The Harry Potter Approach to Procurement (It’s the Long Game) (Marina Nitze)

Procurement reform does not happen overnight, but taking small steps can create large-scale change. Harry Potter beat Voldemort piece (horcrux) by piece, breaking down the unimaginable task into more manageable steps. Instead of spending years trying to document every last requirement for every nook and cranny of your existing legacy system before you update it, spend six months replacing and improving one well-defined piece of it. Then repeat. In much less time, you’ll have made huge progress.

Procurement may not be appreciated as the most sexy topic in public administration, but it is one of the areas most ripe for seismic transformation. Here are six ideas that can be implemented now:

• The best procurements are tied to Key Performance Indicators (KPIs) that impact real people (constituents or employees- ideally, both). Nobody cares how many lines of code are in your application or what programming language it's written in. Can you measure what your contract deliverables are achieving? (That’s deliverable #1: instrumentation) Is this achievement objectively better than before? If not, why are you doing it? You cannot improve an IT system without improving its associated business processes. This makes for a more complicated and intertwined development process, but infinitely better outcomes.
• Beware of poorly-defined goals like “modernization.” In fact, ban that word altogether. In 2008, the U.S. Department of Labor gave states $500M to “modernize” their unemployment systems. It may surprise you to learn that during the COVID-19 pandemic, more than half of states boasted that their UI systems were “fully modernized.” Did it seem that way? No. This disconnect was because the definition of “modernized,” for purposes of receiving this funding, was “capable of calculating an Alternate Base Period” as opposed to getting the majority of deserving claimants their money within days, letting constituents check their own claim status, or having a website without business hours. The Administration for Children and Families (ACF) is similarly hemorrhaging cash by failing to define “modernization” for child welfare IT systems at all, so the same vendors are pocketing hundreds of millions of dollars to rewrite the same bad processes and forms from Pascal into Java, instead of achieving improved outcomes for children or families.
• Internal talent comes first. A great procurement doesn’t require you to have all the expertise on the inside; if you had that, you wouldn’t need to procure it. But you need enough internal expertise to write accurate, achievable, yet ambitious requirements and performance goals; to vet applicants; to fairly negotiate with vendors when changes inevitably need to be made; and to hold vendors accountable for performance. If you don’t have at least that much, focus on filling that internal talent gap before you award $100 million to outsiders whom you literally can’t evaluate.
• Contracting officers are not robots—they’re human beings who want to be part of achieving your goals and mission. The single best piece of professional advice I’ve ever received was from my predecessor at the VA, Peter Levin, who told me I should drive to VA’s Technical Acquisition Center (TAC) in Eatontown, New Jersey and buy everyone lunch. I eventually stumbled my way north to them and fretted far too long on how to determine what they liked to eat. We ended up discovering a shared love of Brussels sprouts and helping veterans, and my team never did a procurement without going in person to New Jersey from that point on.
• The lawyers, budget analysts, contracting officers, and others who work on procurements all have their own risk and incentive frameworks. I have learned the hard way that “number of dead veterans” or “number of homeless foster children” are not criteria on most approval forms. You can complain about this, or you can fill out the required form so there will be fewer dead veterans and homeless children. Find out what criteria will de-risk your mission-critical procurement (proof points from other governments? ISO certifications? adherence to NIST IAL2/AAL2 standards?) from the perspective of your decision-makers, and fulfill them.
• If you want to change how the big vendors perform, you first have to change contract performance criteria. These kinds of antiquated criteria don’t de-risk your project; they introduce risk by guaranteeing you’ll only get old guard applicants. Instead, removing requirements that unfairly keep out new entrants, like the minimum number of years in business or millions of dollars of past performance, will help increase the playing field and create more alternatives to existing vendors. If you don’t know what’s keeping new entrants from bidding, ask them.

We don’t have to change everything all at once. Let’s pick a few of the above practices and start measuring major procurements against them on a dashboard, a form of “positive peer pressure” (Ten points for Gryffindor!) I’ve found it useful in child welfare. Then, make it drop-dead simple to adopt each practice, such as with copy-and-pastable contract language from other states.