reduce the initial scope, or add a temporary workaround that introduces additional technical risk. <br>The team recommends reducing scope. We need a decision by Thursday to protect the current release date.”</strong></p><p>Same problem. Completely different conversation.<br>The first transfers <strong>anxiety</strong>. The second transfers <strong>a decision</strong>. <br>That distinction has become increasingly important to me.</p><h4>Sometimes We Escalate Too Early</h4><p>There is one side of this that is easy to recognize. Something gets difficult and immediately — <em>“Let’s escalate.”</em></p><p>But not every difficult conversation needs leadership intervention. Sometimes the team hasn’t spoken directly to the other team. Sometimes nobody has clarified ownership. Sometimes assumptions haven’t been validated. Sometimes there are perfectly reasonable options that haven’t been explored. <br>Escalating too early creates its own problems.<br>People begin solving issues through hierarchy instead of collaboration.</p><p>Teams become cautious with one another. Leaders get pulled into details that should have been resolved closer to the work. And over time, escalation becomes a substitute for problem-solving.</p><p>That’s not healthy. But there is another side that I think is much more subtle.</p><h4>Sometimes We Escalate Too Late</h4><p>I’ve seen this too. Everyone knows there is a problem. Nobody wants to escalate it.<br><em>“Let’s give them another day.” “They’re working on it.”<br>“I don’t want this to look like we’re blaming them.” <br>“Let’s try to resolve it ourselves.”</em></p><p>All reasonable thoughts. Until another day becomes another week. And eventually leadership asks — <strong>“When did we first know this was at risk?”</strong></p><p>That’s usually an uncomfortable question. <br>Because the answer may be — <strong>A while ago.</strong></p><p>The problem wasn’t that nobody knew. The problem was that everyone kept hoping the situation would resolve before it needed attention.<br>This taught me something important:</p><blockquote><strong><em>Waiting isn’t always patience. Sometimes waiting is simply delaying a decision.</em></strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*cSo3kte3_IrS7hEG" /><figcaption>Photo by <a href="https://unsplash.com/@cherrydeck?utm_source=medium&amp;utm_medium=referral">Cherrydeck</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>Why Do We Avoid Escalation?</h4><p>I think part of the problem is the word itself. <strong>Escalation.</strong></p><p>It sounds confrontational. Someone failed. Someone needs to be called out. Someone senior needs to intervene. So people naturally hesitate. Especially when relationships matter.<br>Nobody wants to damage trust with another team by escalating them.<br>But perhaps that’s because we are escalating the wrong thing.</p><p>Instead of — <strong>“Team X isn’t delivering.”</strong></p><p>What if the conversation became — <strong>“Our current plan depends on X by Friday. The latest estimate is two weeks later. Here are the options available to us.”</strong></p><p>Now we aren’t escalating a person. We aren’t even necessarily escalating a team. We’re escalating a <strong>constraint that requires a trade-off.</strong></p><p>That’s a very different posture.</p><h4>The Moment a Problem Becomes a Decision</h4><p>I’ve started thinking about escalation through a simple question — <strong><em>Can the people currently working on this make the decision required to move forward?</em></strong></p><p>If yes, keep solving.<br>If no, continuing to discuss the issue at the same level may not help.</p><p>For example:<br>Two teams disagree about implementation. They probably need to work through it.<br>But if the disagreement requires choosing between two strategic priorities?Different conversation.<br>A team has too much work. They should first look at sequencing and priorities. But if every item is leadership-mandated and something has to move?<br>That’s a decision.<br>A dependency is delayed. Teams should explore alternatives. But if every alternative changes scope, budget, customer commitment or release date?<br>That’s a decision.</p><p>This is where escalation becomes useful. Not because the team failed.Because <strong>the decision has reached a different boundary.</strong></p><h4>Before Escalating, Do the Thinking</h4><p>This may be the most important lesson for me. Escalation shouldn’t outsource thinking. It should demonstrate it. <br>There is a huge difference between: <strong>“What should we do?” </strong>and: <strong>“Here are the options. Here are the trade-offs. Here is our recommendation. Here is the decision we need from you.”</strong></p><p>The second requires more work. You have to understand the problem. You have to separate facts from assumptions. You have to understand impact. You have to explore alternatives. You have to know what happens if no decision is made.<br>And sometimes you have to recommend an option that isn’t perfect. <br><strong>That’s leadership</strong>.</p><h4>Don’t Bring Ten Options</h4><p>There is another trap here. <br>Sometimes we try so hard not to make the decision for leadership that we present every possible scenario.</p><p>Option A.</p><p>Option B.</p><p>Option C.</p><p>Option D.</p><p>Option E.</p><p>Plus a detailed spreadsheet.</p><p>Plus 17 slides of background.</p><p>Technically, we’ve done our job. But have we?</p><p>A decision-maker shouldn’t need to reconstruct weeks of team discussion to understand what matters. If five options exist but only two are genuinely viable, say that. If one is clearly preferred, say that too.</p><blockquote><strong><em>Good escalation reduces complexity before it moves upward.</em></strong></blockquote><p>That doesn’t mean hiding information. It means doing enough thinking to distinguish signal from noise.</p><h4>The Recommendation Matters</h4><p>I used to think presenting options objectively was enough. Over time, I’ve become much more convinced that a recommendation matters.</p><p><strong>“Here are three options.” </strong>is useful.</p><p>But: <strong>“Here are three options. We recommend Option B because it protects the customer commitment while limiting additional cost. The trade-off is that Feature X moves to the next release.” </strong>is far more useful.</p><p>Why?</p><p>Because the people closest to the work usually understand the implications best. Leadership may have broader context. But the team has deeper context.<br>A recommendation combines the two.</p><p>It says: <strong>This is what we know. This is what we think. This is where we need your decision.</strong></p><h4>What Happens If Nobody Decides?</h4><p>This is one question I think we don’t ask enough. Sometimes a decision appears optional because nobody has explicitly said: <em>“We must decide today.”</em></p><p>So it stays open. And open. And open.</p><p>But not making a decision is still a decision.<br>If the architecture decision isn’t made this week, development may start late. <br>If scope isn’t reduced, the release date may move.<br>If funding isn’t approved, resources may be reassigned.<br>If priorities aren’t clarified, teams may continue working on everything.</p><p>Which usually means making slow progress on everything. So when raising an issue, I like to understand: <strong><em>What happens if we do nothing?</em></strong></p><p>Sometimes that is the most important option on the page.</p><p><strong>Option D: No decision.</strong></p><p>And its consequences should be just as visible as every other choice.</p><h4>There Is a Difference Between Visibility and Escalation</h4><p>This distinction matters too. Not every risk shared with leadership is an escalation. Sometimes leadership simply needs visibility.</p><p><strong>“This dependency is trending late. The teams have a mitigation plan and no decision is required right now.”</strong></p><p>That’s useful information.</p><p>Compare that with: <strong>“The mitigation no longer protects the milestone. We need a scope-versus-date decision by Wednesday.”</strong></p><p>Now we have crossed into escalation.</p><p>I’ve found this distinction helpful because it allows risks to become visible early without turning every risk into a crisis.</p><p>Visibility says: <strong>You should know.</strong></p><p>Escalation says: <strong>We need something from you.</strong></p><p>Those are different messages.</p><h4>Escalation Should Have an Expiration Date</h4><p>One pattern I have learned to avoid is the permanent escalation. <br>A risk appears in the weekly update.</p><p>Week 1: <strong>Escalated.</strong></p><p>Week 2: <strong>Still escalated.</strong></p><p>Week 3: <strong>Leadership aware.</strong></p><p>Week 4: <strong>Monitoring.</strong></p><p>At some point, I have to ask:</p><p><strong>What exactly are we waiting for?</strong></p><p>An escalation without a decision owner or decision date isn’t really an escalation. It’s just a highlighted problem. <br>A useful escalation should make four things clear:</p><ul><li><strong>What needs to be decided?</strong></li><li><strong>Who can make that decision?</strong></li><li><strong>By when?</strong></li><li><strong>What happens if the decision doesn’t happen?</strong></li></ul><p>That turns urgency into action.</p><h4>The Hardest Escalations Are Often About Priorities</h4><p>Some of the most difficult situations I’ve encountered aren’t technical problems at all.</p><p>They’re priority problems. Two initiatives are both important.<br>Both have committed dates. Both need the same people.<br>Nobody wants to move either one. The team is asked to <em>“make it work.”<br></em>And for a while, they try.<br>People stretch. Context switching increases.<br>Work gets fragmented. Progress slows.<br>Then eventually someone asks why execution isn’t predictable. This is exactly where I think good escalation matters.</p><p>Because the team cannot solve a capacity problem created by competing organizational priorities.<br>Working harder isn’t a prioritization strategy. Sometimes the right escalation is simply:<strong><em>“We can deliver A by this date or A+B by this later date. Which outcome matters more?”</em></strong></p><p>That isn’t failure. That’s making reality visible enough for someone to choose.</p><h4>A Small Experiment to Try</h4><p>The next time you’re about to escalate something, pause before sending the message.</p><p>Try completing these five sentences:</p><p><strong>The problem is…</strong></p><p><strong>The impact is…</strong></p><p><strong>We have already tried…</strong></p><p><strong>The realistic options are…</strong></p><p><strong>The decision we need is…</strong></p><p>And then add one more: <strong><em>If no decision is made by ___, then ___.</em></strong></p><p>If you can’t complete those sentences yet, perhaps the issue needs more investigation before escalation.</p><p>But if you can? Then the conversation becomes much easier. You’re no longer walking into the room saying: <strong>“We have a problem.”</strong></p><p>You’re saying:</p><p><strong>“We understand the problem. We understand the choices. We’ve reached the boundary of what we can decide ourselves.”</strong></p><p>And perhaps that’s the real purpose of escalation. Not to push problems upward. Not to assign blame. Not to make someone senior take over.</p><p>But to make sure the <strong>right decision reaches the right person while there is still time for the decision to matter.</strong></p><p><a href="https://medium.com/agile-true-north"><em>Agile True North</em></a><em> is about the lessons that rarely fit neatly into frameworks — the ones we learn while working with real teams, real constraints, and real people.</em></p><p><em>Something to think about: Think of the last issue you escalated — or chose not to escalate. Was the real problem unresolved… or was there a decision nobody was willing or able to make?</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=df168f64ca47" width="1" height="1" alt=""><hr><p><a href="https://medium.com/agile-true-north/dont-escalate-the-problem-escalate-the-decision-df168f64ca47">Don’t Escalate the Problem. Escalate the Decision.</a> was originally published in <a href="https://medium.com/agile-true-north">Agile True North</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Dashboard Became a Blame Tool — and Adoption Collapsed]]></title>
            <link>https://medium.com/agile-true-north/the-dashboard-became-a-blame-tool-and-adoption-collapsed-a394628971ad?source=rss----e69a8c7e233d---4</link>
            <guid isPermaLink="false">https://medium.com/p/a394628971ad</guid>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[project-management]]></category>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[analytics]]></category>
            <dc:creator><![CDATA[Soumya]]></dc:creator>
            <pubDate>Tue, 13 Oct 2026 16:16:01 GMT</pubDate>
            <atom:updated>2026-10-13T16:16:01.536Z</atom:updated>
            <content:encoded><![CDATA[<h4>What happens when visibility starts feeling like surveillance — and teams learn that green is safer than honest.</h4><p>At first, I thought the dashboard was working. Leadership was looking at it. Teams were updating it. Risks were visible. <br>Dependencies were being tracked.</p><p>We finally had the transparency we’d been asking for. Then something strange started happening. <br>The dashboard kept getting greener. But the delivery conversations weren’t getting easier.</p><p>Risks started appearing later. Blockers seemed to stay “under investigation” longer. Amber statuses somehow remained green until there was enough evidence to make green impossible.</p><p>And updates became… careful. Very careful.</p><p>That’s when I realized: <strong>The dashboard hadn’t become more accurate. People had become more cautious about what they put on it.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*36obB-xFO3jWL-6c" /><figcaption>Photo by <a href="https://unsplash.com/@goumbik?utm_source=medium&amp;utm_medium=referral">Lukas Blazek</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>It Didn’t Happen Overnight</h4><p>Nobody announced — <em>“Starting today, this dashboard will be used to judge your team.” <br></em>There was no policy change. No big meeting. No email. It happened gradually.</p><p>A red status appeared.</p><p>Someone asked — <strong>“Why didn’t this team manage the dependency better?” <br></strong>A delivery metric dropped.</p><p>Someone asked — <strong>“What’s going wrong with this team?” <br></strong>An initiative turned amber.</p><p>The conversation quickly moved from — <strong>“What is getting in the way?” <br></strong>to — <strong>“Who owns this?”</strong></p><p>Ownership is important. Accountability is important. But the <em>tone</em> of the questions started changing. And teams noticed. They always do.</p><h4>Then the Dashboard Started Changing</h4><p>Not technically. Behaviorally. <br>People learned that green meant: <strong>No uncomfortable questions. <br></strong>Amber meant: <strong>Explain yourself.<br></strong>Red meant: <strong>Expect escalation.</strong></p><p>So what do you think happened? Nobody falsified the data. Nobody sat down and said — <em>“Let’s hide the truth.”</em></p><p>It was much more subtle than that. A risk became: <em>“We’re still evaluating.”</em></p><p>A dependency became: <em>“In progress.”</em></p><p>A slipping milestone became: <em>“Monitoring closely.”</em></p><p>A blocker became: <em>“Potential concern.”</em></p><p>And my personal favorite: <strong>“Green, with watch items.”</strong>🙂</p><p>At some point, you have to ask — If green needs a paragraph explaining why it might not actually be green… <strong>is it really green?</strong></p><h4>The Coaching Question</h4><p>This experience left me with a question I now find much more useful than: <strong>“Why wasn’t this risk reported earlier?”</strong></p><p>I ask — <strong>“What happens in this organization when someone reports bad news early?”</strong></p><p>Think about that for a moment. Not what the process says should happen. What <em>actually</em> happens? <br>Does someone say: <strong>“Good catch. What do you need?”</strong></p><p>Or: <strong>“How did you let this happen?” </strong>Does leadership help remove the constraint?</p><p>Or does the team immediately have to prepare three slides explaining itself? <br>Does an amber status trigger support? Or scrutiny?</p><p>Because whatever happens next teaches people how transparent they should be the next time.</p><h4>Every Dashboard Has an Invisible Feedback Loop</h4><p>We usually think of dashboards as information flowing upward. <br>Team → Dashboard → Leadership.</p><p>But there’s another flow that matters just as much.<br>Leadership → Reaction → Team behavior → Future data.</p><p>And that loop is powerful.</p><p>Imagine this: <br>A team identifies a dependency risk three weeks before a milestone.<br>They flag it amber.<br>Leadership sees it early, helps resolve the dependency, and thanks the team for surfacing it.<br>What has the organization just taught everyone? <strong>Early visibility is useful.</strong></p><p>Now imagine the same scenario.<br>The team flags amber. They’re immediately pulled into an escalation meeting. They spend hours explaining why the project is “off track.” Someone asks whether the team failed to plan properly. Nothing actually gets unblocked. <br>What has the organization taught them? <strong>Next time, wait until you’re absolutely certain before raising the flag.</strong></p><p>Same dashboard. Completely different behavior.</p><h4>This Is Where Metrics Become Dangerous</h4><p>Metrics themselves aren’t the enemy. Visibility isn’t the enemy. Accountability isn’t the enemy either. <br>The danger begins when we confuse: <strong>a signal </strong>with <strong>a verdict.<br></strong>Cycle time increased — <strong>That’s a signal. </strong><br>It doesn’t automatically mean the team is underperforming. Predictability dropped —<strong> That’s a signal. <br></strong>It doesn’t automatically mean planning was poor. A milestone turned amber — <strong>That’s a signal. <br></strong>It doesn’t automatically mean someone failed. Blocked work increased — <strong>That’s a signal.</strong></p><p>Maybe the team is struggling. Or maybe they are doing exactly what we asked them to do: <strong>making invisible problems visible.</strong></p><p>If every negative signal becomes a judgment, eventually people stop giving you signals.</p><h4>I Had to Rethink What “Good” Looked Like</h4><p>This was uncomfortable for me too. Because as a program leader, I wanted clean dashboards. <br>Current data. Clear ownership. Accurate RAG statuses. Timely updates.<br>All reasonable expectations.</p><p>But I eventually realized that a dashboard full of green isn’t necessarily a sign of healthy delivery. Sometimes the healthier organization is the one showing:</p><p>🟢 Green</p><p>🟡 Amber</p><p>🔴 Red</p><p>because people are comfortable enough to surface reality. That changed how I interpreted dashboards. I stopped asking — <strong>“Why do we have so many ambers?” </strong>and started asking — <strong>“Are these ambers appearing early enough for us to do something about them?”</strong></p><p>That’s a very different conversation.</p><h4>Early Red Can Actually Be Healthy</h4><p>This sounds strange. But sometimes I’d rather see a red status early than a green status late. <br>Why? <br>Because early red gives us options. We can reduce scope. Change sequencing. Resolve a dependency. Add expertise. Escalate a decision. Move a date. Accept the risk consciously. <br>But late red? <br>Late red usually gives us fewer choices. By then, we’re not managing risk. We’re explaining impact. So perhaps one sign of a mature delivery organization isn’t: <strong>“We have very few red projects.” <br></strong>Maybe it’s: <strong>“Our teams surface red while there’s still time to change the outcome.”</strong></p><p>That’s a much harder metric to put on a dashboard. But it’s a much more interesting leadership capability.</p><h4>The Dashboard Principle</h4><p>This experience gave me another principle: <strong><em>A dashboard should make it safer to tell the truth earlier — not more expensive.</em></strong></p><p>Because the moment people believe transparency will be used against them, the dashboard begins losing its most valuable asset.</p><p>Not adoption. Not automation. Not visualization.</p><p><strong>Truth.</strong></p><p>And once that happens, no amount of dashboard redesign will fix the problem.</p><h4>There’s Another Word We Need to Talk About: Accountability</h4><p>Whenever conversations move toward psychological safety, someone eventually says: <strong>“But people still need to be accountable.”</strong></p><p>Absolutely.</p><p>Psychological safety doesn’t mean: “No questions.”</p><p>It doesn’t mean: “No ownership.”</p><p>It doesn’t mean: “Everything is fine.”</p><p>And it certainly doesn’t mean poor delivery should never be challenged.</p><p>But there’s an important difference between asking: <strong>“Who caused this?”</strong></p><p>and: <strong>“Who can help move this forward?”</strong></p><p>Between: <strong>“Why are you red?”</strong></p><p>and: <strong>“What moved you to red, and what decision do we need to make?”</strong></p><p>Between: <strong>“Why wasn’t this handled?”</strong></p><p>and: <strong>“What prevented us from seeing this earlier?”</strong></p><p>Accountability should improve the system. Blame usually improves the storytelling. People get better at explaining the number. Not necessarily at changing the outcome.</p><h4>A Small Change That Made Conversations Better</h4><p>One thing I’ve found useful is changing the language around status reviews. Instead of asking teams to <strong>defend their status</strong>, make the status the beginning of the conversation. <br>For every amber or red item, ask: <br><strong>What changed? <br>What is the impact if we do nothing?<br>What has the team already tried?<br>What decision, support, or escalation is needed?<br>Who can help?</strong></p><p>Now the dashboard has a purpose. <br>The amber isn’t a scarlet letter. It’s a request for attention. <br>Red isn’t failure. It’s a signal that the current plan needs intervention.</p><p>That doesn’t remove accountability. It makes accountability <strong>actionable.</strong></p><h4>And Here’s the Test I Wish More Leaders Would Try</h4><p>Look at your dashboard. Find the teams that are consistently green. Not occasionally. <strong>Consistently. </strong>Now don’t congratulate them yet.</p><p>Ask: <strong><em>“Are they genuinely predictable — or have they learned how to stay green?”</em></strong></p><p>That’s not an accusation. It’s a diagnostic question.</p><p>Because there are at least two ways to create a beautifully green dashboard.</p><p><strong>Option 1 — </strong>Build a healthy delivery system. Surface risks early. Manage dependencies. Make trade-offs. Learn continuously.</p><p><strong>Option 2 — </strong>Move the definition of red. One is considerably easier.</p><h4>Try This Tomorrow</h4><p>At your next status review, pick <strong>one amber or red item. <br></strong>Before asking the owner to explain it, say: <strong><em>“I’m glad we’re seeing this now. What can we do while we still have options?”</em></strong></p><p>Then notice what happens. Does the conversation change? Do people offer information more freely? Does someone mention another risk? Does the discussion move from explanation to problem-solving?</p><p>Now take it one step further. At the end of the meeting, ask: <strong><em>“Is there anything we’re currently calling green that you’re uncomfortable calling green?”</em></strong></p><p>Then wait. Don’t fill the silence. You might be surprised by what comes next.</p><h4>The Most Valuable Dashboard Might Be the One People Aren’t Afraid to Make Ugly</h4><p>We spend a lot of time making dashboards beautiful. Clean layouts. Consistent colors. Executive summaries. Automated updates. Perfect visualizations.</p><p>But after years of working around delivery teams, I’ve come to value something else much more. <br><strong>A dashboard that people are willing to make ugly. </strong>Amber. Red. Blocked. Unknown. Decision needed. Dependency unresolved. Confidence dropping.</p><p><strong><em>Because messy truth is far more useful than beautiful reassurance.</em></strong></p><p>And if your dashboard makes problems visible while there’s still time to solve them… it’s doing its job.</p><h4>Your Turn</h4><p>Think about the last time someone changed a status from green to amber or red. What happened next? Did the organization become curious? Or critical? Did people help? Or interrogate? Did the team leave the conversation thinking: <strong>“I’m glad we raised that early.”</strong></p><p>Or: <strong>“Next time, let’s be more careful before we put something in red.”</strong></p><p>Because that reaction may tell you more about the quality of your dashboard than any metric displayed on it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a394628971ad" width="1" height="1" alt=""><hr><p><a href="https://medium.com/agile-true-north/the-dashboard-became-a-blame-tool-and-adoption-collapsed-a394628971ad">The Dashboard Became a Blame Tool — and Adoption Collapsed</a> was originally published in <a href="https://medium.com/agile-true-north">Agile True North</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Your Communication Plan Is Not Your Meeting Calendar]]></title>
            <link>https://medium.com/agile-true-north/your-communication-plan-is-not-your-meeting-calendar-89300d27092a?source=rss----e69a8c7e233d---4</link>
            <guid isPermaLink="false">https://medium.com/p/89300d27092a</guid>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[project-management]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[communication]]></category>
            <dc:creator><![CDATA[Soumya]]></dc:creator>
            <pubDate>Fri, 09 Oct 2026 15:41:01 GMT</pubDate>
            <atom:updated>2026-10-09T15:41:01.491Z</atom:updated>
            <content:encoded><![CDATA[<h4>We had stand-ups. We had status meetings. We had dashboards. Somehow, people still didn’t know what was happening.</h4><p>We had meetings. Lots of them…</p><p>Daily stand-ups. Weekly status meetings. Cross-team syncs. Leadership updates. Sprint reviews. Dashboards. Status reports.</p><p>And yet, one day, someone asked a question that stopped me — <strong>“Why am I hearing about this only now?”</strong></p><p>My immediate reaction was — <em>But we talked about this. </em>It had been mentioned in a meeting. It was on the dashboard. It appeared in the status update. <br>Technically, the information had been communicated.</p><p>Or had it?</p><p>That was one of those moments that changed how I think about communication. Because I slowly realized — <strong><em>Sending information and creating understanding are not the same thing.</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*77khWIY7w3bsT2M0" /><figcaption>Photo by <a href="https://unsplash.com/@mariogogh?utm_source=medium&amp;utm_medium=referral">Mario Gogh</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>“But We Discussed This Last Week”</h4><p>I think most of us have been in this situation. A dependency becomes critical. A stakeholder is surprised.</p><p>Someone says: <em>“Why wasn’t this raised earlier?”<br></em>And someone else immediately responds: <em>“It was. We discussed it last Tuesday.”</em></p><p>Both statements might actually be true. The issue <em>was</em> discussed. And the stakeholder still didn’t understand its significance. That is where communication gets interesting.</p><p>Because mentioning something isn’t necessarily communicating it. Putting something on slide 17 isn’t communicating it. Adding a red icon to a dashboard isn’t communicating it. Sending an email to 25 people isn’t communicating it.<br>Even discussing something in a meeting doesn’t guarantee that the right people understood:</p><p><strong>What happened? Why does it matter? What happens if we do nothing?</strong></p><p><strong>Who needs to act? By when?</strong></p><p>Without those answers, we may have shared information. But we may not have created clarity.</p><h4>More Communication Isn’t Always Better Communication</h4><p>When teams experience communication problems, the instinctive response is often — <strong>Let’s add another meeting.</strong></p><p>I’ve seen this many times. Something falls through the cracks. So we create a sync. <br>Another dependency gets missed. We create another sync.<br>Leadership wants more visibility. Now there’s another status meeting.<br>Before long, calendars are full and everyone is spending hours talking about the work instead of doing the work.<br>And somehow the original communication problem still exists.<br>Because the problem wasn’t necessarily the absence of communication.<br>Sometimes the problem was that <strong>the communication had no clear purpose.</strong></p><h4>Every Audience Needs Something Different</h4><p>This was probably one of the biggest shifts in how I approached communication over time.<br>I stopped thinking: <strong>“What status do I need to send?”</strong></p><p>And started thinking: <strong>“What does this audience need to understand?”</strong></p><p>Those are very different questions. The team closest to the work may need — dependencies, technical constraints, sequencing, ownership and immediate blockers.</p><p>A product partner may care about — scope, trade-offs, customer impact and decisions.</p><p>A leader may need — overall progress, major risks, investment implications and where intervention is required.</p><p>Someone dependent on your work may simply need: <strong>Are you still delivering what I need, when I need it?</strong></p><p>Same program. Different information needs. Sending everyone the same detailed status update may feel efficient. But efficiency isn’t the goal. <strong>Understanding is.</strong></p><h4>The Dashboard Isn’t the Conversation</h4><p>I love dashboards. They can make invisible problems visible. They can show trends. They can expose aging work. They can reveal dependencies. They can give teams and leaders a shared view of what’s happening. But dashboards have a limitation. They show you <strong>what</strong>. <br>They don’t always tell you <strong>so what?</strong></p><p>A milestone turns amber. Okay. What changed?<br>A dependency is late. What does it affect?<br>Capacity drops. What are we no longer able to deliver?<br>Three items are blocked. Which one actually matters?<br>A good dashboard should help start the conversation.</p><p>It shouldn’t be expected to have the conversation for us.</p><h4>One Question Changed My Status Updates</h4><p>At some point, I started asking myself this before sending an update — <strong><em>“What do I want the person reading this to know, decide, or do?”</em></strong></p><p>That small question changed a lot. Because suddenly — <strong>“API dependency delayed.” </strong>wasn’t enough.</p><p>Instead — <strong>“The API dependency is two weeks behind. If it isn’t available by Friday, testing moves into the next release window. We need confirmation from Team X by Thursday on whether the revised date is achievable.”</strong></p><p>Same issue. Very different communication.<br>One reports information. The other creates a path to action.</p><h4>The Most Important Information Is Often Buried</h4><p>Another pattern I’ve noticed: Teams frequently communicate everything they know instead of emphasizing what people need to know.</p><p>A status update might contain — 12 accomplishments. 8 upcoming activities. 6 metrics. 4 dependencies. 3 risks. 2 decisions.</p><p>And one genuinely important issue. But everything is formatted exactly the same way. So the reader has to figure out what matters.<br>That works when someone is deeply involved in the program.<br>It doesn’t work when a leader has five minutes between meetings.</p><p>One of the simplest communication improvements I’ve learned is — <strong><em>Don’t make the audience discover the message.</em></strong></p><p>If something needs attention, say it. If a decision is required, say it. If the timeline is at risk, say it. If nothing needs action, say that too.</p><p>Clarity is a kindness.</p><h4>“Everyone Was in the Meeting” Is Not an Ownership Model</h4><p>This one deserves special attention. A decision gets discussed. Everyone agrees. The meeting ends.</p><p>A week later, nothing has happened. Then comes the question — <em>“Who was supposed to do that?”</em></p><p>And somebody says — <em>“I thought they were handling it.”</em></p><p>Someone else says — <em>“I thought you were.”</em></p><p>This is where I have learned that every important conversation should eventually answer — <strong>Who owns the next move?<br></strong>Not — <em>“The team will look into it.” <br></em>Not — <em>“We’ll follow up.”<br></em>Not — <em>“Let’s take this offline.”</em></p><p>Who? What? By when?</p><p>Those three pieces can prevent an incredible amount of confusion.</p><h4>Sometimes Silence Is Misleading</h4><p>There is another communication trap I’ve learned to watch for. You send an update. Nobody responds.</p><p>So you assume — <strong>Everyone is aligned.</strong></p><p>Maybe. Or maybe nobody read it. <br>Maybe they read it and didn’t understand the impact. <br>Maybe they assumed someone else would respond.<br>Maybe they disagreed but didn’t think a response was expected.</p><p>This becomes especially important when the communication contains a decision, trade-off or significant risk.</p><p>For those situations, silence isn’t always alignment. Sometimes you need explicit confirmation. <strong>“We are proceeding with Option B unless there are objections by Thursday” <br></strong>is very different from — <strong>“Here are the options we discussed.”</strong></p><p>Communication should make the expected response clear.</p><h4>Good Communication Reduces Meetings</h4><p>This might sound counterintuitive. But I have found that better communication often means <strong>fewer meetings</strong>, not more.</p><p>When ownership is clear… When decisions are documented… When dashboards answer routine questions… When risks are visible… When written updates explain what changed and why it matters…</p><p>Then meetings don’t have to become information-sharing sessions.</p><p>They can be used for the things that actually benefit from having people together — <strong>decisions. trade-offs. problem solving. alignment. difficult conversations.</strong></p><p>That’s a much better use of everyone’s time.</p><h4>Before You Add Another Meeting…</h4><p>The next time someone says: <strong>“We need better communication.”<br></strong>Try not to immediately schedule another sync.</p><p>Instead, ask — <br><strong>Who isn’t getting the information they need? What do they need to understand? What decision or action should this communication enable? What is the simplest way to deliver that information?</strong></p><p>Sometimes the answer will be a meeting. Sometimes it will be a dashboard. Sometimes a short written update. Sometimes a quick conversation between two people. <br>And sometimes the solution is simply writing one sentence more clearly.</p><h4>A Small Experiment to Try</h4><p>Take one recurring status meeting or update you own. Before preparing it, write this sentence — <strong><em>“After this communication, I want this audience to ______.”</em></strong></p><p>Maybe the answer is: Understand the risk. Make a decision. Confirm a dependency. Choose between two options. Know that no action is required.</p><p>Then look at everything you planned to communicate.</p><p>If something doesn’t help achieve that outcome, ask whether it really needs to be there. You may discover that your 12-slide update needs six slides. Your one-hour meeting needs 30 minutes. Or your meeting doesn’t need to happen at all.</p><p>Because communication isn’t measured by how many meetings we hold, emails we send or dashboards we build. It’s measured by something much simpler: <strong><em>Did the right people understand what mattered — and know what needed to happen next?</em></strong></p><p>That’s the communication plan. The calendar is just one of the tools.</p><p><a href="https://medium.com/agile-true-north"><em>Agile True North</em></a><em> is about the lessons that rarely fit neatly into frameworks — the ones we learn while working with real teams, real constraints, and real people.</em></p><p><em>Have you ever been in a situation where everyone said, “But we already communicated this,” and somehow nobody was actually aligned?</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=89300d27092a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/agile-true-north/your-communication-plan-is-not-your-meeting-calendar-89300d27092a">Your Communication Plan Is Not Your Meeting Calendar</a> was originally published in <a href="https://medium.com/agile-true-north">Agile True North</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>