<?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 Rafael Proenca on Medium]]></title>
        <description><![CDATA[Stories by Rafael Proenca on Medium]]></description>
        <link>https://medium.com/@rafaelproenca?source=rss-f89411522ce7------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*O-gK9NCzcmT1L7eL.jpg</url>
            <title>Stories by Rafael Proenca on Medium</title>
            <link>https://medium.com/@rafaelproenca?source=rss-f89411522ce7------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 17:38:34 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@rafaelproenca/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[Takeaways from DPI670M]]></title>
            <link>https://medium.com/@rafaelproenca/takeaways-from-dpi670m-1c44007b28e8?source=rss-f89411522ce7------2</link>
            <guid isPermaLink="false">https://medium.com/p/1c44007b28e8</guid>
            <category><![CDATA[procurement]]></category>
            <category><![CDATA[digital-government]]></category>
            <dc:creator><![CDATA[Rafael Proenca]]></dc:creator>
            <pubDate>Wed, 14 Mar 2018 18:57:58 GMT</pubDate>
            <atom:updated>2018-03-15T21:31:06.642Z</atom:updated>
            <content:encoded><![CDATA[<ol><li><strong>Summarize</strong></li></ol><p><strong>Writing RFPs</strong></p><p>Before writing an RFP, the two main questions a government worker must ask is: i) What am I really trying to achieve with this purchase? and ii) How much time and resources do I have to put into the procurement/development process?</p><p>The first question is fundamental to make the procurement process more strategic. It might look like an obvious question, however, far too many government procurement projects fail to consider deeply what they are really trying to achieve and end up with an inputs/process-focused model. Ignoring this question, projects tend to fall into the “business as usual” model of operation. In business as usual, procurement officials take the tried path and rely heavily on past RFPs as templates for new RFPs. This seems like the easy way to go, as it demands less time from overworked procurement personnel and minimizes risks of a catastrophic failure –what has worker in the past will rarely fail spectacularly. However, this approach curbs innovation and ends up delivering less public value. By focusing on inputs and processes rather than outcomes, it relegates the user to after-thought.</p><p>On the other hand, asking “what am I really trying to achieve?” forces procurement officials to put users at the center, making they think about how what they are buying will create public value. Starting with this question, procurement officials might feel compelled to go ask user about what they really need. Having user interviews and testing assumptions early on is a core principle of agile. Frequently, the questioning process will show that mockups or an MVP is necessary to really understand user needs and how they would interact with the new system the government is trying to buy. A productive back-and-forth between procurement officials, users and potential suppliers will ensue. In this process, procurement officials will most likely be surprised by how different from their initial assumptions reality is.</p><p>The second fundamental question is “am I the right person/agency to conduct this procurement?”. Too many government procurement projects “reinvent the wheel”. In Brazil, its 5,565 municipalities have bought 3,876 personalized fiscal management software suits in the past 5 years. Conducting agile procurement process demands a time from qualified and motivated government officials. These small municipalities frequently lack the expertise and the time to properly supervise agile development. Further, low-value purchases attract less suppliers, decreasing competition and the likelihood the best vendors will bid. Municipalities end-up paying too much for low-quality software. Bundling together the needs of several municipalities would have saved a huge amount of public funds and made agile development more likely.</p><p>Even if a buyer concludes he is the right person to conduct that procurement process, he still gains by asking “am I the right person/agency to conduct this procurement?” at the start. The question compels government buyers to research what other government agencies conducted similar procurement processes recently, uncovering valuable lessons from colleagues’ experience, increasing learning and producing better outcomes.</p><p><strong>Implementation and accountability</strong></p><p>The main questions to ask in the area are: i) am I testing early enough? ii) is there enough space to speak truth to power?</p><p>Testing early in settings as close as possible to real-world applications is one of the most crucial steps for the successful roll-out of software. This is accentuated if the system has to operate at scale. As pointed by Steven Brill in “America’s better pill” and confirmed by Henry Chao, lack of early testing was one of the main reasons for the disastrous launch of healthcare.gov.</p><p>Asking “Is there enough space to speak truth to power?”, and acting to create this space if the answer is negative, guarantees problems will surface and be addressed early on. Again, healthcare.gov illustrates how a failure in this aspect can lead to disastrous outcomes. Henry Chao, for example, knew of the problems with the website far in advance of the launch. However, the culture he was embedded in rewarded optimist and “pushing through” difficulties, rather than honestity and realism when it came to the chances the website would work as expected. This mentality prevented CMS and CGI from obtaining the resources they needed to make healthcare.gov work — time, the power to decrease requirements and focus on the core, the power to launch in phases, etc.</p><p>The rescue team, on the other hand, tackled this issue right at the start. Standing meetings led by Mickey</p><p><strong>2. Analyze</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DcfDzHKpzPA84ZkiYbTpiA.png" /><figcaption>Building successful software requires more than writing good code</figcaption></figure><p>To ensure success in a government IT project, we must remember that software is a system that comprises computer code and human behavior, and that neither part can work without the other. Even if we have the perfect code, the prettiest interface, and the best hardware, a new software system will not work unless users understand, like and adapt their behavior to it. Most IT failures are neither in the IT side or in the human side of the system, but in the interaction of the two.</p><p>In the failure of healthcare.gov it was clear that problems only became obvious once people started trying to buy insurance in the website. That’s why agile emphasizes early testing so much: only by having real users interact with code can developers see what users really need and how their behavior affects the underlying software/hardware infrastructure.</p><p>Rohan Bhobe — CEO of Nava Technologies, an innovative government IT vendor who relies heavily on agile — reports that in many of their projects writing code only starts at the second half of the project. They would spend the initial half only talking to users and devising new processes to achieve the desired outcome. For example, when helping the Department of Defense improve their system for storing documents, they first spent time understanding why paper storage had been so important to them, and how to optimize the flow of documents to make storage more efficient. Only after having a good draft of a better process and having experimented in the field with paper-based mockups they would start writing code.</p><p><strong>3. Synthesize</strong></p><p>To me, the HKS class where I most used the lessons from DPI670 is “PED 250 — Second Year Policy Analysis Seminar”. As part of my SYPA (Second Year Policy Analysis Exercise), a colleague and I helped the Brazilian Office of the Controller General of the Union (CGU) — an auditing and consulting unit under the Ministry of Transparency (MoT) — improve efficiency of procurement of simple goods and services. The lessons of DPI670 around cultural change in government procurement have been extremely valuable to this task.</p><p>After interviewing 15 procurement experts in Brazil, we identified that a culture of fear — <em>cultura de medo, </em>as expressed by the experts — was one of the main causes of public money waste in procurement. This problematic lack of trust between controllers and procurement agents stems from an over-emphasis on fighting corruption. A succinct quote from an expert from the Ministry of Planning captures this lack of trust: “Procurement agents are treated as ‘guilty’ until proven innocent”. Auditors from CGU agreed that there was a high level of distrust towards procurement agents in the field.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aA6Z6HhzCYJAeu9Xpw5mPg.png" /><figcaption>Key quotes on the culture of fear in Brazilian procurement</figcaption></figure><p>Specifically, experts from units we interviewed cited <em>cultura de medo</em> as a major source of waste. In follow-up questions, we asked if the distrust was warranted (i.e. there was a lot of active waste) or if it caused more waste than it solved. The majority of experts believed that, for procurement of standardized goods and services, lack of trust creates more waste than the corruption it prevents, because: i) Procurement agents focus on following formal processes instead of getting the best possible outcome, ii) Procurement agents lose motivation and have little space to innovate and increase efficiency, iii) The efforts of auditors could often be better spent improving overall processes instead of trying to catch corruption in individual cases.</p><p>C<em>ultura de medo</em> was frequently cited as stifling innovation. We heard stories of experimentation that were made incredibly difficult due to this lack of trust. For example, running a small pilot for outcome-based payments for cleaning services required authorization from the Minister of Planning, and demanded many rounds of discussion with legal advisors. In another case, substituting security guards for much cheaper cameras exposed procurement agents to huge personal financial liability. This environment makes inertial –just copying previous contracts with few adjustments — the safe option for procurement officials.</p><p>An expert at the CGU stated that “Some people are bad, but most are good. We spend too much time trying to find the bad ones”. This statement resonates with international evidence showing that corruption usually accounts for about 20% of procurement waste while inefficiency accounts for the remaining 80% (Bandiera, Prat, and Valletti, 2009).</p><p>We learned that Lei 8.666 — the main Brazilian procurement law — was designed with large public-works projects in mind, and in this type of project active waste is indeed a large concern. However, for standardized goods and services, the numerous controls in the legal framework and the emphasis on active waste appears to be above optimum. Experts believed that the <em>cultura de medo</em> could have been caused by the prescriptive nature of Lei 8.666. Together with its accompanying regulatory environment, it creates incentives for auditors to catch specific instances where rules were not followed instead of systematically reviewing these rules, aggravating a power imbalance between auditors and buyers. This power imbalance is particularly large in Brazil since auditors tend to have passed more difficult entry-exams to join the public service, are significantly better paid, and have access to more resources to do their work.</p><p>The theme of a legalist culture in procurement generating waste also appears in the US experience. When the US Digital Services tried to implement FLASH, an innovative agile procurement framework for the Department of Homeland Security (DHS), the whole project was killed by appeals from vendors who did not win. The culture of enforcing fairness and fighting corruption at any cost can be expensive. Maybe we should do a cost-benefit analysis of corruption and fairness enforcement, keeping in mind that the ultimate goal of public policy should be creating public value to citizens.</p><p>The lessons of DPI670 can help transform the Brazilian culture of fear into a culture of performance in procurement. Agile procurement, with laser-sharp focus on delivering public value, would help increase efficiency and promote innovation in Brazilian procurement.</p><p>Discussions on how to change government procurement culture and make agile possible in DPI670 are also closely related to the lessons I took from “MLD 102: Getting Things Done —Management in Development Contex”. The class emphasizes the importance of being problem-driven, i.e., of starting any intervention by convening stakeholders and constructing together a problem. Problem construction is hard. It involves challenging assumptions, challenging authority and going to deep causes of government challenges. Instead of saying that we have fiscal problems because our fiscal management software is old, we must uncover why people do not use the old system as intended, and how exactly a new system will help improve the process. PDIA also emphasizes iteration. A project should evolve in cycles of planning, experimenting, measuring performance and planning again. Finally, PDIA emphasizes adaptation. Instead of imposing cookie-cutter solutions that worked elsewhere, truly innovative and effective solutions must be constructed in a local context through this iterative process I just described. These principles are very close to the principles of agile, and together empowered me to go back to government work and innovate effectively.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1c44007b28e8" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>