<?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[Stories by Konstantin Rohleder on Medium]]></title>
        <description><![CDATA[Stories by Konstantin Rohleder on Medium]]></description>
        <link>https://medium.com/@rohlederkonstantin?source=rss-7d025f6fd75c------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*AhCqaSOiLoYKp8gn</url>
            <title>Stories by Konstantin Rohleder on Medium</title>
            <link>https://medium.com/@rohlederkonstantin?source=rss-7d025f6fd75c------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 20:25:34 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@rohlederkonstantin/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Taming Complexity: HelloFresh’s Playbook for Managing Large-Scale Change (Part 3/3)]]></title>
            <link>https://engineering.hellofresh.com/taming-complexity-hellofreshs-playbook-for-managing-large-scale-change-part-3-3-ec0fd8bc6cd9?source=rss-7d025f6fd75c------2</link>
            <guid isPermaLink="false">https://medium.com/p/ec0fd8bc6cd9</guid>
            <category><![CDATA[program-management]]></category>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[engineering-management]]></category>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[change-management]]></category>
            <dc:creator><![CDATA[Konstantin Rohleder]]></dc:creator>
            <pubDate>Wed, 25 Jun 2025 07:14:21 GMT</pubDate>
            <atom:updated>2025-06-25T07:14:21.506Z</atom:updated>
            <content:encoded><![CDATA[<h3>Part 3: Technical Complexity &amp; the ASEI Framework</h3><p>⏳ <em>6 min read</em></p><p><strong><em>Final stretch!</em></strong><em> We’ve handled people and processes, now let’s crack the code side and stitch everything together into a four-stage framework you can reuse.</em></p><h3>Technical Complexity</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*wokmW2u4izwrjDr8" /><figcaption><em>A tightly linked 3 × 3 cluster of puzzles on the right and a 2 × 1 mini-cluster on the left, each internally connected by white dashed lines</em></figcaption></figure><p>Technical complexity stems from the intricacies of systems, integrations, and infrastructure. This might include legacy monoliths, microservices, data platforms, or specialized vendor solutions. Each system introduces dependencies that can make a large-scale change more complicated, especially if multiple teams share those systems.</p><h3>High Technical Complexity Examples</h3><p><strong>Incremental Renaming &amp; Migration of Data Streams</strong></p><p>Imagine a situation where core data streams or message queues must be renamed or reorganized to support new business requirements. While a bridging mechanism might allow old and new versions to coexist temporarily (e.g. running both “old-named” and “new-named” streams in parallel), coordination remains critical to ensure consuming services switch in sync. Delays, mismatched updates, and additional infrastructure costs can occur if teams don’t thoroughly plan and communicate each step.</p><p><strong>Complex Data Dependencies Across Multiple Services</strong></p><p>In many organizations, root data tables feed into dozens of downstream applications or analytical workflows. Even if a phased approach permits partial coexistence of old and new schemas, leaders must still orchestrate each step to avoid data inconsistencies. If one team migrates before others are ready, a cascade of incomplete or inaccurate data can hamper critical services.</p><p><strong>Next-Generation Multi-Cluster Platform Adoption</strong></p><p>Transitioning to a more resilient, multi-cluster platform may reduce operational risk in the long run, but it introduces immediate complexity. Core Workstreams building the platform must define infrastructure blueprints, traffic-routing policies, and continuous delivery pipelines, while each team migrating a service has to manage its own roadmap for switching over. Although the architecture might permit incremental migrations, each application must adjust configurations, monitoring setups, and deployment procedures to ensure compatibility and find workarounds for edge-cases.</p><p>Each scenario highlights how high technical complexity demands careful cross-team coordination, thorough dependency mapping, and a balanced approach between incremental, decoupled and unified changes. Delayed updates or partial transitions can inflate operational risks and business costs, especially when bridging mechanisms (e.g. parallel environments) are extended too long. Ultimately, they illustrate the interplay between architecture, data flows, and team ownership, underscoring that without synchronized planning and robust governance, large-scale transitions are highly complex and can quickly become unwieldy and costly if the complexity is not managed and reduced.</p><h3>Unification vs. Decoupling</h3><p>The key to addressing technical complexity from a program perspective often boils down to deciding:</p><ul><li><strong>What must be done centrally, in parallel or in a unified way?</strong> (e.g. building a new solution)</li><li><strong>What can be done at the edges or by individual teams independently?</strong> (e.g. adopting the platform at one’s own pace)</li></ul><p><strong>Example: A cross-alliance platform initiative</strong></p><p>Consider a new platform initiative that brings together multiple core workstreams into a “uber-workstream” building the solution. If these squads reside in the same alliance or tribe, alignment, while still complex, is relatively manageable. But the broader organization must then adopt the platform on their own timelines. Because the underlying architecture supports <strong>decoupled</strong> migrations, each team can transition at its own pace rather than orchestrating a massive, simultaneous big-bang or “lift-and-shift”. Echoing our earlier point from Part 1 this scenario illustrates how technical design choices can shape organizational strategy, reducing overall complexity through careful decoupling.</p><h3>Automation and Documentation</h3><p>Where possible, reduce cognitive load and manual effort by providing scripts and tooling to minimize manual tasks teams have to perform during migrations. Establish documented “Golden Paths” that cover 80% of common scenarios, along with a clear plan for handling edge cases. Additionally, standardize processes and guides with clear steps, ensuring consistency and efficiency. Don’t forget to “pilot” these guides first as well, before rolling them out to the wider audience.</p><h3>Core Techniques for Reducing Technical Complexity</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*XOmnf2MXVm_9EQfZ" /></figure><p>Below are further ways to consider to reduce technical complexity:</p><ul><li><strong>Deprecate</strong>: Identify systems that are no longer adding value and phase them out. This can drastically reduce the surface area of your tech estate.</li><li><strong>Modularize</strong>: Implement a modular architecture that allows teams to upgrade or replace individual components without taking down the entire system.</li><li><strong>Split</strong>: Segment a large, monolithic system into smaller components (e.g. microservices or sub-networks) to reduce the global scope of each change.</li><li><strong>Merge</strong>: If multiple systems or tools fulfill the same purpose, merge them into a single, unified platform or solution.</li><li><strong>Standardize</strong>: Adopt common processes, data formats, and interfaces, so teams can work consistently across multiple systems. This reduces friction and fosters reuse.</li><li><strong>Automate</strong>: Automate repetitive steps, from testing to deployment. But remember: <strong>Don’t automate complexity. First simplify, then automate</strong>.</li><li><strong>Reorganize</strong>: Rethink the order in which tasks, data flows, or systems are arranged. Small changes in workflows can significantly reduce complexity.</li></ul><h3>Putting It All Together — The ASEI Framework to Reduce Complexity</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6q9aGKB0zbeeJWQaEWm9vA.png" /></figure><p>A single, systematic approach often delivers far more than a series of ad-hoc fixes. When the key levers outlined in this series are pulled within one coherent framework, their effects compound instead of competing. The four-stage model that follows is one way to organise those levers. Use it as a template, blend it with your own rituals, and most importantly, adapt it to suit your context.</p><h3>ASSESS</h3><p><em>Objective: A clear picture of the existing organizational, operational, and technical landscape, plus any early insights into dependencies and risk levels.</em></p><ul><li><strong>Identify Overall Goals &amp; Boundaries</strong>: Are you dealing with a single alliance or multiple alliances? Are you tackling an entire platform or a localized feature?</li><li><strong>Evaluate Types of Complexity</strong>: Is the dominant challenge organizational (e.g. many stakeholders), operational (e.g. unclear prioritization), or technical (e.g. legacy systems)?</li><li><strong>Check Existing Alignment</strong>: What was revealed during (Yearly) Planning or through early discovery sessions?</li></ul><h3>STRATEGIZE</h3><p><em>Objective: A program blueprint (or lightweight alternative) with clear priorities, alignment on KRs, an understanding of organizational vs. operational vs. technical constraints, and a rollout strategy.</em></p><ul><li><strong>Program Evaluation</strong>: From the <em>Assess</em> stage, decide if a formal program with governance, steering committees, and KRs is needed or if a lighter-weight working group suffices.</li><li><strong>RACI &amp; Accountability: </strong>Who must be involved and who is accountable for what?</li><li><strong>Prioritization: </strong>Which work packages provide the most value for the lowest effort? Consider the previously mentioned ICE scores and tech real estate data.</li><li><strong>Decide on Unification vs. Decoupling</strong>: Identify where centralized efforts are critical (e.g. building the core platform) vs. where teams can move independently.</li><li><strong>Explore Approaches to Reduce Technical Complexity: </strong>How can you break down the technical real estate to reduce the technical complexity the most?</li><li><strong>Communication Strategy: </strong>Define official communication channels</li><li><strong>Rollout Strategy:</strong> Determine which rollout strategy is required or provides the most value in relation to your goals and deadlines?</li><li><strong>Key Results</strong>: Utilize KRs as a transparency and alignment mechanism</li></ul><h3>EXECUTE</h3><p><em>Objective: Teams execute the plan under a clear governance structure with aligned best practices, automation, and strong communication channels.</em></p><ul><li><strong>Core &amp; Adoption Workstreams</strong>: Assigned Squads build the central solution and program participating squads adopt changes</li><li><strong>Change Champions &amp; SMEs</strong>: Domain experts and “Change Champions” are involved early to clarify feasibility and reduce rework.</li><li><strong>Decision-Making: </strong>Utilize decision makers to pivot program strategy where necessary.</li><li><strong>Enable Incremental Changes: </strong>Ensure teams can work in small increments</li><li><strong>Standardize &amp; Automate</strong>: Provide scripts, tooling, and “Golden Paths” so teams aren’t each reinventing the wheel and can save time during execution</li><li><strong>Introduce Gating Mechanisms</strong>: Stop new tech debt from creeping in (e.g. no more deploying to a legacy environment).</li></ul><h3>ITERATE</h3><p><em>Objective: A cycle of improvement, with reduced complexity over time, plus a scalable model for applying these lessons to other programs across the organization.</em></p><ul><li><strong>Continuous Reassessment</strong>: Have new complexities emerged due to organizational changes, newly discovered dependencies, or tech limitations?</li><li><strong>Review &amp; Adapt</strong>: Are we meeting our goals? Is the chosen rollout strategy still optimal?</li><li><strong>Capture Lessons Learned</strong>: Document successes/failures for future programs. Update guidelines, tooling, and processes.</li></ul><h3>Final Thoughts &amp; Key Takeaways</h3><p>Complexity is inevitable but <strong>unwanted complexity</strong> isn’t. By recognizing whether an initiative is primarily facing organizational, operational, or technical challenges, you can choose the right strategies (unification vs. decoupling or big-bang vs. incremental approaches) to tame it.</p><p>When organizations actively work to reduce complexity, whether by retiring or consolidating systems, embedding SMEs, standardizing processes, or changing organisational structure, they create an environment that supports faster innovation, lower risk, and stronger alignment on what really matters. In the end, simpler is almost always better: fewer moving parts, fewer risks, and more time to spend on delighting customers and pushing the boundaries of what your product and teams can achieve.</p><p><strong><em>Thanks for joining the journey!</em></strong><em> If these ideas helped, clap or leave a question. We’re keen to hear your own war stories of taming complexity.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ec0fd8bc6cd9" width="1" height="1" alt=""><hr><p><a href="https://engineering.hellofresh.com/taming-complexity-hellofreshs-playbook-for-managing-large-scale-change-part-3-3-ec0fd8bc6cd9">Taming Complexity: HelloFresh’s Playbook for Managing Large-Scale Change (Part 3/3)</a> was originally published in <a href="https://engineering.hellofresh.com/">HelloTech</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Taming Complexity: HelloFresh’s Playbook for Managing Large-Scale Change (Part 2/3)]]></title>
            <link>https://engineering.hellofresh.com/taming-complexity-hellofreshs-playbook-for-managing-large-scale-change-part-2-3-516dc3961e26?source=rss-7d025f6fd75c------2</link>
            <guid isPermaLink="false">https://medium.com/p/516dc3961e26</guid>
            <category><![CDATA[program-management]]></category>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[change-management]]></category>
            <category><![CDATA[engineering-mangement]]></category>
            <category><![CDATA[software-architecture]]></category>
            <dc:creator><![CDATA[Konstantin Rohleder]]></dc:creator>
            <pubDate>Wed, 18 Jun 2025 07:17:04 GMT</pubDate>
            <atom:updated>2025-06-18T07:17:04.962Z</atom:updated>
            <content:encoded><![CDATA[<h3>Part 2: Operational Levers: From Big-Bang to Incremental Wins</h3><p>⏳<em>6 min read</em></p><p><strong>Welcome back!</strong> In <a href="https://engineering.hellofresh.com/taming-complexity-hellofreshs-playbook-for-managing-large-scale-programs-part-1-3-cdf06c5a6ed9">Part 1</a> we saw why complexity deserves attention. Today we’ll get strategic, deciding when a heavyweight program is justified, how to slice scope, and which rollout strategy minimises drama.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*u8JhYaacHngIJMk3J3xCLQ.png" /><figcaption>Two hands push and pull five labelled levers, Program Evaluation, Key Results, Prioritisation, Roll-out Strategy, Gating Mechanisms, on a green control panel</figcaption></figure><h3>Operational Complexity</h3><p>Operational complexity revolves around how the work itself gets done, from planning and sequencing to governance and execution. This includes:</p><ul><li><strong>Program Evaluation</strong>: Deciding whether a strategy requires a more formal program with set governance and KRs or not.</li><li><strong>Prioritization</strong>: Determining which initiatives are “must-haves” vs. “nice-to-haves.” and how they are sequenced.</li><li><strong>Rollout Strategy</strong>: Choosing between big-bang launches or phased/incremental deployments.</li></ul><h3>Program Evaluation</h3><p>Not every large initiative requires the formal overhead of a structured program with a dedicated TPM. A key determinant is the level of interdependency and the need for synchronized coordination across multiple teams or alliances.</p><p>If each team can execute the work following a straight-forward shared documentation standard the risk of misalignment can be relatively low. As a result, a program might not be strictly necessary; a working group or SME contact might be enough for occasional queries.</p><p>On the other hand, for complex initiatives requiring new technologies and multi-step adoption, robust oversight becomes indispensable. In high-stakes environments where critical tech real estate directly affects service availability and continuity, comprehensive dependency mapping, structured decision-making, and centralized governance are paramount. By proactively identifying risks and implementing gating mechanisms, we safeguard stability and efficiency at scale, ensuring clear visibility into both successes and failures.</p><h3>Key Results (KRs) as transparency and alignment mechanism</h3><p>Each tribe or squad participating in the program should have a KR tied to the initiative. This fosters transparency, alignment, and makes it easier for leadership and the program owner to track progress during, e.g. Monthly Alliance Reviews which include KR progress reports. KRs create a shared language for success. If a squad commits to a well-defined KR, everyone knows exactly what must be done, and resources can be allocated accordingly.</p><p>More importantly, by discussing these changes in MARs, we can uncover issues at scale, leverage information to feed into the Program Strategy and derive lessons learned for future large-scale efforts.</p><h3>Prioritization</h3><p>Prioritization is a powerful lever for reducing operational complexity. While it’s a complex discipline in its own right, below are some short examples of how you can leverage it effectively to streamline large-scale initiatives:</p><ol><li><strong>Align Early &amp; Often: </strong>Incorporate prioritization efforts into formal planning cycles, like Yearly Planning, so that the highest-impact items receive appropriate attention and resourcing from the outset. Early alignment on priorities prevents overcommitting teams and mitigates the risk of “scope creep.”</li><li><strong>Start with Objective Criteria: </strong>Define clear parameters to gauge both the <em>value</em> of an initiative (e.g. user impact, revenue potential, risk mitigation) and the <em>cost or effort</em> (e.g. engineering capacity, timeline constraints). The <strong>ICE</strong> framework (Impact, Confidence, Effort) is a popular way to quickly categorize and compare different initiatives.</li><li><strong>Distinguish “Must-Haves” vs. “Nice-to-Haves”: </strong>Within each initiative, label the requirements that deliver the highest business value or mitigate critical risks as <em>must-haves</em>. Less essential features go under <em>nice-to-haves</em>. This distinction forces clarity on what really moves the needle and what can be deferred if time or resources run short.</li><li><strong>Let the data pick the battles: </strong>Data turns hunches into evidence: Some examples of tech real estate data that can additionally feed your ICE scores and prioritisation: service-mesh traces or static graphs surfacing tight inter-service knots; ownership density showing squads with few repos ready for quick wins versus teams that will need heavier lifting; service criticality, service maturity and incident logs spotlight Tier-1 systems where changes can not only pay off fastest but enable the biggest impact. Together they temper gut feel with facts.</li><li><strong>Map Out Dependencies &amp; Sequence Work: </strong>Once initiatives are sorted by ICE scores and criticality, revisit technical dependencies to identify tasks that must logically precede others. This step ensures you don’t block high-value work or create last-minute surprises by uncovering a hidden prerequisite too late.</li><li><strong>Keep Scope Manageable: </strong>By focusing on top-impact initiatives first and limiting the scope of each iteration, you <strong>reduce operational complexity</strong> in tangible ways: <br> — <strong>Resource Efficiency</strong>: Avoid overloading engineering capacity, especially when multiple tribes or alliances are involved.<br> — <strong>Fewer Collisions</strong>: Minimize cross-team scheduling headaches and coordination overhead.<br> — <strong>Less Firefighting</strong>: Cutting down on half-finished projects and last-minute crises frees teams to maintain quality and adapt proactively.</li><li><strong>Exercise Extreme Caution with Large Programs: </strong>In programs involving 50+ Tribes or 100+ squads, each additional requirement can exponentially increase effort, cost, and coordination overhead. <strong>Keep increments small</strong> and make sure each incremental step provides substantial value relative to its complexity.</li></ol><p>Bottom Line: Effective prioritization isn’t just about picking what to do first; it’s about reducing the <em>operational burden</em> on teams. By systematically scoring initiatives, aligning on must-haves, and limiting scope, we create a more predictable, agile environment where resources aren’t spread too thin, allowing teams to deliver the right outcomes at the right time with fewer surprises.</p><h3>Rollout Strategy</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*YZVFLIgKAU6T-OcQ" /><figcaption><em>Split graphic: top shows a rocket and explosion under the caption ‘BIG-BANG’; bottom shows boxes moving along a conveyor belt under ‘PHASED’, illustrating one-shot vs. incremental rollouts.</em></figcaption></figure><p>Rollout Strategy often hinges on risk tolerance, user adoption patterns, and system interdependencies. Where a big-bang launch may suit highly cohesive systems that must change all at once, it also demands meticulous alignment across teams, introducing significant risk. Conversely, phased approaches allow smaller pilots to gather feedback, reduce blast radius, and facilitate incremental course corrections. In practice, we often favor phased rollouts for large-scale initiatives to balance technical complexity with operational agility while still driving meaningful progress.</p><p>Even after decoupling, prioritizing, and opting for a phased rollout, further opportunities for operational efficiency remain. This goes beyond operational efficiency alone: Slicing work into smaller increments and delivering them, even if only internally at first, is the very essence of successful agile product delivery, as big-bang releases often introduce unnecessary risk. By structuring each work package for incremental execution, teams can swiftly adapt, detect issues early, and maintain momentum. This approach not only streamlines planning and resource allocation but also maximizes the value realized from managing complexity over time.</p><h3>Gating Mechanisms</h3><p>Operational complexity often arises when teams continue to create new technical debt while an initiative is trying to reduce it. By introducing gating mechanisms, e.g. blocking new deployments to outdated environments, you effectively “stop the bleeding”, preventing further entrenchment of deprecated patterns. Failing to do so creates a never-ending cycle, where you continually chase and retrofit newly added legacy elements, expanding the scope of the original initiative indefinitely.</p><p>The cornerstone of successful change management lies in making it easy for teams to adopt modern practices while making it hard to persist with suboptimal ones.</p><p>Concretely, you might implement gating in three stages:</p><ol><li>Inform users about the upcoming gating plans, gating ETA and provide comprehensive documentation on how to “follow the new pattern”.</li><li>Introduce a “soft gating” period where you alert non-compliant users</li><li>Fully enforce gating</li></ol><p><em>We’ve lined up the people and the plan, now it’s time for the plumbing. Part 3 will dissect technical complexity and unveil the ASEI framework, a recipe to turn hard lessons into repeatable wins.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=516dc3961e26" width="1" height="1" alt=""><hr><p><a href="https://engineering.hellofresh.com/taming-complexity-hellofreshs-playbook-for-managing-large-scale-change-part-2-3-516dc3961e26">Taming Complexity: HelloFresh’s Playbook for Managing Large-Scale Change (Part 2/3)</a> was originally published in <a href="https://engineering.hellofresh.com/">HelloTech</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Taming Complexity: HelloFresh’s Playbook for Managing Large-Scale Programs (Part 1/3)]]></title>
            <link>https://engineering.hellofresh.com/taming-complexity-hellofreshs-playbook-for-managing-large-scale-programs-part-1-3-cdf06c5a6ed9?source=rss-7d025f6fd75c------2</link>
            <guid isPermaLink="false">https://medium.com/p/cdf06c5a6ed9</guid>
            <category><![CDATA[program-management]]></category>
            <category><![CDATA[change-management]]></category>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[engineering-mangement]]></category>
            <category><![CDATA[leadership]]></category>
            <dc:creator><![CDATA[Konstantin Rohleder]]></dc:creator>
            <pubDate>Wed, 11 Jun 2025 08:03:00 GMT</pubDate>
            <atom:updated>2025-06-11T08:03:00.914Z</atom:updated>
            <content:encoded><![CDATA[<h3>Taming Complexity: <strong><em>HelloFresh’s Playbook for Managing Large-Scale Change</em></strong> (Part 1/3)</h3><h3>Part 1: Why Complexity Matters &amp; Where To Begin</h3><p>⏳<em>6 min read</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ToJF1JlTw2UFr8I7axdj2A.png" /><figcaption><em>Chef juggling tangled and neat spaghetti</em></figcaption></figure><p><strong><em>Welcome to part 1 of our three-part dive into complexity.</em></strong></p><p>If you’ve ever watched either a seemingly simple or highly complex cross-team initiative slow-crawl through tangled dependencies, this series is for you. Drawing on lessons we’ve accumulated at HelloFresh over years of untangling exactly those snarls, Part 1 unpacks what “complexity” really means, why cutting it pays off, and how to take the very first swing at the knot. Parts 2 and 3 dive into the nuts-and-bolts of operational and technical tactics, then stitch them together into a repeatable playbook you can tailor to your own organisation.</p><h3>Introduction</h3><p>Complexity is inevitable in modern organizations, especially when managing cross-functional programs. At HelloFresh, we’ve found structured ways to address it. To achieve top-level goals like C-level or alliance-level OKRs, multiple tech teams must align across diverse tech stacks, roadmaps, and priorities, inevitably multiplying complexity. While yearly or quarterly planning and dependency mapping processes help with initial alignment, actively reducing complexity remains key to success.</p><p>Leaders across HelloFresh, including those managing large groups, known internally as “alliances” and “tribes” (check out our previous “<a href="https://engineering.hellofresh.com/how-hellotechs-working-and-knowledge-sharing-culture-supports-a-company-on-scale-cb37f1901947">HelloTech ways of working</a>” blog for reference), as well as TPMs, often juggle overlapping priorities, numerous stakeholders, and intricate systems. Yet, effectively managing complexity is essential for delivering on time, keeping teams aligned, and maintaining focus on outcomes. While these methods are illustrated primarily through our experiences with large-scale migrations, they are equally effective for any cross-team product delivery involving complex and diverse systems.</p><h3>Why bother reducing complexity?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Tdqsu2P4ICFh7KdF8kmQoQ.png" /><figcaption><em>Shield / Rocket / Coins / Lightning mapping to Risk, Success, Cost, Innovation</em></figcaption></figure><ol><li><strong>Reduced Risk</strong>: Fewer moving parts reduce the probability of critical failures or incidents, leading to more reliable operations.</li><li><strong>Higher Success Rates</strong>: Simpler programs are more likely to succeed, succeed faster and deliver intended outcomes.</li><li><strong>Cost Savings</strong>: Avoiding or removing unnecessary interdependencies cuts down on wasteful engineering capacity and lowers incident costs by mitigating risks and issues caused by unmanaged complexity.</li><li><strong>Faster Innovation</strong>: Eliminating complicated processes or systems speeds up decision-making and allows for more agile pivots.</li></ol><p>In an ideal world, tech products would be built to foster organic adoption. No need for massive, orchestrated programs or “forced” migrations. Same for products involving multiple teams: Those teams would all be part of the same tribe or alliance, making cross-alliance collaboration unnecessary. Unfortunately, real-world conditions often require orchestrated, cross-functional change. That’s why understanding complexity, its sources and types, and how to reduce it is crucial for large organizations looking to remain efficient, scalable, and innovative.</p><h3>Where to begin?</h3><p>HelloFresh’s <strong>Yearly Planning</strong> (YP) process is a powerful, company-wide mechanism that lets alliances crystallise goals, test feasibility with engineers, and outline high-level dependencies for the upcoming year. But YP can only drill so deep. The real hazard is the “unknown unknowns”, hidden couplings or legacy quirks no one realises exist until they bite. Even a thorough discovery phase that maps inter-team touchpoints, platform-adoption paths, and hard architectural limits won’t flush out every blind spot, yet it shrinks the risk envelope and reveals where program-level muscle is essential versus where squads can move on their own, prompting questions such as:</p><ul><li>Which strategies need a dedicated program?</li><li>Which squads must unify to build the core product?</li><li>Which teams can decouple and adopt the platform at their own tempo?</li><li>Where are the genuine technical interdependencies that force synchronous action versus areas suited to gradual transitions?</li></ul><h3>Organizational Complexity</h3><p>Organizational complexity manifests in how teams are structured, how decision-making authority is distributed, and how communication channels are configured. A matrixed or siloed organization often sees multiple squads or alliances each operating with different processes, tools, and timelines. Even within a single alliance, you might have sub-teams working at different cadences or with different skill sets.</p><p><strong>Key Challenges</strong></p><ul><li><strong>Multiple Streams of Work</strong>: Without a cohesive program structure, parallel initiatives can overlap or conflict.</li><li><strong>Undefined Boundaries</strong>: Ambiguity around what each squad owns, how initiatives are scoped and how they relate to another can lead to confusion and stalled progress.</li><li><strong>Too Many Cooks</strong>: Over-involving leaders or wrongly allocated stakeholders in the decision-making process and problem/solution definition can create slowdowns and muddy accountability.</li><li><strong>Communication Silos:</strong> When different teams rely on separate communication tools, processes, or forums, critical information can get lost or delayed.</li></ul><h3>Balancing Unification vs. Decoupling</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*86QJD7yYit2zDFOJSu24Bw.png" /><figcaption><em>Side-by-side graphic. Left: four linked squares climb upward together under one arrow (UNIFY). Right: one square rises independently while three remain, showing independence (DECOUPLE).</em></figcaption></figure><ul><li><strong>Unify</strong> teams when a coordinated push is needed for a successful outcome (e.g. deeply interdependent systems or big-bang changes).</li><li><strong>Decouple</strong> when teams can adopt solutions independently without compromising functionality or incurring unnecessary risk.</li></ul><p>As a foundational principle, decoupling workstreams and minimizing interdependencies fosters agility, reduces overhead, and mitigates the risk of large-scale failures. By limiting how one team’s deliverables can block another’s, the Program Strategy remains more adaptable, with each unit able to progress at its own pace and capacity.</p><p>Technical complexity heavily influences how far decoupling can realistically go. Architectural constraints or monolithic dependencies can naturally force certain teams to move in lockstep. To reduce organizational complexity, teams and leadership must thoroughly understand the technical underpinnings of each initiative. This enables them to identify potential bottlenecks, plan effectively for dependencies, and manage the organizational structure in a way that aligns with the actual capabilities and constraints of the systems involved.</p><p>We’ll explore an example of this interrelation in more detail in the Technical Complexity section in Part 3 of this series.</p><h3>Accountability, Decision-Making and Communication</h3><p>Reducing organizational complexity also means establishing clear accountability structures without overcomplicating them. The simplest model is often a “single-threaded leader,” one individual who holds ultimate responsibility yet need not directly manage all contributors. While a RACI framework can help sidestep the “too many cooks” dilemma and ensure each task or decision has an explicit owner, ensure that it does not become unwieldy and bloated with excessively detailed role definitions and tasks. Likewise, committees must remain lean and composed of true decision-makers to avoid diffusing authority. Top-down and bottom-up communication loops enable continuous feedback. Leadership sets strategic direction, while teams highlight practical insights and potential roadblocks. Having clear communication channels and processes with agreed-upon best practices become important in reducing alignment and information flow complexity. Embracing the principle that ‘transparent communication is the best communication,’ keep channels and documents as open as circumstances permit, barring sensitive or classified information.</p><h3>Change Champions &amp; Subject Matter Experts</h3><p>A great way to reduce organizational complexity is to embed “Change Champions” and SMEs into cross-functional teams as tribe representatives. These roles can participate in dedicated training and cascade knowledge down to their teams as well as raise issues uncovered by their teams or through their domain expertise to the Program Core Team. By embedding these experts directly, teams spend less time searching for information or looping in specialists as consultants and more time making informed decisions when executing the work. The Core Workstreams building the product also no longer need to respond to a constant queue of ad-hoc requests, as Change Champions can handle much of the load on the ground.</p><p><strong><em>Up next:</em></strong><em> We’ll leave the org chart behind and tackle operational levers: prioritisation, phased roll-outs, and gating in Part 2. Subscribe to get it the moment it drops.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cdf06c5a6ed9" width="1" height="1" alt=""><hr><p><a href="https://engineering.hellofresh.com/taming-complexity-hellofreshs-playbook-for-managing-large-scale-programs-part-1-3-cdf06c5a6ed9">Taming Complexity: HelloFresh’s Playbook for Managing Large-Scale Programs (Part 1/3)</a> was originally published in <a href="https://engineering.hellofresh.com/">HelloTech</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>