<?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 David Pritchett on Medium]]></title>
        <description><![CDATA[Stories by David Pritchett on Medium]]></description>
        <link>https://medium.com/@pritchdf?source=rss-fa12629cba30------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*Ftf2pOBE5jC0xP21</url>
            <title>Stories by David Pritchett on Medium</title>
            <link>https://medium.com/@pritchdf?source=rss-fa12629cba30------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 15:51:25 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@pritchdf/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[You Can’t Integrate the Contract Without Integrating the Government]]></title>
            <link>https://medium.com/@pritchdf/you-cant-integrate-the-contract-without-integrating-the-government-31ff6ef9297f?source=rss-fa12629cba30------2</link>
            <guid isPermaLink="false">https://medium.com/p/31ff6ef9297f</guid>
            <category><![CDATA[organizational-change]]></category>
            <category><![CDATA[it-leadership]]></category>
            <category><![CDATA[digital-transformation]]></category>
            <category><![CDATA[it-service-management]]></category>
            <category><![CDATA[government-contracting]]></category>
            <dc:creator><![CDATA[David Pritchett]]></dc:creator>
            <pubDate>Thu, 19 Nov 2026 02:12:58 GMT</pubDate>
            <atom:updated>2026-11-19T02:12:58.659Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Beyond the Contract: What It Takes to Own Federal Technology — Article 3</em></p><p>Over several years at the Environmental Protection Agency (EPA), we changed how we contracted for technology services. We started with separate contracts for hosting, networking, applications, and security, with government staff tying the pieces together. Over time, we consolidated related services under fewer contracts and added a requirement that contractors coordinate across the remaining boundaries.</p><p>It worked. Our contractors were now responsible for delivering integrated services. But our government teams were still organized around individual technologies.</p><p>Hosting, networking, applications, and security each had separate responsibilities and established ways of operating. People in those areas generally worked together when needed. But decisions were often made inside one technical area without fully considering their effect on other services. Working together is not the same as sharing accountability for a service. We had improved how our contractors worked together. Our government organization hadn’t kept pace.</p><p>Cooperation isn’t integration. When priorities conflicted, resources were tight, or a major incident crossed technical areas, the question became:</p><p><strong>Who owns the service?</strong></p><p>Not the server. Not the network. Not the application. The service.</p><p><strong>Creating the Enterprise Operations Division</strong></p><p>The Enterprise Operations Division (EOD) was formed as part of the reorganization of EPA’s Office of Mission Support. It was more than a merger of the existing hosting and networking divisions. We reconsidered how the organization should be structured to support enterprise service delivery.</p><p>Rather than combining existing branches under common management, we reorganized their responsibilities into four new branches:</p><ul><li><strong>Customer Service</strong> managed customer relationships, service requests, service delivery coordination, and project management.</li><li><strong>Infrastructure Engineering</strong> managed engineering, architecture implementation, technical standards, and modernization across all our technical stacks.</li><li><strong>Infrastructure Operations</strong> managed general-purpose infrastructure and cloud operations, including our substantial on-site systems footprint, shared storage, web applications, and supporting infrastructure.</li><li><strong>Platform Operations</strong> managed specialized enterprise platforms and services, including Salesforce, cloud.gov, high-performance computing, and geospatial capabilities.</li></ul><p>Engineering had one of the most challenging assignments. Years of technical silos had produced independent technology stacks that needed integration into a better service delivery model. Engineering was responsible for identifying common requirements, integrating capabilities, and, where appropriate, retiring redundant systems.</p><p>Monitoring was one example. Rather than maintaining separate technologies providing similar capabilities across different environments, we wanted to establish common approaches that better supported the enterprise.</p><p><strong>Build, Then Run</strong></p><p>Separating engineering from operations was deliberate.</p><p>Engineering needed to focus on modernization and integration, but the services still had to run every day. Operations was responsible for sustaining those services. Engineering was accountable for developing and modernizing capabilities, while Operations was accountable for sustaining them. Neither worked independently. Engineering involved Operations in designing capabilities that could be supported, and Operations engaged Engineering when changes required broader technical integration.</p><p>This separation was intended to give modernization the attention it needed without compromising ongoing operations.</p><p>There was another reason, and it mattered to me as much as the first.</p><p>The two branches offered different kinds of technical work. Operations developed deep expertise in the technologies that kept the environment running. Engineering worked across technology stacks, designing how those pieces fit together. Moving between branches let personnel broaden their experience and develop new skills.</p><p>That mattered because of how the reorganization happened. Staff didn’t choose their initial placements, and we had limited time to fill each position. I wanted people to have a path toward different kinds of work if they wanted one, rather than staying wherever the reorganization placed them.</p><p><strong>Coordination by Design</strong></p><p>But the most important change wasn’t how we divided responsibilities. It was how we expected the branches to work together.</p><p>Just as we had required contractors to work across contract boundaries, we required our branches to work across organizational boundaries to ensure service delivery. That expectation applied to every branch, not just engineering.</p><p>Engineering had to include Customer Service, Infrastructure Operations, and Platform Operations in its workflows so anything it created, modernized, or eliminated was viable from every service-delivery perspective. When Operations planned a change affecting customers, Customer Service needed to be involved so customers understood what was happening. When Platform Operations needed a new authentication capability, Engineering ensured integration into the broader enterprise environment. When Customer Service received a request for a specialized capability, Engineering, Operations, and potentially Platform Operations needed to participate before implementing anything.</p><p>These weren’t simply opportunities for consultation after decisions had been made. The appropriate branches were expected to participate throughout the work. Each branch retained responsibility for its area of expertise, but that responsibility also required working with the others.</p><p>The coordination wasn’t only about avoiding mistakes. It was also intended to encourage innovation. When designers, operators, and customer-facing teams work together on the same problem, they can identify opportunities that might otherwise remain hidden within individual organizations.</p><p>We weren’t trying to eliminate specialization. We were changing how those specialties were expected to work together.</p><p><strong>Changing How Technical Decisions Were Made</strong></p><p>Establishing expectations for cross-branch coordination was one thing. Changing practices built over years of operating independently was another.</p><p>Many of our technical personnel had deep experience managing their environments and longstanding relationships with particular contractors and colleagues. Those relationships were valuable. Sometimes they also reinforced approaches that no longer served the broader organization.</p><p>One team used a contractor to maintain a spreadsheet of customers and associated tasks. The environment had grown to thousands of customer names and tens of thousands of configuration items. Multiple versions circulated through email, and information became outdated almost as soon as it was updated.</p><p>Establishing a reliable view of the work was increasingly difficult.</p><p>Meanwhile, we had a database team that could help develop a centralized reporting solution that would allow customers to enter and validate their own information.</p><p>The spreadsheet continued anyway because it was familiar and relied on established working relationships. The problem wasn’t the spreadsheet itself. The approach no longer supported the scale of the work, and the team wasn’t considering resources available elsewhere in the organization.</p><p>I want to tell you how that one ended. I can’t. The effort was still underway when I left. We brought in multiple project managers to support it, and as best I know, the result was a more complex spreadsheet.</p><p>That wasn’t the ending I wanted.</p><p>But it illustrates how difficult established practices can be to change. A new organizational structure, access to technical expertise, and additional project management weren’t enough to overcome a familiar way of working.</p><p>The experience reinforced a principle we were trying to establish:</p><p><strong>Routine technical decisions belong with the people closest to the work. Decisions affecting other services or shared capabilities cannot be made in isolation.</strong></p><p>The answer wasn’t to elevate every decision to senior management. That would have created another bottleneck. However, some decisions did move. Consolidation meant certain decisions were now made at a more senior level for the entire contract or enterprise service.</p><p>Some technical personnel felt they were losing control of systems they had managed for years, even though the objective was to improve how those systems supported the enterprise. Not all of that pushback was obstruction. Some reflected legitimate concerns about changing responsibilities and decision rights. We needed to understand those concerns without allowing them to prevent the organization from adopting better ways of working. We needed their technical expertise. We also needed them to consider the broader implications of their decisions.</p><p>Any integration effort that ignores what people believe they are giving up will encounter resistance it doesn’t fully understand.</p><p><strong>Changing the Org Chart Is Only the Beginning</strong></p><p>Creating EOD was one step in a transformation that took years, and we couldn’t stop delivering services while we changed how the organization worked. Combining organizations doesn’t automatically create service ownership, just as combining contracts doesn’t automatically create contractor coordination. If decision rights, management expectations, and established ways of working don’t change, old silos can persist within a larger organization. Looking back, consolidation wasn’t the solution on its own. It gave us a structure that better supported integrated service delivery.</p><p>The real work was changing how people made decisions and worked together.</p><p>And it raised the next question.</p><p>Once an organization is responsible for an integrated service, it can’t let every team or contractor make decisions based solely on what works best in its own environment. Someone has to understand how the entire technology environment fits together.</p><p>That takes architecture, and it leads to another lesson:</p><p><strong>You can’t outsource responsibility for understanding how your enterprise fits together.</strong></p><p><em>Next: Architecture Before Acquisition.</em></p><p><em>David Pritchett retired in December 2025 as Director of the Enterprise Operations Division at the U.S. Environmental Protection Agency, where he led enterprise technology operations, cloud modernization, cybersecurity initiatives, and service delivery organizations. He writes about enterprise technology leadership, organizational transformation, and the human side of leading change.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=31ff6ef9297f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Contract Boundaries Aren’t Service Boundaries]]></title>
            <link>https://medium.com/@pritchdf/contract-boundaries-arent-service-boundaries-27782504de3c?source=rss-fa12629cba30------2</link>
            <guid isPermaLink="false">https://medium.com/p/27782504de3c</guid>
            <category><![CDATA[digital-transformation]]></category>
            <category><![CDATA[it-service-management]]></category>
            <category><![CDATA[it-leadership]]></category>
            <category><![CDATA[government-contracting]]></category>
            <category><![CDATA[it-governance]]></category>
            <dc:creator><![CDATA[David Pritchett]]></dc:creator>
            <pubDate>Fri, 13 Nov 2026 03:37:58 GMT</pubDate>
            <atom:updated>2026-11-13T03:37:58.631Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Beyond the Contract: What It Takes to Own Federal Technology — Article 2</em></p><p>In the first article of this series, I described how our approach to federal technology contracting created an unintended challenge. We had separate contracts supporting different parts of our enterprise technology environment, and each contractor was responsible for meeting the requirements within its own scope of work. The arrangement was intended to encourage competition and provide access to specialized expertise, but it also left government personnel responsible for coordinating the contractors whenever a problem crossed those boundaries.</p><h4><strong>Reducing the Number of Contract Boundaries</strong></h4><p>As we gained experience managing this environment, we began looking for ways to reduce the coordination burden. One of the first changes we made was combining services with natural operational dependencies. We brought hosting and application support together, as well as networking and security services. We thought that if related technologies were supported under the same contract, some of the coordination problems we had experienced would be resolved within the contractor organization rather than requiring government intervention.</p><p>The approach helped. Troubleshooting became easier when the same organization supported more of the technologies involved in a problem. It also gave us a single contractor program manager to hold accountable for work that previously involved multiple contracts.</p><p>However, consolidation introduced its own challenges. Our contracting office raised concerns about how larger contracts could affect small-business participation. Given the size of our contracts, small-business subcontracting requirements were an important consideration in our acquisition strategies. Combining requirements could improve operational coordination, but it could also reduce opportunities for smaller companies to compete independently as prime contractors for portions of the work. We weighed those considerations as we developed our acquisition strategies.</p><p>We also discovered that combining related services didn’t eliminate the coordination problem. We still depended on other contractors supporting enterprise services outside the boundaries of our newly consolidated contracts. No matter how much we consolidated, situations still arose where multiple providers needed to work together.</p><h4><strong>Making Coordination a Contractual Requirement</strong></h4><p>Rather than continuing to assume that coordination would occur when responsibilities overlapped, <strong>we began explicitly requiring it.</strong></p><p>Working with our contracting office, we modified the language in our requests for proposals (RFP) and requests for quotations (RFQ) to require contractors to coordinate with other contractors supporting the service. We incorporated those requirements into the performance work statements (PWS) of the resulting contracts. We also worked to make this a consistent requirement across our acquisitions rather than something we addressed only in particular contracts.</p><p>The intent wasn’t to make one contractor responsible for another contractor’s performance. It was to establish that resolving a service problem could require working directly with another provider, even when the underlying issue might ultimately fall outside an individual contractor’s responsibilities.</p><p>Over time, we continued consolidating more of our technology services. Applications, hosting, networking, and other capabilities eventually operated within a more integrated contractor delivery model. That further reduced the number of contractual boundaries, but it also reinforced the importance of the coordination requirements we had established.</p><p>One example involved our enterprise productivity services, which were supported through a separate end-user services contract. When users experienced connectivity problems, the issue could involve either the end-user environment or the enterprise network. Previously, government personnel might have needed to coordinate separate investigations, exchange information between contractors, and keep troubleshooting moving. With the new coordination requirements, the contractors were expected to work directly with one another. Neither could report that its own systems appeared healthy and consider its involvement complete.</p><p>Requiring contractors to work directly with one another didn’t remove the government from the process. Contractors had to keep government personnel informed while they investigated and worked to resolve an issue, and report back when the work was completed.</p><p>What changed was the government’s role in those interactions. We no longer needed to serve as the intermediary for every technical conversation, passing information from one contractor to another or coordinating each step of the investigation. The contractors could work directly with one another while the government maintained visibility into the work and its outcome.</p><p>We were changing how the work was coordinated, <strong>not who was ultimately accountable for the service.</strong></p><h4><strong>Coordination Is Not a One-Time Activity</strong></h4><p>What I learned, however, was that establishing this expectation was only the beginning. Coordination between working organizations is not something you implement once and consider complete.</p><p>People change on both sides of a contract. Government personnel move into new positions, contractor personnel change assignments, and new people arrive with experiences and expectations shaped by the organizations they previously worked in. We had to continually monitor how the contractors interacted and ensure that new personnel understood how their responsibilities fit into the larger service-delivery environment.<br>That didn’t mean we expected everyone to adopt our existing practices without question. Some of the people who joined our organization brought ideas that improved how we operated, and we eventually incorporated some of those ideas into contract modifications.</p><p>However, we couldn’t allow individual teams to operate independently of the rest of the service-delivery organization. A team might develop an approach that worked well within its own area. Still, that approach had to align with its contractual responsibilities and the needs of the other organizations supporting the service.</p><p>Finding that balance between encouraging improvement and maintaining coordination became an important part of managing our contracts.</p><h4><strong>Maintaining Visibility Without Managing Every Detail</strong></h4><p>We also needed a way to understand whether these expectations were being met and whether the overall program was healthy.</p><p>Our contracts included management reporting requirements within individual tasks, with reporting performed weekly or biweekly depending on the nature of the work. The intent was to keep tactical and localized issues visible to the people responsible for addressing them. We wanted problems resolved at the lowest appropriate level rather than having every issue escalated to senior management.</p><p>We incorporated quarterly contractor self-assessments into the contract requirements to provide a broader view of the program. These were completed before the award fee board and allowed the contractor to showcase its accomplishments, highlight improvements, and keep important or recurring issues in front of the government.</p><p>I wanted the self-assessments to do more than summarize completed activities. They helped us understand the program’s overall health, including how the contractor coordinated with other providers and addressed issues that extended beyond individual tasks.</p><p>The weekly and biweekly reporting helped us understand what was happening within the individual areas of work, while the quarterly self-assessments brought that information together at the contract level. Not every problem needed senior management attention, but recurring problems or issues affecting multiple areas of service delivery needed to remain visible.</p><p>This reporting structure helped us maintain oversight without taking responsibility away from the people closest to the work.</p><h4><strong>Integration Doesn’t Mean One Contract</strong></h4><p>Looking back, I wouldn’t conclude from this experience that putting everything under one large contract is necessarily the best approach. Larger contracts can increase dependence on a single provider, complicate transitions, and reduce opportunities for new or specialized companies to compete. Poor performance can also affect a much larger portion of the enterprise.</p><p>The objective shouldn’t be to eliminate contract boundaries. It should be to determine where those boundaries make sense and establish how work will be coordinated across them.</p><p>One challenge we faced was how slowly we could implement these changes. We could adjust existing contracts through modifications when appropriate, but the major structural changes we wanted generally had to wait until the next acquisition. Our contracting model evolved over several recompete cycles rather than through one comprehensive redesign.</p><p>Each change addressed problems we had encountered with the previous approach. Combining related services reduced some of the operational boundaries. Explicit coordination requirements helped us manage the remaining boundaries. Our reporting and oversight practices helped us sustain those improvements as personnel and circumstances changed.<br>But as our contractor delivery model became more integrated, another problem became increasingly apparent.</p><p>The government organizations responsible for overseeing those services were still largely organized around the individual technologies. We had spent years improving how our contractors worked together, but we hadn’t made equivalent changes to how the government managed the work.<br>In many respects, the contractor delivery model had become more integrated than the government organization responsible for it.</p><p>That realization led to the next stage of our transformation.<br>In the next article, I’ll discuss why integrating contractor delivery wasn’t enough and how we began changing the government organization to better support enterprise service ownership.</p><p><strong>You Can’t Integrate the Contract Without Integrating the Government.</strong></p><p><strong>About the Author</strong><br>David Pritchett retired in December 2025 as Director of the Enterprise Operations Division at the U.S. Environmental Protection Agency, where he led enterprise technology operations, cloud modernization, cybersecurity initiatives, and service delivery organizations. A Navy veteran based in North Carolina, he writes about enterprise technology leadership, organizational transformation, service delivery, and the human side of leading change. He is currently exploring senior leadership opportunities in federal technology, government contracting, and other mission-driven organizations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=27782504de3c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[When Every Contract Is Green but the Service Is Red]]></title>
            <link>https://medium.com/@pritchdf/when-every-contract-is-green-but-the-service-is-red-d250d8a05893?source=rss-fa12629cba30------2</link>
            <guid isPermaLink="false">https://medium.com/p/d250d8a05893</guid>
            <category><![CDATA[digital-transformation]]></category>
            <category><![CDATA[it-leadership]]></category>
            <category><![CDATA[government-contracting]]></category>
            <category><![CDATA[technology-leadership]]></category>
            <category><![CDATA[it-service-management]]></category>
            <dc:creator><![CDATA[David Pritchett]]></dc:creator>
            <pubDate>Sun, 08 Nov 2026 03:02:33 GMT</pubDate>
            <atom:updated>2026-11-08T03:02:33.450Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Beyond the Contract: What It Takes to Own Federal Technology — Article 1</em></p><p>Early in my 15 years managing federal technology contract requirements, we supported the major components of our enterprise services through separate contracts. One company supported hosting, another handled the network, another supported applications, and another handled security.</p><p>There was a reasonable business argument for structuring contracts this way. The intent was to encourage competition, provide access to specialized expertise, and potentially deliver better pricing. Each contractor had a defined scope of work, performance requirements, and responsibilities. From an acquisition perspective, the approach made sense.</p><p>What wasn’t fully accounted for was the amount of coordination required to make those individual services function as a single enterprise environment. We expected government personnel to bring everything together. That worked reasonably well when operations were normal, but it became much more difficult when something went wrong.</p><p>I remember one incident involving an application operating across two geographically separated data centers. The environment was designed with redundancy so maintenance could be performed on one side without interrupting the application and failover in the event of a problem. The databases operated in a primary-and-passive configuration, with replication between locations and a third system determining which database should be active.</p><p>During the incident, both databases became active, disrupting the application. The environment was designed to prevent that situation, but something had gone wrong with the configuration that maintained the intended relationship between the databases.</p><p>We brought together the teams responsible for the network, databases, security, and application to investigate. What made the situation frustrating was that each initially reported that the infrastructure within its area of responsibility appeared to be functioning normally. The network was available, the databases were running, and the security controls appeared to be operating as expected. Looking at the individual components, there wasn’t an obvious failure that explained why the application wasn’t working.</p><p><strong>The problem existed in how those components interacted.</strong></p><p>As I recall, the investigation ultimately identified a database configuration issue. Correcting it restored the intended operating arrangement and allowed the application to function normally.</p><p>What stayed with me wasn’t the specific configuration problem. It was how much coordination the government had to provide to identify and resolve an issue that crossed several technical and contractual boundaries.</p><p>This was not an example of contractors deliberately avoiding responsibility. Each organization was examining the technologies it had been contracted to support, and the information it provided was generally consistent with what it could see. The difficulty was that no single contractor was responsible for investigating the entire service across all those boundaries.</p><p><strong>That responsibility fell to the government.</strong></p><p>Our personnel had to bring the teams together, coordinate communications, determine what additional information was needed, and keep the troubleshooting effort moving. Rather than having the technical teams work directly across organizational boundaries as part of normal service delivery, the government often became the intermediary.</p><p>This wasn’t limited to one incident. We encountered variations of the same problem whenever an issue involved multiple contractors. Each company understood its contractual responsibilities, but the service depended on several companies working together. When that coordination didn’t happen naturally, government had to make it happen.</p><p>The arrangement also reflected how our government technology organizations were structured. Hosting, networking, applications, and security had separate organizational responsibilities. Those groups generally cooperated, but they also had their own priorities, management structures, and technical perspectives. In many respects, we had created similar boundaries on both sides of the government-contractor relationship.</p><p><strong>Over time, I began to question whether we were measuring the right things.</strong></p><p>Contract performance measures were necessary. We needed to know whether infrastructure was available, incidents were resolved within established timeframes, changes were completed successfully, and contractors were delivering the services we purchased. Those measures provided important information for managing performance and holding contractors accountable.</p><p>But they didn’t always tell us whether the customer was receiving a functioning service.</p><p>The database incident was a good example. Individual components could appear healthy while the application remained unavailable. Each contractor could point to performance information showing it met its responsibilities, yet the customer still had a problem.</p><p>That experience helped me recognize a distinction I hadn’t fully appreciated earlier in my career: <strong>a contractor can meet its component-level performance requirements without the overall service meeting the customer’s needs.</strong></p><p>As enterprise environments became more interconnected, this distinction became increasingly important. Applications depended on infrastructure, networks, databases, identity systems, security controls, and other shared capabilities. A change or failure in one area could affect services managed by several different organizations.</p><p>Customers didn’t see those organizational or contractual boundaries. They expected the application or service to work, and when it didn’t, they expected us to resolve the problem.</p><p>Looking back, I don’t believe the original contracting approach was inherently wrong. Competition and specialization were legitimate considerations, and separate contracts can make sense for many services. The problem was that we had focused heavily on defining and measuring the individual components without giving equal attention to how they would operate together.</p><p>We had effectively made the government responsible for integrating the contractors without fully accounting for what that responsibility required.</p><p>Our initial response was to start combining related services. We brought hosting and application support closer together, as did networking and security. We expected that reducing the number of contractual boundaries would improve coordination, and to some extent it did.</p><p>But we eventually discovered that combining services didn’t eliminate the need for coordination across the boundaries that remained. That led us to make another change: explicitly requiring contractors to work with other contractors supporting the same service.</p><p>It was an important step in how we managed enterprise technology delivery, but it wasn’t the last.</p><p>In the next article, I’ll discuss how our contracting model evolved, what improved as we consolidated services, and why we eventually concluded that contractor coordination needed to be a contractual responsibility, not an assumption.</p><p><strong>Contract Boundaries Aren’t Service Boundaries.</strong></p><p><strong>About the Author</strong></p><p>David Pritchett retired in December 2025 as Director of the Enterprise Operations Division at the U.S. Environmental Protection Agency, where he led enterprise technology operations, cloud modernization, cybersecurity initiatives, and service delivery organizations. A Navy veteran based in North Carolina, he writes about enterprise technology leadership, organizational transformation, service delivery, and the human side of leading change. He is currently exploring senior leadership opportunities in federal technology, government contracting, and other mission-driven organizations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d250d8a05893" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What Two Decades of Leading Enterprise Transformation Taught Me About Leadership]]></title>
            <link>https://medium.com/@pritchdf/what-two-decades-of-leading-enterprise-transformation-taught-me-about-leadership-7b8d3cfd6356?source=rss-fa12629cba30------2</link>
            <guid isPermaLink="false">https://medium.com/p/7b8d3cfd6356</guid>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[org-transformation]]></category>
            <category><![CDATA[federal-it]]></category>
            <category><![CDATA[technology-leadership]]></category>
            <category><![CDATA[continuous-improvement]]></category>
            <dc:creator><![CDATA[David Pritchett]]></dc:creator>
            <pubDate>Wed, 04 Nov 2026 23:44:55 GMT</pubDate>
            <atom:updated>2026-11-04T23:44:55.466Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>What Two Decades of Leading Enterprise Transformation Taught Me About Leadership</strong></p><p><em>The technology kept evolving. The leadership principles that mattered most proved far more durable — and I only fully understood them once the “fog of busy” finally lifted.</em></p><p><em>Part 12 of 12 — Enterprise Technology Transformation Series</em></p><p>Over this twelve-part series, I explored how enterprise technology organizations evolved from infrastructure management to service delivery, platform services, DevSecFinOps, and product-centric operating models. Each step reflected responses to changing technology, changing customer expectations, and changing business demands.</p><p>Looking back, however, those changes taught me something far more important than new operating models.</p><p>They changed how I think about leadership.</p><p>At the time, I was focused on solving the next problem, delivering the next project, responding to the next outage, or preparing for the next budget review. Like many leaders, I was caught in what I now call the fog of busy. There was little time to step back and recognize the broader patterns emerging across the organization.</p><p>Retirement has provided that perspective.</p><p>What follows are not lessons retirement taught me. They are lessons leadership taught me that I only fully understood once I had time to reflect.</p><h4>The Half-Million Dollar Lesson I Almost Learned the Wrong Way</h4><p>Early in my career, my instinct was to fix problems.</p><p>If something wasn’t working, I wanted to understand it, solve it, and move on. That approach served me well as an engineer and later as a technical manager.</p><p>Over time, I realized that solving today’s problem isn’t always the highest-value contribution a leader can make.</p><p>Developing people who can solve tomorrow’s problems is.</p><p>One experience shaped that realization more clearly than any formal training. It occurred during my first assignment managing a technology service at the EPA. My responsibilities included customer billing and reviewing billing disputes.</p><p>When a customer had questions about line items on their quarterly bill, they submitted a dispute, and when it was approved, it eliminated that expense. I believed the customer understood their billing, so I approved those disputes. I was providing good customer service. In reality, approving disputes without investigation would have resulted in approximately half a million dollars in lost revenue in a single quarter.</p><p>It immediately caught our business manager’s attention.</p><p>What happened next stayed with me for the rest of my career. He didn’t blame me. He sat down with me and explained how the business actually worked. Most billing disputes weren’t requests to eliminate charges; they were requests to understand them. The work had already been performed. The contractors had already delivered the service. Customers often wanted help validating the work, understanding the costs, or connecting the invoice to the value they had received.</p><p>He didn’t simply correct the billing errors caused by my decision; he developed my understanding.</p><p>That distinction between “correcting a decision” and “developing the person who made it” fundamentally changed how I approached my role as a leader.</p><h4>What Changed Because of That Conversation</h4><p>Rather than viewing billing only as a financial transaction, I began seeing it as part of service delivery. To improve our customers’ understanding, we invested in better dashboards, more meaningful metrics, more frequent customer reviews, and clearer explanations of how services were delivered and why they cost what they did. We helped customers understand that every identity service, network connection, logging platform, cloud resource, and infrastructure component they consumed contributed to the capabilities they relied on every day.</p><p>We also helped them understand the financial consequences of their own decisions. Leaving cloud resources running “just in case,” maintaining redundant or unnecessary environments, or delaying key decisions all carried costs. Once they could see those costs, conversations changed. Customers were no longer reacting to invoices. They were making better-informed business decisions.</p><p>What began as a mistake by one new account manager ultimately improved how the entire organization communicated with every customer.</p><h4>What I Learned About Developing People</h4><p>My instinct is still to solve problems. I haven’t outgrown it, and I don’t think I ever will. But I learned that leaders have a different and higher-value responsibility.</p><p>They develop people.</p><p>When someone on my team made a mistake, my first questions gradually shifted from what went wrong to what understanding was missing — the understanding that, if built now, would help that person make better decisions for the next five or ten years.</p><p>This also became my lesson to resist the temptation to fix the problem myself. I needed to create an environment where the right decision became the easiest decision to make — not through surveillance or control, but through shared understanding and clear principles.</p><p>It also meant creating an environment where thoughtful mistakes became learning opportunities rather than events to manage or minimize.</p><p>Most of the failures I witnessed over twenty years were not the result of bad people or bad intentions. They resulted from good people working inside systems that made the wrong decision the easiest one to make. In those situations, the leader’s job is not to find whom to hold accountable. It is to understand what made the wrong decision easy and to change that.</p><p>The technology discussed throughout this series continues to evolve. Leadership principles are far more durable.</p><h4>The People Behind the Work</h4><p>Looking back over more than two decades with the EPA, I am grateful for the opportunities I was given, the challenges I was trusted to solve, and — most importantly — the people I had the privilege to work alongside.</p><p>I worked with remarkable public servants who cared deeply about their mission and the people they served. Time and again, I watched talented people willingly step forward to support work that was far larger than their individual responsibilities — responding to September 11th, Hurricane Katrina, the Deepwater Horizon oil spill, the East Palestine train derailment, the California Wildfires, the Fukushima nuclear disaster, and countless other events that demanded more than any organizational chart required.</p><p>Those experiences shaped me as much as any technology initiative. I learned far more from the people I worked for, the colleagues I worked beside, and the people I had the privilege to lead than I ever learned from the projects themselves.</p><p>Perhaps the most important lesson was that leadership isn’t about having all the answers. It’s about creating an environment where good people can make good decisions, thoughtful mistakes become learning opportunities, today’s challenges become tomorrow’s organizational improvements, and where you provide the support needed to implement those improvements.</p><h4>The Fog of Busy</h4><p>Writing this series reinforced something I wish I had practiced more intentionally throughout my career.</p><p>The demands of leadership create what I now call the “fog of busy”. Another meeting, another outage, another budget decision, another personnel issue demanding immediate attention. It becomes genuinely difficult — not from lack of desire but from lack of capacity — to step back and recognize the larger patterns developing around you.</p><p>This series began as an attempt to organize my own thinking after stepping away from day-to-day leadership. What I did not fully anticipate was how clearly the individual experiences would connect into a coherent perspective once I had the time and space to examine them.</p><p>Reflection turns experience into judgment. I now see it as part of the operating discipline, not a retirement activity.</p><p>The fog of busy is real. It is not a character flaw or a failure of prioritization. It is the predictable consequence of leading organizations that have real operational demands every day. But it has a cost — patterns recognized too late, lessons relearned, and perspectives available only in retrospect.</p><p>If I were doing it again, I would build reflection into the operating rhythm, perhaps even daily. I would protect the time needed to examine patterns, challenge assumptions, and ask what the organization is teaching us before the next crisis consumes the lessons.</p><h4>Looking Forward</h4><p>As I begin the next chapter of my career, I am interested in helping organizations navigate the same challenges that shaped my leadership — organizational transformation, enterprise operations, DevSecFinOps, product-centric operating models, and developing the next generation of technology leaders. If you are working through any of these challenges and would find value in that conversation, I would welcome the connection.</p><p>The technologies will continue to evolve. The need for thoughtful leadership will not.</p><p>Thank you to everyone who followed this series, challenged my thinking, and shared your own experiences. The conversations reminded me that while every organization is different, many of the leadership challenges are remarkably similar.</p><p>I am looking forward to continuing those conversations.</p><p><strong>About the Author</strong></p><p>David Pritchett retired in December 2025 as Director of the Enterprise Operations Division at the U.S. Environmental Protection Agency, where he led enterprise technology operations, cloud modernization, cybersecurity initiatives, and service delivery organizations. A Navy veteran based in North Carolina, he writes about enterprise technology leadership, organizational transformation, service delivery, and the human side of leading change. He is currently exploring senior leadership opportunities in federal technology, government contracting, and other mission-driven organizations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7b8d3cfd6356" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Difference Between Running a Service and Owning One]]></title>
            <link>https://medium.com/@pritchdf/the-difference-between-running-a-service-and-owning-one-86944d04807f?source=rss-fa12629cba30------2</link>
            <guid isPermaLink="false">https://medium.com/p/86944d04807f</guid>
            <category><![CDATA[devsecops]]></category>
            <category><![CDATA[federal-it]]></category>
            <category><![CDATA[technology-leadership]]></category>
            <category><![CDATA[technical-debt]]></category>
            <category><![CDATA[product-ownership]]></category>
            <dc:creator><![CDATA[David Pritchett]]></dc:creator>
            <pubDate>Sun, 20 Sep 2026 17:58:56 GMT</pubDate>
            <atom:updated>2026-09-20T17:58:56.084Z</atom:updated>
            <content:encoded><![CDATA[<p><em>“You build it, you run it” was the right idea. But running a service is only one part of what long-term accountability actually requires — as two experiences from my career made unmistakably clear.</em></p><p><em>Part 11 of 12 — Enterprise Technology Transformation Series</em></p><p>For years, DevOps popularized a principle that not only improved operational efficiency but also highlighted the importance of long-term ownership in technology organizations.</p><p>“You build it, you run it.”</p><p>The premise was straightforward: the teams that build services should also operate them. Engineers who know their deployment decisions will wake them up at two in the morning will make better deployment decisions. Reliability improves. Response times shorten. The long-standing divide between the team that builds something and the team that has to live with what was built begins to close.</p><p>It was the right idea. It solved a real problem.</p><p>Over time, I came to believe that it was also incomplete.</p><p>Running a service means keeping it operational, but owning a service means being accountable for all aspects that sustain its long-term value, including those rarely visible on dashboards and whose effects may appear only years later.</p><p>Ownership doesn’t begin when you deploy a product. Ownership is what remains after everyone else has moved on.</p><h3>What Running Doesn’t Cover</h3><p>The responsibilities that come with running a service are visible and immediate. Availability. Performance. Security incident response. Keeping the lights on.</p><p>The responsibilities that come with owning a service extend into less visible territory.</p><p>Technical debt accumulates through decisions that were reasonable at the time — a workaround that deferred a harder architectural choice, an upgrade pushed to the next sprint, then the next quarter, then the next year. In a project model, whoever comes after inherits that debt. In a product ownership model, the team that made the decision is still present when the debt comes due — which profoundly changes how those decisions get made.</p><p>Workforce continuity is another dimension that product ownership makes visible. When a team disbands at project close, the institutional knowledge that made the system understandable and supportable leaves with the people. In product ownership, maintaining that knowledge — through documentation, cross-training, and deliberate knowledge transfer — is a continuous responsibility rather than a closeout task.</p><p>Vendor relationships also look different through a product ownership lens. A project team evaluates a vendor at a point in time. A product team has to evaluate what that vendor relationship will look like in year three or year five — what happens if pricing changes, the product roadmap shifts, or the vendor is acquired.</p><p>Customer expectations change. The service that met customer needs at launch will face different expectations within a year. Technology that was modern at deployment becomes antiquated. Capabilities that seemed optional become standard. Staying close enough to the customer to recognize those shifts before they become gaps, and having the authority and the sustained funding to respond, is the work of product ownership.</p><h3>What a Legacy Document System Taught Me About Technical Debt</h3><p>One of the clearest examples of long-term ownership failure I witnessed involved a legacy document-management system that had successfully served its original mission for years.</p><p>By the time modernization began, the underlying technology had become increasingly difficult to support. The original developers had moved on, and much of the architectural knowledge behind earlier decisions had left with them. Modernization required significant investment while the legacy environment still had to be operated, licensed, secured, and supported.</p><p>The key lesson wasn’t about one technology. It was about continuity. When knowledge preservation, lifecycle planning, and modernization are treated as future-project concerns instead of current ownership responsibilities, the organization eventually pays for the gap. Scarce legacy expertise becomes a risk concentration, and modernization becomes harder precisely when it is most necessary.</p><p>The original project had long since been completed. The responsibility had not.</p><h3>What an Orphaned Infrastructure Layer Taught Me About Ownership</h3><p>Technical debt was not the only ownership failure I experienced. Sometimes the failure wasn’t in the application layer at all; it was in infrastructure nobody realized they had stopped owning.</p><p>A mission-critical regulatory reporting system operated successfully for years. The application team implemented new workflows, maintained the software diligently, and did their jobs well. What they couldn’t do was fully administer or modernize the underlying infrastructure the application depended on.</p><p>After the original implementation contract ended, the transition left a critical infrastructure layer without the administrative access and institutional knowledge needed to manage it fully. The environment continued operating, but upgrades and modernization repeatedly hit constraints the application team lacked the authority or access to resolve.</p><p>The application team was doing its job. Accountability was missing below the application layer.</p><p>The implementation had ended. Operational responsibility remained. Ownership had not been preserved.</p><p>Product ownership cannot stop at the application layer. It must extend across the entire ecosystem that enables a service to operate, including the hardware and infrastructure the application depends on. When those layers are orphaned, the application team inherits constraints they did not create and cannot fully resolve.</p><h3>The Time Horizon That Separates Running from Owning</h3><p>Perhaps the greatest difference between running a service and owning one is the time horizon that accountability requires.</p><p>Running a service focuses primarily on today — availability, incident response, current service levels. Owning a service requires thinking about today, next year, and five years from now at the same time. It requires balancing the urgency of current operational demands against the longer-term investments that keep the service viable. That tension never fully resolves. Managing it deliberately, rather than defaulting to whichever pressure is loudest at the moment, is one of the defining challenges of product leadership.</p><p>The teams that navigated this transition most successfully shared one characteristic. They stopped thinking about the service as something they had built. They began thinking about it as something they were continuously responsible for improving.</p><p>The go-live date no longer represented the finish line.</p><p>It became the starting line.</p><p><em>Next in the series: The Leadership Lessons Organizational Transformation Taught Me — a capstone reflection on what two decades of leading enterprise change revealed about leadership, people, and the cost of the fog of busy.</em></p><p><strong>About the Author</strong></p><p>David Pritchett retired in December 2025 as Director of the Enterprise Operations Division at the U.S. Environmental Protection Agency, where he led enterprise technology operations, cloud modernization, cybersecurity initiatives, and service delivery organizations. A Navy veteran based in North Carolina, he writes about enterprise technology leadership, organizational transformation, service delivery, and the human side of leading change. He is currently exploring senior leadership opportunities in federal technology, government contracting, and other mission-driven organizations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=86944d04807f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why the Project Mindset Eventually Becomes the Obstacle]]></title>
            <link>https://medium.com/@pritchdf/why-the-project-mindset-eventually-becomes-the-obstacle-12be4ed200ec?source=rss-fa12629cba30------2</link>
            <guid isPermaLink="false">https://medium.com/p/12be4ed200ec</guid>
            <category><![CDATA[digital-transformation]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[technology-leadership]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[devsecfinops]]></category>
            <dc:creator><![CDATA[David Pritchett]]></dc:creator>
            <pubDate>Fri, 18 Sep 2026 03:09:38 GMT</pubDate>
            <atom:updated>2026-09-18T03:09:38.976Z</atom:updated>
            <content:encoded><![CDATA[<p><em>When the same modernization effort appears on the project list year after year, that is not a planning failure. It is a structural one.</em></p><p><em>Part 10 of 12 — Enterprise Technology Transformation Series</em></p><p>For most of my career, technology work was organized around projects.</p><p>A need was identified. A project was initiated. Requirements were gathered, funding was secured, a team was assembled, and work began. When the project delivered its agreed scope on time and within budget, it was declared successful.</p><p>Then the project ended, the team celebrated and moved on, and the funding cycle closed.</p><p>The technology just delivered became someone else’s responsibility to operate, secure, support, and eventually modernize. The organizational model then moved on to the next project, leaving behind a growing portfolio of systems that someone needed to maintain but no persistent team owned.</p><p>That model made sense for a long time. In the era of relatively stable infrastructure, infrequent software releases, and technology whose useful life could be measured in years, organizing work around discrete projects with defined beginnings and endings was appropriate.</p><p>It makes far less sense today.</p><h3>Why the Project Model Fails Modern Technology</h3><p>Cloud platforms do not have end dates. Security requirements evolve continuously. Customer expectations keep rising. AI capabilities that barely existed a few years ago are already becoming foundational to how organizations deliver services. Technology that is modern at deployment will require significant investment to remain effective within a few years—not because it failed, but because the world around it did not stand still.</p><p>The project model was built on an assumption the current technology environment no longer supports: that technology work has a natural completion point. That once a system is built and deployed, the major investment is behind it.</p><p>What actually happens is that the handoff from project to operations marks the beginning of a long, expensive, and often underfunded lifecycle. Development considers the work complete. Operations inherits responsibility for keeping it running. Security continues introducing new requirements. Business priorities continue to evolve. The technology itself never stops changing, but the organizational model behaves as if the work is finished.</p><p>Technical debt accumulates almost immediately after project closeout — not through negligence, but because the funding cycle that could have supported ongoing investment has closed and the team that understood the system well enough to make good decisions about it has disbanded.</p><h3>The Lesson an Invoice Taught Us</h3><p>One experience made the project-to-product distinction clearer than any framework ever could.</p><p>For one of our platform services, customers purchased a block of engineering hours. Internally, we managed that work using Agile practices and organized it into sprints. From the technology team’s perspective, the model worked well. Priorities could shift between sprints, work was continuously refined, and engineering effort was visible and accounted for.</p><p>From the customer’s perspective, the experience was very different.</p><p>Customers received invoices for engineering hours but understood progress through the capabilities delivered at sprint reviews. Many struggled to connect the hours they were paying for with the outcomes they expected. The gap between how the technology team measured work and how the customer experienced value was significant — and it created friction that consumed considerable time and eroded trust.</p><p>That experience crystallized something fundamental for me:</p><p>Customers may be billed for hours, but they judge value through outcomes.</p><p>The conversation that followed — working with our business office and our customers to reframe how we communicated value, estimated work, and connected engineering effort to business outcomes — changed how I think about product management. Technology naturally measures effort. Business naturally measures outcomes. Product management translates between the two.</p><h3>What Product Thinking Changes</h3><p>Shifting from project to product thinking changes more than the funding model, though it changes that too.</p><p>Project funding creates start-and-stop cycles. Technology receives investment during development and is then expected to sustain itself with minimal ongoing support. The obligation to operate, secure, modernize, and evolve the technology continues long after project funding closes. Product funding recognizes that technology is a sustained business capability rather than a one-time delivery — and funds it accordingly.</p><p>The harder conversation that product thinking requires is the one about total cost of ownership. Project thinking asks “what does it cost to build?” Product thinking asks “what does it cost to own?” That is a longer time horizon, a more complex answer, and a conversation that requires technology, finance, procurement, and business leaders to think together rather than sequentially.</p><p>Ownership also changes how teams define success. A project team succeeds when it delivers scope on time and on budget. A product team succeeds when the service continues to deliver value — which requires ongoing investment, ongoing attention to customer needs, and ongoing willingness to make the less urgent but more important improvements that sustain long-term capability.</p><h3>The Clearest Sign an Organization Is Ready</h3><p>Looking back, one of the most reliable signals that an organization is ready for product-centric thinking is when the same modernization effort appears on the project list year after year.</p><p>That is not a planning failure. The organization has correctly identified that the capability needs sustained investment. The failure is structural — it is still funding a continuous need as though it were temporary.</p><p>Projects can deliver capabilities. Product thinking creates the accountability and funding model needed to sustain their value.</p><p>That distinction, simple as it sounds, changes almost everything about how organizations plan, fund, and govern technology work — and why the shift toward product-centric operating models has become one of the most important changes in modern enterprise technology leadership.</p><p><em>Next in the series: From “You Build It, You Run It” to “You Build It, You Own It” — what long-term product ownership actually demands from teams and leaders.</em></p><p><strong>About the Author</strong></p><p>David Pritchett retired in December 2025 as Director of the Enterprise Operations Division at the U.S. Environmental Protection Agency, where he led enterprise technology operations, cloud modernization, cybersecurity initiatives, and service delivery organizations. A Navy veteran based in North Carolina, he writes about enterprise technology leadership, organizational transformation, service delivery, and the human side of leading change. He is currently exploring senior leadership opportunities in federal technology, government contracting, and other mission-driven organizations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=12be4ed200ec" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[DevSecFinOps Is Not DevOps with Finance Added]]></title>
            <link>https://medium.com/@pritchdf/devsecfinops-is-not-devops-with-finance-added-7ba03f265020?source=rss-fa12629cba30------2</link>
            <guid isPermaLink="false">https://medium.com/p/7ba03f265020</guid>
            <category><![CDATA[org-transformation]]></category>
            <category><![CDATA[tech-leadership]]></category>
            <category><![CDATA[devsecops]]></category>
            <category><![CDATA[devsecfinops]]></category>
            <category><![CDATA[finops]]></category>
            <dc:creator><![CDATA[David Pritchett]]></dc:creator>
            <pubDate>Fri, 18 Sep 2026 02:59:33 GMT</pubDate>
            <atom:updated>2026-09-18T02:59:33.467Z</atom:updated>
            <content:encoded><![CDATA[<p><em>The term is becoming common. What it actually represents — a fundamental shift in how enterprise technology organizations make decisions together — is less well understood.</em></p><p><em>Part 9 of 12 — Enterprise Technology Transformation Series</em></p><p>Any framework that acquires a name carries a risk.</p><p>Once it has a name, organizations can adopt the name without adopting what the name was meant to describe. DevOps became a job title before many organizations understood it as a cultural change. Agile became a process before many organizations understood it as a mindset. DevSecFinOps is following a similar path in some corners of the industry — being treated as an organizational chart change or a tooling decision rather than the fundamental shift in accountability it actually represents.</p><p>Having helped lead an organizational evolution that brought these disciplines into the same decision process — before we had a convenient label for it — I came to see DevSecFinOps less as a framework than as an accountability model.</p><h3>The Chain of Dependencies That Made It Necessary</h3><p>The case for DevSecFinOps does not start with a framework. It starts with a pattern that reveals itself as organizations mature their technology operating models.</p><p>Improving software delivery requires closer collaboration with operations — because the team that operates a service has essential knowledge about what makes it operable. Improving operational consistency requires platform services — because teams solving the same problems independently create the complexity that undermines consistency. Platform services require common security guardrails — because a shared platform’s value is undermined if every team implements security differently on top of it. Cloud adoption requires financial accountability — because consumption-based pricing makes the financial consequences of technology decisions immediate rather than deferred. Procurement decisions influence technology architecture — because the contract structures that govern how work gets done constrain what is technically possible. Business priorities influence every investment decision — because technology exists to serve mission outcomes, not as an end in itself.</p><p>None of those disciplines can optimize enterprise outcomes on its own. Each optimization in isolation creates a new constraint somewhere else in the system. The only way to make the system work is to have all relevant perspectives — development, operations, security, finance, procurement, business leadership — contributing to decisions together rather than sequentially.</p><p>That is what DevSecFinOps describes. Not a tool. Not a process. A way of making decisions.</p><h3>The Meeting Problem</h3><p>Getting those disciplines into the same decision-making process is harder than it sounds, and I want to be direct about why.</p><p>People are busy. Priorities shift. Operational demands rarely pause for strategic conversations. When a meeting competes with an urgent customer issue or an executive request, the meeting loses — and the urgent issue usually has more immediate and visible consequences than the strategic decision that was supposed to be made.</p><p>When decision-making meetings did happen in our organization, two patterns emerged that consistently undermined their effectiveness.</p><p>The first was absent decision-makers. Key stakeholders with the authority to actually make a decision were represented by someone without that authority. The meeting could analyze, discuss, and recommend — but could not decide. The decision got deferred, the analysis had to be repeated in a subsequent meeting, and the organizational cost of the delay accumulated.</p><p>The second was the opposite problem — over-attendance. Technology teams sometimes brought large numbers of staff to meetings on the premise that something relevant might come up. The intention was reasonable. The result was meetings of twenty people where six needed to make decisions, and the dynamic of a large group actively worked against the focused conversation that good decisions require.</p><p>The standard we eventually worked toward was straightforward in principle and difficult to maintain in practice. Every decision meeting needed a clear decision, the authority to make it, and the expertise to inform it — no more, no less.</p><p>When that standard held, decisions happened faster, everyone in the room understood the tradeoffs, and the organization moved forward with shared clarity about why. When it didn’t, the cost was repeated discussions, delayed decisions, and a growing reluctance among senior leaders to invest time in meetings that might not produce outcomes.</p><h3>What Shared Accountability Actually Means</h3><p>One of the most common misunderstandings about DevSecFinOps is that shared accountability means diffused accountability — that when everyone is responsible, no one is.</p><p>That is a real risk, but it is not what the model requires. Shared accountability means something more specific.</p><p>Business leaders remain accountable for business outcomes. Technology leaders remain accountable for technical implementation. Security leaders remain accountable for managing risk. Finance leaders remain accountable for financial stewardship. Each domain retains its own specialized accountability. What changes is that those leaders share responsibility for the overall outcome — not just their piece of the process — and that shared responsibility requires them to engage with each other’s constraints and priorities rather than optimizing their own domain at the expense of the whole.</p><p>The shift is from sequential accountability — where each function handles its piece and passes the result to the next — to concurrent accountability, where the functions that ultimately share responsibility for an outcome participate in the decisions that shape it.</p><h3>Why It Matters More Now</h3><p>Enterprise technology decisions have always had business consequences. What has changed is how fast and direct those consequences are.</p><p>In a consumption-based cloud environment, a technology decision made this morning can have measurable financial consequences by afternoon. A security posture decision made during system design can have regulatory consequences years later. A procurement structure that made sense for a stable technology environment can create constraints that slow organizations when the technology environment changes faster than the contract anticipated.</p><p>The disciplines that DevSecFinOps brings together are not new. The current technology environment demands bringing them together early — before decisions are made, not after.</p><p>Enterprise value is created at the intersection of those disciplines, not within the boundaries of any single one.</p><p><em>Next in the series: From Projects to Products — why the project mindset eventually becomes the obstacle, and what replaces it.</em></p><p><strong>About the Author</strong></p><p>David Pritchett retired in December 2025 as Director of the Enterprise Operations Division at the U.S. Environmental Protection Agency, where he led enterprise technology operations, cloud modernization, cybersecurity initiatives, and service delivery organizations. A Navy veteran based in North Carolina, he writes about enterprise technology leadership, organizational transformation, service delivery, and the human side of leading change. He is currently exploring senior leadership opportunities in federal technology, government contracting, and other mission-driven organizations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7ba03f265020" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Organizational Change That Improved Our Technology Decisions More Than Any Tool We Adopted]]></title>
            <link>https://medium.com/@pritchdf/the-organizational-change-that-improved-our-technology-decisions-more-than-any-tool-we-adopted-68f1e4caea2a?source=rss-fa12629cba30------2</link>
            <guid isPermaLink="false">https://medium.com/p/68f1e4caea2a</guid>
            <category><![CDATA[technology-leadership]]></category>
            <category><![CDATA[federal-it]]></category>
            <category><![CDATA[finops]]></category>
            <category><![CDATA[org-transformation]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <dc:creator><![CDATA[David Pritchett]]></dc:creator>
            <pubDate>Fri, 04 Sep 2026 12:16:54 GMT</pubDate>
            <atom:updated>2026-09-04T12:16:54.850Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Most technology leaders can tell you their cloud architecture. Fewer can tell you what it costs. Even fewer can explain why the business should care. Here is what changed when we stopped separating those conversations.</em></p><p><em>Part 8 of 12 — Enterprise Technology Transformation Series</em></p><p>Most technology leaders can tell you their cloud architecture.</p><p>Fewer can tell you what it costs.</p><p>Even fewer can explain why the business should care.</p><p>For most of the history of enterprise IT, that gap was manageable. Technology organizations were funded through capital budgets, costs were relatively predictable, and the financial consequences of individual technology decisions played out slowly enough that the separation between technology planning and financial planning was sustainable.</p><p>Cloud computing changed that equation in ways that many organizations are still working through. A configuration decision made on a Tuesday afternoon can change the cost structure by Friday. A new service spun up without proper lifecycle management can quietly accumulate costs for months before anyone notices. In a consumption-based model, the relationship between technology decisions and financial outcomes is immediate — and it requires the organizations responsible for each to be in much closer communication than traditional structures typically provide.</p><h3>What Embedding Actually Means</h3><p>One of the most valuable organizational changes we made during our transformation was not a new technology platform or architectural pattern.</p><p>It was ensuring that our Working Capital Fund business office worked directly alongside the technology organization — not as an external review function that evaluated technology investments after the fact, but as an embedded partner present in the daily work of planning, evaluating, and making decisions.</p><p>In practice, this meant daily interaction. Formal planning meetings where technology and business perspectives were present at the same table. Shared workspaces that made informal conversation natural rather than scheduled. And perhaps most importantly, the countless brief exchanges — what I came to think of as drive-by conversations — where a quick question got answered, an assumption got challenged, and a decision got better before it became a problem.</p><p>Those informal interactions turned out to be as valuable as any formal process we designed. They created a mutual understanding that formal processes rarely produce on their own.</p><h3>What Each Side Learned</h3><p>Technology teams that worked alongside business and procurement professionals developed a much deeper appreciation for the constraints those professionals operate within — budgeting cycles, acquisition regulations, fiscal year obligations, and the financial stewardship responsibilities that govern how federal organizations can spend money.</p><p>That understanding changed how technology teams approached decisions. Rather than viewing procurement timelines as bureaucratic obstacles, they began understanding why those timelines exist and how to work within them more effectively. Rather than treating budget conversations as something that happened elsewhere and arrived as a constraint, they began participating in budget formation in ways that made the resulting budgets more accurate and more useful.</p><p>Business and procurement teams, for their part, developed a much better understanding of technology lifecycles, cloud adoption patterns, and the pace of enterprise technology evolution. They became better equipped to ask the right questions about technology investments and to evaluate the answers. They developed enough context to distinguish between a technology team requesting flexibility because the work genuinely requires it and one requesting it because the planning was inadequate.</p><p>Neither side became experts in the other’s discipline. That was never the goal. The goal was for each to develop enough understanding to be a better partner — and that goal was achieved.</p><h3>The Governance That Followed</h3><p>Major technology investments in our environment required justification to the Working Capital Fund Board, executive leadership, the CIO, senior agency leadership, internal customers across the agency, and external partners including the General Services Administration.</p><p>That level of governance is significant. It demands that every investment demonstrate both mission value and financial responsibility — not one or the other, but both, in the same conversation, to stakeholders who care about both.</p><p>The embedding model made that possible in a way that a traditional separated structure could not. When technology and business professionals work together daily, they can explain each other’s reasoning, anticipate each other’s questions, and present a unified perspective on an investment rather than two separate perspectives that the governance body must reconcile.</p><p>The persistent tension — and it was never fully resolved — was the mismatch between the pace of technology evolution and the pace of procurement. Cloud services, platform capabilities, security tooling, and AI capabilities move faster than acquisition cycles were designed to accommodate. Technology organizations have a responsibility to work within procurement constraints. Procurement organizations have a responsibility to understand that technology constraints are sometimes driven by genuine operational urgency rather than impatience. Managing that tension together, rather than around each other, required the mutual understanding that the embedding model was designed to build.</p><h3>What It Actually Produces</h3><p>Today, many organizations describe this kind of integration as FinOps or DevSecFinOps. For us, it did not begin as a framework or a methodology. It began as a decision to put the right people in the same room and trust that proximity would produce the understanding that formal structures alone had not.</p><p>It produced better budgets. It produced better technology decisions. It produced a governance capability that could explain investments credibly to a wide range of stakeholders.</p><p>Most importantly, it produced better enterprise decisions — decisions that balanced mission value, technical capability, operational requirements, security, cost, and long-term sustainability in the same conversation rather than reconciling them after the fact.</p><p><em>Next in the series: DevSecFinOps — Where the Disciplines Finally Converge — why this concept represents something more fundamental than a new acronym.</em></p><p><strong>About the Author</strong></p><p>David Pritchett retired in December 2025 as Director of the Enterprise Operations Division at the U.S. Environmental Protection Agency, where he led enterprise technology operations, cloud modernization, cybersecurity initiatives, and service delivery organizations. A Navy veteran based in North Carolina, he writes about enterprise technology leadership, organizational transformation, service delivery, and the human side of leading change. He is currently exploring senior leadership opportunities in federal technology, government contracting, and other mission-driven organizations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=68f1e4caea2a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Hardest Part of Our Technology Transformation Wasn’t the Technology]]></title>
            <link>https://medium.com/@pritchdf/the-hardest-part-of-our-technology-transformation-wasnt-the-technology-b05ec50d63e1?source=rss-fa12629cba30------2</link>
            <guid isPermaLink="false">https://medium.com/p/b05ec50d63e1</guid>
            <category><![CDATA[change-management]]></category>
            <category><![CDATA[technology-leadership]]></category>
            <category><![CDATA[it-leadership]]></category>
            <category><![CDATA[digital-transformation]]></category>
            <category><![CDATA[leadership]]></category>
            <dc:creator><![CDATA[David Pritchett]]></dc:creator>
            <pubDate>Fri, 04 Sep 2026 01:41:11 GMT</pubDate>
            <atom:updated>2026-09-04T01:41:11.664Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Leadership gaps, staffing constraints, and resistance to change slowed an organizational transformation whose technical direction was sound. The lesson was about leadership capacity, not architecture.</em></p><p><em>Part 7 of 12 — Enterprise Technology Transformation Series</em></p><p>When I think back over the years of leading enterprise technology transformation at the EPA, the challenges that stand out most clearly are not the technical ones.</p><p>The architecture decisions were difficult. The platform migrations were complex. The security integrations required careful coordination across teams that had not previously worked closely together. All of that was hard, and all of it required real expertise.</p><p>But those challenges followed a recognizable logic. Engineering problems, however complex, respond to engineering approaches. You can analyze them, design solutions, test the solutions, and iterate. Progress is measurable.</p><p>The organizational challenges were of a different kind.</p><h3>What the Org Chart Couldn’t Fix</h3><p>Early in our transformation, one of the first significant leadership decisions was to redesign the organizational structure to align with the future operating model rather than the existing one. The intent was clear and reasonable — to signal the strategic direction, establish the leadership roles that would guide the transformation, and create the organizational clarity that people needed to understand where they were headed.</p><p>What happened next was a lesson I did not anticipate.</p><p>At one point, six of the nine key leadership positions in the new organizational structure remained unfilled for nearly a year.</p><p>Those vacancies were not a minor inconvenience. They created a leadership gap at exactly the moment the transformation needed leadership most. The organization still had to deliver services, respond to incidents, and support customers every day. Existing managers and technical leaders did what responsible people do — they focused on keeping operations running. The transformation work, which required sustained leadership attention to champion the vision, coach people through new expectations, and reinforce the direction when daily pressures pulled against it, lost the momentum it needed.</p><p>Looking back, I underestimated how much communication the transformation required. Not just explaining what was changing — which is the easier part — but continuously reinforcing why the organization needed to change and how each person’s role contributed to the future state. That kind of sustained communication requires leadership presence. Without it, the vision loses ground to the day’s urgency.</p><h3>The People Behind the Resistance</h3><p>One of the more important lessons I learned about organizational change is that resistance is rarely uniform.</p><p>Most people, in my experience, are willing to move forward once they understand where they are going and can see how they fit into the future organization. They have concerns, they ask questions, and they need reassurance — but they are fundamentally oriented toward moving forward.</p><p>A small number of people, for various reasons, are oriented toward preserving the existing model. Sometimes that is because they have built significant expertise in the current way of working and genuinely believe the new direction is wrong. Sometimes it is because the change threatens something specific about their role or their standing. Sometimes the reasons are more complex than either of those.</p><p>What made this dynamic consequential was not the number of people who resisted. It was the influence that a few skeptical voices could have when leadership capacity was already thin, and the future operating model had not yet become familiar.</p><p>I do not think those individuals were resistant to technology. Looking back, I believe most of them were trying to understand where they fit in the future organization while simultaneously managing the very real responsibility of keeping the current organization running. That distinction matters, because it changes how a leader should respond. The answer is not enforcement. It is clarity, communication, and enough leadership presence to make the future feel navigable rather than threatening.</p><h3>The Fog That Operational Demands Create</h3><p>There is a dynamic in any large organization that works against transformation, and it is neither malicious nor particularly unusual.</p><p>When people are already operating at capacity — and in most healthy technology organizations, they are — the time and mental energy required to learn new approaches, mentor teams through new responsibilities, and reinforce organizational change are genuinely scarce. Day-to-day operational demands are concrete, urgent, and immediately consequential if neglected. Transformation work is important, but it is rarely as urgent as the incident that needs to be resolved by the end of the day.</p><p>The result is that transformation consistently loses in the daily competition for attention, even in organizations where the leadership genuinely wants it to succeed. This is not a failure of will. It is a predictable consequence of how operational organizations are structured and incentivized.</p><p>The organizations that navigate this most successfully are the ones that build explicit capacity for transformation work — not by asking already-stretched teams to add it to their existing load, but by protecting time, providing resources, and ensuring that the leadership reinforcement needed to sustain momentum is consistent and visible.</p><h3>The Lesson That Changed How I Think About Leadership</h3><p>Technology transformations succeed because people choose to adopt them — not because organizational charts or new tools require them.</p><p>That sentence seems obvious once you say it. Acting on it consistently, in the middle of an organization that needs to deliver services every day while simultaneously changing how it works, is considerably harder than it sounds.</p><p>What it requires from leaders is a different kind of attention than technical leadership demands. Less focus on getting the architecture right, more focus on helping people understand why the architecture is changing and what it means for them. Less emphasis on the organizational chart as the driver of behavior, more investment in the human relationships and communication that actually shape how people make decisions.</p><p>The platforms we built were important. The automation we implemented was valuable. But the transformation that mattered most happened not in the technology itself but in how the people responsible for it understood their work and their role.</p><p><em>Next in the series: When Technology and Business Stopped Working in Separate Rooms — how embedding finance and procurement alongside the technology organization produced better enterprise decisions than any technology investment alone.</em></p><p><strong>About the Author</strong></p><p>David Pritchett retired in December 2025 as Director of the Enterprise Operations Division at the U.S. Environmental Protection Agency, where he led enterprise technology operations, cloud modernization, cybersecurity initiatives, and service delivery organizations. A Navy veteran based in North Carolina, he writes about enterprise technology leadership, organizational transformation, service delivery, and the human side of leading change. He is currently exploring senior leadership opportunities in federal technology, government contracting, and other mission-driven organizations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b05ec50d63e1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How We Stopped Solving the Same Problems Over and Over Again]]></title>
            <link>https://medium.com/@pritchdf/how-we-stopped-solving-the-same-problems-over-and-over-again-3a15a315c42e?source=rss-fa12629cba30------2</link>
            <guid isPermaLink="false">https://medium.com/p/3a15a315c42e</guid>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[devsecops]]></category>
            <category><![CDATA[technology-leadership]]></category>
            <category><![CDATA[engineering-services]]></category>
            <category><![CDATA[platform-engineering]]></category>
            <dc:creator><![CDATA[David Pritchett]]></dc:creator>
            <pubDate>Sun, 30 Aug 2026 22:30:39 GMT</pubDate>
            <atom:updated>2026-08-30T22:30:39.735Z</atom:updated>
            <content:encoded><![CDATA[<p><em>When every team independently builds its own monitoring, logging, and security tooling, the organization pays for it in complexity, cost, and technical debt. Platform services offer a different model.</em></p><p><em>Part 6 of 12 — Enterprise Technology Transformation Series</em></p><p>There is a pattern that appears in almost every large technology organization, and it tends to become visible only after it has already created significant problems.</p><p>Different teams, working independently, solve the same operational challenges in different ways. Each builds its own monitoring capability. Each deploys its own logging solution. Each manages identity differently, selects different security tooling, and implements configuration management using different tools and processes. Each solution addresses a legitimate business need. Each was built by capable people making reasonable decisions with the information available to them at the time.</p><p>Collectively, those independent decisions compound into something the organization never chose and never designed — an enterprise environment of duplicated capabilities, inconsistent processes, and technical complexity that grows harder to manage with every addition.</p><p>This is the problem that platform services exist to solve.</p><h3>The Cost of Independent Solutions</h3><p>The visible cost of independent solutions is duplication. Multiple teams paying for overlapping tools. Multiple sets of engineers maintaining overlapping capabilities. Multiple procurement processes for similar products. Those costs are real and worth addressing on their own.</p><p>The less visible cost is complexity. Every independently implemented capability creates its own set of integration challenges, its own operational processes, and its own learning curve for anyone who needs to work across team boundaries. Security audits become more difficult when every team has implemented security controls differently. Incident response becomes more complicated when logging is inconsistent across the environment. Onboarding new teams takes longer when there are no common platforms to build on.</p><p>In our environment — spanning multiple data centers, multiple cloud providers, and multiple Platform as a Service offerings — the cost of that complexity was particularly evident. The same operational challenge, solved a dozen different ways, created a maintenance burden that grew with every addition and made it increasingly difficult to deliver services consistently across the enterprise.</p><h3>The Question That Reframes the Problem</h3><p>At some point, the right question stopped being “how do we build and operate better services?” and became “how do we provide common capabilities that enable every service team to succeed?”</p><p>That shift in framing represents a significant change in thinking about what operations is for.</p><p>The answer was not to centralize decision-making — which tends to slow everything down and remove the local knowledge that makes teams effective. The answer was to centralize the capabilities that every service team depended on, while preserving each team’s ability to innovate where they created unique business value.</p><p>Instead of each team maintaining its own monitoring stack, a shared monitoring platform is available to every team — with consistent capabilities, shared operational overhead, and the ability to deliver enterprise-wide visibility that independent solutions cannot. The same logic applies to logging, identity management, security tooling, deployment automation, and the other foundational capabilities that every service needs but none needs to build independently.</p><h3>How the Journey Actually Unfolded</h3><p>The transition to platform services did not start with a grand strategy. It started with practical problems.</p><p>Enterprise logging and monitoring became one of the first shared platforms because the pain of inconsistency was most visible there. Incident response was harder when logs looked different across environments. Root cause analysis took longer when monitoring tools didn’t share a common language. Solving that specific problem — building a shared logging and monitoring capability that every team could use — turned out to be the foundation of something larger.</p><p>As that capability matured, it became easier to make the case for the next one. Identity management. Common security tooling. Standardized operational services. Cloud adoption accelerated the evolution by making shared platforms, automation, and reusable services both practical and cost-effective in ways that were much harder to achieve in pure on-premises environments.</p><p>Interestingly, at no point did we describe what we were doing as “building a platform.” Each initiative addressed a specific operational need. It was only in retrospect that the pattern became clear. Each shared capability was creating the foundation for the next, and collectively they were becoming an enterprise platform strategy that no single team had set out to build.</p><p>That observation matters for any organization considering this journey. Our experience taught me not to wait for the perfect enterprise platform strategy. A platform often begins when an organization recognizes that one capability is being solved repeatedly and decides to solve it once, well, for everyone. Strategy can grow from that evidence.</p><h3>What Changed Most</h3><p>The technology evolved. Shared platforms reduced fragmentation and improved enterprise consistency. But the more important change was what those platforms allowed teams to stop doing: repeatedly rebuilding undifferentiated capabilities that did not create unique mission value.</p><p>More importantly, the role of operations evolved with it.</p><p>Operations teams that had defined their purpose around managing infrastructure increasingly found themselves in the business of enabling other teams. The shift from “we run the servers” to “we provide the capabilities that let every team run their services effectively” is not a small change. It requires a different understanding of what operations is for, a different set of skills, and a different relationship with the development and security teams that platform services are meant to support.</p><p>That evolution — in the role of operations, in the identity of the people doing that work, and in the organizational culture that surrounds it — turned out to be considerably more difficult than the technology transition itself.</p><p><em>Next in the series: The Human Side of Transformation — why building the technology platforms proved easier than transforming the organization that would use them.</em></p><p><strong>About the Author</strong></p><p>David Pritchett retired in December 2025 as Director of the Enterprise Operations Division at the U.S. Environmental Protection Agency, where he led enterprise technology operations, cloud modernization, cybersecurity initiatives, and service delivery organizations. A Navy veteran based in North Carolina, he writes about enterprise technology leadership, organizational transformation, service delivery, and the human side of leading change. He is currently exploring senior leadership opportunities in federal technology, government contracting, and other mission-driven organizations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3a15a315c42e" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>