Latest / The Diversity Career Podcast with Fexingo: Inclusion, Equity, and Underrepresented Professionals / How One Company Retained Underrepresented Engineers With Cohort Mentorship
Transcript
- Lucas: Luna, I want to talk about a retention experiment that ran at a mid-sized fintech company called BridgePay — roughly three thousand employees, mostly engineers. They noticed something frustrating: they were hiring Black and Latino software engineers at a decent clip, but losing about a third of them within eighteen months. Luna: That's a brutal leak in the pipeline. And it's not uncommon — a lot of companies hit that wall after they get past the initial recruiting push. Lucas: Exactly. So they tried something that wasn't a typical mentor-match program. Instead of assigning each junior engineer one senior mentor, they created small cohorts — five to seven engineers per group, all from underrepresented backgrounds, paired with two senior leaders who shared a similar identity. The cohorts met monthly for a full year. Luna: What was the thinking behind the group format versus one-on-one? Lucas: Their data showed that one-on-one mentoring often fizzled after three or four months — the mentor got busy, the mentee felt like they were bothering someone. But a cohort creates peer accountability. If you skip a session, your peers notice. And the senior leaders are more motivated to show up when there's a group waiting. Luna: So it's structured social pressure in a good way. Did they measure the impact? Lucas: They did. Over two years, turnover among Black and Latino engineers who participated in the cohort dropped by 40 percent compared to a control group who received standard one-on-one mentoring. Promotion rates also improved — cohort members were promoted at roughly twice the rate of non-participants within the same timeframe. Luna: Those numbers are hard to ignore. What were the cohort sessions actually like? Were they just career advice sessions? Lucas: They started with a structured curriculum the first six months — things like navigating performance reviews, building a technical brand, dealing with microaggressions in code reviews. The second half was more open-ended, driven by whatever the group needed. But the key feature was what they called 'sponsorship triggers.' Luna: I like that phrase. What does it mean in practice? Lucas: At the end of each month, the senior leaders had to identify one concrete action they would take for a cohort member — nominating them for a high-visibility project, introducing them to a director, or recommending them for a stretch assignment. It turned passive mentorship into active sponsorship. Luna: So the cohort became a forcing mechanism for sponsorship, not just advice-giving. That's smart because research shows sponsorship matters more than mentorship for career advancement, especially for underrepresented groups. Lucas: Right. And they tracked those sponsorship actions. Over the year, each senior leader made an average of about eight sponsorship moves per cohort member. That's a lot more than you'd get from a casual mentoring relationship. Luna: What about the senior leaders themselves — did they feel burned out? That's a common concern with any program that asks more from already busy people. Lucas: BridgePay surveyed them quarterly. The majority said the cohort format actually felt less draining than one-on-one mentoring because they could share the load with the other senior leader in the room. And they reported higher satisfaction because they could see the group dynamic working — it felt more like teaching a class than coaching an individual. Luna: That's interesting — the group format might be more sustainable at scale. Did BridgePay expand it beyond the pilot? Lucas: They did. After the first year, they rolled it out to all engineering teams globally, and they're now adapting it for product management and data science. They also opened it up to allies — engineers from majority backgrounds who wanted to be part of the cohort as peer supporters, not leaders. Luna: I want to pick up on that ally piece. How did that work? Lucas: Allies could join as what they called 'cohort contributors' — they attended sessions, shared their own experiences, and helped with things like resume reviews or practice interviews. But they didn't have the sponsorship trigger responsibility. That was reserved for senior leaders from underrepresented backgrounds, because the company wanted those sponsorship actions to come from people who understood the specific barriers. Luna: Makes sense. So the ally role was more about amplifying and supporting, not leading. I can see that being a good entry point for people who want to help but aren't sure how. Lucas: Exactly. And BridgePay tracked engagement among allies too — they found that allies who participated were more likely to advocate for inclusive policies in their own teams later. It created a ripple effect. Luna: So besides the retention and promotion numbers, what were some of the softer outcomes that the company noticed? Lucas: They administered an inclusion index survey before and after the program. Among cohort participants, scores on 'I feel my voice is heard in meetings' jumped by 28 percent. And scores on 'I have a clear path to advancement' rose by 35 percent. The control group didn't see any significant change. Luna: Those are the kinds of metrics that actually predict whether someone stays long-term. If you don't feel heard and you don't see a path, you're already halfway out the door. Lucas: Right. And one thing that struck me — they also measured something called 'network density.' They asked participants to name colleagues they could turn to for career advice. Before the program, the average was about two people. Afterward, it was nearly seven. That's a huge shift in social capital. Luna: And that network effect is exactly what underrepresented professionals often miss out on in homogeneous workplaces. The cohort essentially fast-tracked relationship-building. Lucas: Yeah. And the company didn't spend a ton of money on it. The main cost was the senior leaders' time — about four hours per month per cohort, plus some training for them on how to be effective sponsors. No expensive software, no outside consultants. Luna: So it's a relatively low-cost, high-impact intervention. I wonder why more companies don't try this model. Lucas: I think part of it is that traditional mentoring is so ingrained — companies assume one-on-one is the gold standard. They don't realize that group dynamics can actually amplify the benefits. And there's also a fear that cohorts might create silos, but BridgePay's data showed the opposite: participants reported feeling more connected to the broader company, not less. Luna: You know, it reminds me of something we've talked about before on this show — that inclusion isn't just about bringing people in, it's about making sure they have the relationships and sponsorship to thrive. This cohort model seems to operationalize that. Lucas: That's exactly right. And speaking of making sure people have what they need to thrive — if you're listening and this kind of practical inclusion work is something you care about, we want to keep having these conversations. This show is ad-free and listener-supported, and if you find value in episodes like this one, you can support us at buy me a coffee dot com slash fexingo. It helps us keep digging into real-world case studies like BridgePay's. Luna: Yeah, totally. Every bit of support goes straight into research and production time. And we're grateful for the listeners who make that possible. Lucas: So back to BridgePay — one last thing I want to mention. They published their results internally, and it actually changed how the company thought about retention. HR started using cohort participation as a leading indicator for engagement. If someone dropped out mid-program, it triggered a check-in. Luna: That's a smart operational move. Treating program dropout as a signal rather than just an administrative note. Lucas: Exactly. And I think that's the takeaway here — it's not just about the cohort model itself. It's about building feedback loops so you can constantly improve. BridgePay didn't just set it and forget it. They iterated based on data. Luna: And the results speak for themselves: 40 percent less turnover, double the promotion rate, and a measurable increase in inclusion. That's the kind of case study every talent leader should read. Lucas: Absolutely. So for anyone thinking about retention — especially for underrepresented engineers — consider whether a cohort model could work in your organization. It might be the simplest thing you haven't tried yet. Luna: Great episode, Lucas. See you next time. Lucas: Thanks, Luna. See you then.