We at Interactive Mechanics launched our fellowship program this year, and I’ve been getting questions about how we made it happen and how it connects to hiring.

This nine-month, paid, part-time program, which aims to build capacity for representation and inclusion in tech, provides portfolio development, hard and soft skill building and networking opportunities for entry- and junior-level designers and developers. While we didn’t design it as a direct pipeline for new hires, parts of it were deliberately targeted at teaching us more inclusive hiring practices.

Here are some of our key takeaways.

Do your research

Our fellowship kickoff was this past September, but we first started talking about this idea in the fall of 2015. We talked as a team about how it connected to our company values and goals, and we shared stories of good and bad experiences with similar programs. We researched, we dreamed and we talked to people who’d managed and participated in similar programs.

We then used this information to figure out the intersection between what we wanted to achieve, what we could reasonably take on and what we had to offer. We landed on a fellowship program that’s paid, part-time, and mostly remote, without any client-related work.

For instance, a part-time program allows fellows to continue to handle their existing obligations, whether that’s work, school or family, so that they can afford to participate. It also allows our staff to limit their time mentoring, so they can continue to work on client projects.

Not including client work allows fellows to really focus on learning, without being tied to external deadlines or deliverables. (Editor’s note: Chris Alfano of Northern Liberties web dev firm Jarvus has said the same thing about not giving client work to apprentices.) When we talked to people who’d been through programs that did involve client work, they often felt stressed out and underpaid. I also have some ethical concerns with it — if we’re asking for client-quality work, we should be paying client-quality salaries. I also think it’s crucial to make clients aware of who’s doing the work. (Azavea’s Summer of Maps fellowship does this well by connecting specific client projects to the fellowship program and by offering this work to clients pro bono.)

Start with unconscious bias training

If you’ve ever caught yourself saying, “But we’re already trying hard to find diverse applicants,” you yourself are the biggest barrier to finding diverse applicants.

Here are the basics from our training:

• Diversity is a competitive advantage. This isn’t about being fair or nice; it’s about the fact that we can’t see some of our best candidates.
• Bias is unconscious, ubiquitous and insidious — we are all bad at this and we all need to change.
• We need clear, conscious processes to address our biases.
• This is about progress, not perfection. We’re not going to get this right, certainly not the first time.

If you think this doesn’t apply to you, try out one of the anonymous assessments at Harvard’s Project Implicit. I learned I have a strong preference for abled over disabled and a moderate association between men and careers vs women and home. We’re all biased, and the only way to get around our biases is to talk about it.

Try blind applications and then try again

The first time we tried blind applications, we had a friend of the company review all application materials and redact any information related to name, age, race and gender. She was amazing at it, but it took a lot of time and it didn’t feel right to involve someone outside the team.

So for our fellowship program, we used Screendoor, an online tool for application forms and review rubrics, which I knew from Azavea’s Summer of Maps and used for Al Jazeera’s Media in Context Hackathon that I helped organize with Secondmuse. We built our application as a form, without attachments, so that we could hide name fields from reviewers and make it easier to compare specific information on applications. We’re already planning on asking more targeted questions in our next job application process, instead of asking for a resume or cover letter.

This was much easier than redacting manually, but we still had challenges. Hiding name fields doesn’t do a much good if you ask for links to an online portfolio or Github. Next time, we’re going to consider whether we need those links at all or if they can be part of a later round of review. If we need to assess a designer’s work, could we ask them to include an attachment without names, instead of a link to a personal portfolio site? If we need to review a developer’s Github, could we leave that until after we’ve narrowed down our pool of applicants based on other criteria?

Targeted outreach really, really matters

We ended up with three Black women as our fellows, in part because Black women were a large part of our applicant pool, which happened for a reason. Our outreach focused on the fellowship program as being “designed to build capacity for representation and inclusion in the technology field” and prominently featured benefits for fellows.

We made sure that outreach included groups like women’s education nonprofit Girl Develop It and The ITEM, which aims to organize people of color in the tech and entrepreneurship scene. We contacted people who we know are personally connected to communities that we’re not, even if it meant reaching beyond the traditional tech communities, and we asked them to help us spread the word. We had one-on-one conversations and meetings. We attended events that were new to us.

But it’s clear to me now that we could do a lot better, and that we need to define and measure more. Who counts as “diverse”? I want to track our applicant demographics more accurately, without it feeling cumbersome or unwelcoming to applicants. We also plan to list out specific target demographics internally and include at least three points of outreach for each one. We’re pretty strongly aware of and connected to the women in tech community in Philly, but we have a lot of room to grow to reach other communities, especially in terms of attending events. Sending an email to a listserv can reach new people, but showing up in person gives you the opportunity to have deeper conversations, to hear new perspectives, and to demonstrate your commitment to diversity.

Continue to evaluate

I’m thrilled to say all of that work brought fellows Ebonie Butler, Madilynn Whittle and Candace Worthen on board. You can read about them and their projects here. We still have a way to go before we can assess our program’s overall success, but we’re planning on check ins with them throughout, exit interviews in May and an internal retrospective. I look forward to sharing updates, and you’ll be hearing from each of our fellows on our blog soon.