My business is about investing in extraordinary founder talent so it pains me that I can’t write a Homebrew Term Sheet to Laura Weidman Powers. But alas, she’s changing the world through another awesome channel right now, CODE2040. I’ve gotten to know CODE2040 as a Mentor, Panelist and Donor. And here I get to know her a little better via Five Questions:

Hunter Walk: We met during CODE2040’s first class of students. I was at YouTube and fortunate enough to be a CODE2040 Mentor that summer. Can you share how the organization got started and what the year “2040” refers to?

Laura Weidman Powers: Tristan Walker and I started CODE2040 after working in tech and recognizing 1) how much opportunity there is in this space but 2) how little diversity there is – this was particularly true when we started the organization in early 2012. We chose the name CODE2040 as a call to action: the year 2040 is the year that people of color will be the majority in the United States, and we want to see a tech industry that reflects the country by that time.

No one was really talking about diversity in tech in 2012 and there wasn’t much happening by way of solving the problem so we decided to just jump in and pilot and learn. We launched the CODE2040 Fellows Program, which places top black and Latino/a software engineering students in internships, with five students in summer 2012. This year, in 2016, we will have about 60 Fellows – and we run three other programs for students, entrepreneurs, and tech companies, as well, reaching about 1,000 people a year across all programs.

HW: You’ve been very deliberate in expanding CODE2040 program size over time – not wanting to move too quickly in the early years. What foundation did you seek to put in place and how do you decide if you’re ready to grow?

LWP: We were really deliberate in our growth. For the first three years, CODE2040 was synonymous with the CODE2040 Fellows Program. It is our flagship. We had five students, then 18, then 25. Throughout that time we weren’t just running and growing the program, we were learning. We knew people wanted us to do more – students, companies – but we wanted to make sure that we weren’t just throwing darts, but rather that we were leveraging existing knowledge to run smart experiments.

In 2014 we felt like we’d learned enough about what additional supports college students needed to prep for tech’s idiosyncratic interview processes that we could start piloting some new programming. That lead to TAP (Technical Applicant Prep Program), the pilot of which was funded by Google.org. We partnered with Google for Entrepreneurs in 2015 to begin to bring our work national, with a focus on entrepreneurial ecosystems.

This has always been core to what we care about – increasing representation for blacks and Latino/as in tech – and in the Valley it took the form of the Fellows Program. TAP extended our student-focused work. As we go into other markets, we can choose whether it makes sense to be more student-oriented or ecosystem-oriented with the CODE2040 Residency.

And now, in 2016, we’re growing our programming for companies. We found that the companies we were working with (about 70 partners to date, 42 this year) want talent, yes, but they also want process improvements, culture shifts, coaching – so we’re building out programming that is just for company partners to empower them to create their own change internally.

HW: How do you structure CODE2040’s funding base? Is there a revenue generating aspect to the organization that makes it self-sustainable?

LWP: One of my main goals for CODE2040 is for it to be self-sustaining from a revenue perspective. We are addressing a market failure with CODE2040, and so philanthropic capital is completely necessary at this moment in time. However, earned revenue is great too because it gives us security as well as freedom to experiment, learn, fail, and ultimately succeed.

The Fellows Program is a great example of an earned revenue strategy that’s gone well. It took four years, but the revenue we get from company partners participating in the Fellows Program just about covers program costs – we expect to be able to cover Fellows Program costs, allocated overhead, and begin to support the development of new resources for company partners starting this year.

The CODE2040 Residency is an example of another funding model that works well. It’s a program partnership with Google for Entrepreneurs. We work closely with the GFE and Google Ventures teams on design, development, and implementation, and they underwrite the program in exchange for us implementing it with their partners.

Our other programs like TAP are newer, so they still have a ways to go until they are self-sufficient – they require philanthropic capital. We think of that money as an investment – like seed funding or a Series A. It’s necessary to get a program on its feet and to provide a runway to develop an earned revenue model down the line – sometimes that takes time.

In addition to launching and sustaining newer programs, philanthropic capital is vital to CODE2040’s growth as an organization. Even when a program like the Fellows Program can cover its own costs, we still need to be able to invest in the growth and sustainability of the organization as a whole. We’re hiring a lot right now – we brought on a COO in the fall, we’re hiring for more program staff and a director of talent now. These are investments that may require philanthropic capital to start, but that we can then factor into our revenue models.

HW: When you track the process of CODE2040’s Fellows post-program and post-graduation, what trends are you starting to see? How do you think about efficacy and ROI?

LWP: We’re so excited by what we’re seeing with our alums. This past summer, 90% of our Fellows received return offers from their summer employers – a really strong vote of confidence in our partner companies’ believe that these are talented developers that belong in the Valley full time. For the first time this past summer we had a CODE2040 alum working at a partner company who was managing the partner’s CODE2040 summer interns!

We also see some of our Fellows starting their own companies. One of our first five runs a company called Wise Systems. They just went through Techstars in Boston. We have another group of Fellows in Y Combinator’s fellowship program with Shypmate. It’s been so thrilling to see what our Fellows have done and continue to do!

HW: Are you tired of hearing people talk about diversity? Do we talk too much with not enough action and accountability, or is the increased visibility of the problem an important foundational step?

LWP: I definitely don’t think we talk about diversity too much. It’s such a hard topic for people to get comfortable with that I think the more it’s on the tips of our lips the better. That said, talk does not replace action, of course. And if talk is all we’re doing, we’ll never see progress.

I see all the talk as opening the door. When tech companies started releasing their diversity data in Spring 2014 that not only changed the conversation, it changed the actions companies took. Transparency breeds accountability in this case, I think.

Now companies need to be willing to experience and change. Some really are – I’ve started to hear more and more companies talk about shifting policies, practices, culture – that’s really exciting. So I’m optimistic. Change is hard and change is slow. But I think we’re on the right track.