How I decided to go all-in on a federal contract — before it was assigned

After the American Rescue Plan Act passed in March 2021, the U.S. government expanded child tax credits to provide relief for American families during the pandemic. The legislation allowed some families to nearly double their tax benefits per child, which was especially critical for low-income families, who disproportionately bore the financial brunt of the pandemic.

For Amanda Renteria, the CEO of Code for America , the new legislation meant the work had only just begun. Renteria understood that making tax credits available is one thing, but helping families navigate bureaucracy to get them is another. Renteria also knew that Code for America , a nonprofit that helps the government use technology to better serve communities, could play a critical role in addressing that access problem.

In an interview with Protocol, Renteria explained how she made the decision to develop a proof of concept for an app that would help people claim the Child Tax Credit. Code for America developed this mock-up despite limited resources and no guarantee that it would be selected by the government as its technology partner. The leap of faith ended up paying off, and Code for America reported in May that it helped 115,000 families successfully claim the tax benefit throughout the 2021 tax filing season.

Renteria’s story, as told to Protocol, has been edited for clarity and brevity.

What we knew about the IRS, having worked with the Volunteer Income Tax Assistance Program [last year], is just how hard it is for people to get through [the system to get benefits]. With VITA, we helped create an online, mobile-friendly, simplified application that actually walks you through [the process] in both English and Spanish.

When Child Tax Credit came up, we knew that this policy — though it passed and everyone was excited about it — would actually hit a wall. We had seen the barriers over time on much easier processes, like food assistance in some states.

We knew filling out forms is extremely difficult. So there wasn't even a question of whether or not we could make this process better. Again, [this was] largely because we had been doing it for a while. We helped VITA sites go online the year before, and we could actually see how you improve throughput if you have a simplified process.

The big question for us was whether we wanted to build it, knowing we weren't sure whether it would land. A lot of the question was how we bring the White House, the Treasury, and the IRS up to speed with our lessons learned. We are a small nonprofit. There were no guarantees that anyone would believe us — trust not only our data, but our ability to do it.

We felt our need to jump in because we didn’t see it happening anywhere else. During that period of time, we were still in transition to a new administration, so we knew that the federal government was still building up its digital talent. The private sector was also focused on higher-end clients and how they were going to live in a new online world. Everything was changing during a pandemic, and the only way we were going to reach [families] was through their phones, and trying to be as simplified and trustworthy as possible.

So in very short order, our engineering team put together a mock-up of what it would look like. [It showed] what the experience would be, how we would triage, how we would ensure that clients are being taken care of through chat. We were able to turn this around in a couple of weeks, but it was based on 10 years of research. We were really confident in our ability to create a front door that spoke to the clients we were trying to serve. We were less confident in whether or not it would be enough to show [the White House] what's possible.

In a meeting with the White House and Treasury, we walked through the actual process someone would go through. Within that, not only did we build trust in terms of the relationship we already had, but the tangibility of seeing the vision of what was possible was a key difference for them. But that took some risk on our part — we had to build it before we knew whether it would actually be a relationship that could lead to the scale we have now proved to have.