Latest / The Diversity Career Podcast with Fexingo: Inclusion, Equity, and Underrepresented Professionals / How One Company Built a Truly Inclusive Recruiting Pipeline
Transcript
- Lucas: So there's this company called RelayTech — mid-sized, about 800 employees, builds customer service software. Not a household name, but they did something in 2024 that a lot of bigger companies have only talked about. Luna: What'd they do? Lucas: They redesigned their entire recruiting pipeline — from sourcing to offer — with the explicit goal of increasing hires from underrepresented groups. And they didn't just change a job description and call it a day. They went deep. Luna: Okay, I'm curious. What was the actual change that had the biggest impact? Lucas: The thing that really moved the needle was their sourcing overhaul. They stopped relying solely on the usual channels — LinkedIn, referrals from existing employees — because those channels tend to replicate the existing demographic makeup of the company. Instead, they partnered with five community organizations that focus on training and placing underrepresented tech talent. Think groups like Code2040, Year Up, and others focused on Black and Latinx engineers, women returning to the workforce, and first-generation college grads. Luna: Right, so they diversified the top of the funnel. But that only works if the rest of the process doesn't filter those candidates out. Lucas: Exactly. And that's where most companies stumble. RelayTech realized that their job descriptions were loaded with unnecessary requirements. For example, they had a senior software engineer posting that required a computer science degree and five years of experience with a specific framework. They looked at the actual top performers in that role and found that nearly half didn't have a CS degree. So they removed degree requirements from about 60 percent of their roles and instead focused on demonstrated skills and a work-sample test. Luna: That's a big shift. I think a lot of companies are afraid to drop degree requirements because they worry about lowering the bar. But the data actually shows the opposite — you get a broader, often stronger, pool. Lucas: Yeah, and they coupled that with structured interviews. Every interviewer had to go through a two-hour cognitive bias training before they could be on a panel. And the interviews themselves were standardized — same questions, same scoring rubric, no off-the-cuff 'tell me about yourself' stuff that can lead to affinity bias. Luna: The Harvard Business Review had a study a couple years ago — I think it was by Francesca Gino and colleagues — that found structured interviews reduce bias by about 28 percent compared to unstructured ones. But a lot of companies still resist because it feels less 'organic'. Lucas: Right, but organic is just code for 'gut feel,' and gut feel is where bias lives. RelayTech also instituted a four-person diverse slate policy for every final-round candidate. Meaning, for every role, the final pool of candidates had to include at least two people from underrepresented groups. That forced the hiring managers to keep looking if the initial pool was too homogeneous. Luna: That policy gets a lot of pushback — some people call it a quota. How did they handle that internally? Lucas: They framed it not as a quota, but as a search quality standard. The CEO sent a note saying, 'We are not lowering standards. We are expanding our search until we find qualified candidates we otherwise would have missed.' And they backed it up with data showing that homogeneous slates actually lead to worse hiring decisions over time. Luna: So what happened over the 18 months? Did they actually get results? Lucas: They did. Underrepresented hires increased by 40 percent — from about 22 percent of new hires to 31 percent. But here's the part I found really interesting: the first-year attrition rate for those hires dropped from 22 percent to 9 percent. Because they didn't just change the hiring — they also changed the onboarding and support structure. Luna: That retention piece is crucial. A lot of companies focus on getting people in the door but then leave them to sink or swim. Lucas: Yeah, RelayTech launched a formal sponsorship program — not mentorship, sponsorship. Each new hire from an underrepresented background was paired with a senior leader whose job was to actively advocate for them in promotion conversations, put them on high-visibility projects, and introduce them to key stakeholders. That's the kind of thing that makes people stay. Luna: Sponsorship is so much more powerful than mentorship because it's action-oriented. A mentor gives advice; a sponsor uses their political capital to open doors. Lucas: Exactly. And the senior leaders were held accountable for retention. Their bonuses had a diversity retention metric attached. So it wasn't just a feel-good program — it was tied to compensation. Luna: That's the kind of structural change that actually lasts. If you don't make people accountable, it fades after the initial enthusiasm. Lucas: If today's episode was useful to you and you want to keep it ad-free, buy me a coffee dot com slash fexingo helps. Luna: Yeah, listener support is what lets us spend time digging into these case studies. So if it's valuable, that's the place. Lucas: So back to RelayTech — one more thing they did that I think is replicable even for small teams. They created a 'recruiting scorecard' that tracked not just fill time and cost per hire, but also diversity at each stage of the funnel. They could see, for example, that their screen to interview conversion rate for Black candidates was 30 percent lower than for white candidates. That prompted them to look at the screening questions and find that the phone screen was asking about 'culture fit' in a way that penalized candidates from different backgrounds. Luna: Culture fit is often a euphemism for 'looks like us.' So they replaced it with 'culture add' — what does this candidate bring that we don't already have? Lucas: Exactly. They changed the question to 'What perspective do you bring that would challenge our thinking?' That small language shift made a big difference in who advanced. Luna: It's amazing how tweaking the words in an interview can change outcomes. But it also requires training the interviewers to actually listen for those answers instead of defaulting to their usual patterns. Lucas: Right. So they did quarterly calibration sessions where interviewers would review anonymized interview scores and discuss discrepancies. Over time, that improved inter-rater reliability — the consistency of how different interviewers scored the same candidate. Luna: That's the kind of operational discipline that most companies lack. They'd rather just complain that they can't find diverse talent than examine their own process. Lucas: And that's the takeaway. RelayTech showed that the talent is out there — you just have to build a pipeline that doesn't filter them out before they get a fair shot. The 40 percent increase in underrepresented hires and the drop in attrition prove that the system, not the candidates, was the barrier. Luna: I'd love to see more companies adopt the scorecard approach. It makes the problem visible, and you can't fix what you don't measure. Lucas: Yeah. And on that note, I think the next step for RelayTech is looking at promotion rates — are these hires advancing at the same rate as their peers? Because hiring is only the first chapter. Luna: Right. Inclusion is about advancement, not just entry. Hopefully they're tracking that too. Lucas: I'd bet they are. This team seems to think in metrics.