Our referral bonus pilot program.
To drive diverse hiring in top-of-funnel roles, we rolled out an underrepresented referral bonus pilot. We stopped the program after it didn't drive meaningful change, and we're still working to improve our diverse hiring practices. Published April 2019.
Piloting a referral bonus
When we launched this blog, we said we’d be candid about our inclusion and diversity work, showing both what’s worked and what hasn’t. We’re hoping to prompt other companies to share their successes and failures, so we can learn from each other and move the industry forward.
Late last year, we launched an underrepresented referral bonus pilot. Our goal was to increase the number of women and underrepresented minorities* (“URMs”) in the top of our funnel for tech roles, and senior business roles. We’ve seen good progress on the hiring front by investing in conferences focused on underrepresented talent and providing dedicated diversity sourcing support for teams, and we’re committed to trying new tactics in pursuit of progress.
In October 2018, we announced the referral bonus pilot internally.
We offered a $4,000 referral bonus to any Square who submitted a qualifying referral in the fourth quarter of 2018, where said referral was subsequently hired and stayed at Square for at least six months.
Qualifying referrals included women and URMs in technical roles, and women and URMs in senior business roles. While we feel strongly that gender, race, and ethnicity don’t capture diversity in full, we limited the pilot to EEO-1 groups because they are relevant, measurable indicators of how effective we are at recruiting and supporting underrepresented talent. We also chose to focus on specific seniority levels after evaluating the diversity of our current employee population and agreeing that this is where we needed to make a greater effort.
One major challenge was tracking the referrals submitted as part of the pilot so we could accurately measure our results. Structurally, our Applicant Tracking System didn’t support two central needs of the pilot: notifying the Talent team when referrals were submitted, and tracking this subset of referrals separately from the overall referral population. This was critical because if you’re asking for more referrals, you need a scalable way to manage the potential influx while maintaining a positive experience for both the referrals and referrers.
Accordingly, we built a new, lightweight, entirely separate referral submission process which allowed us to track our pilot referrals specifically. Multiple teams, including Talent, Inclusion & Diversity, Internal Communications, Compensation, and Legal all coordinated on the design and rollout of this pilot. An added benefit was that this allowed us to unlock new levels of data, overcoming our previous limitations in understanding and tracking URM interview performance as compared to our entire candidate population.
When trying something new, especially with inclusion and diversity work, robust communication is essential. We knew there was potential for misinterpretation of the goals here, and we wanted to be sure our messaging was clear, consistent, and repeated. So we intentionally erred on the side of overcommunication:
- First, we announced the pilot to the Talent team, equipped with a robust FAQ in anticipation of recruiters’ questions about how this would impact their roles. This also ensured Talent was clear on the intention of the pilot and consistent in any conversations about the pilot with the broader company.
- We then announced the pilot to the entire company at our all-hands. Our Head of People delivered the presentation, explaining this was a pilot, and reinforcing the importance of inventing and learning, one of our operating principles.
- We immediately followed up with a company-wide FAQ so folks could dig into the details.
- We committed to following up with the company on the outcome and learnings at the pilot’s end.
Square is proud to champion a culture of transparency and two-way communication, so we were anticipating a lot of discussion and potentially pushback around this initiative. Surprisingly, we received hardly any questions—quite unusual for Square.
Our takeaway was that what might feel like overcommunication can help head off perceived opacity. Executive buy-in is also key. Having a member of the leadership team champion the effort and deliver the message to the entire company helped show this was something Square takes seriously, and regular updates on the process drove this home.
Over the course of the pilot and the quarter that followed, we did not see a notable increase in the number of hired underrepresented (UR) referrals. Additionally, we found that quarter over quarter, the pilot didn’t increase the number of referrals submitted overall.
There was good news: We reviewed our data across both pilot hires and non-pilot hires, and found that we hired UR talent at the same rate as non-UR talent. Of the referrals that came through the pilot, 20% were invited onsite for an interview (a 7% higher conversion than our non-pilot referrals), and of those who came on site, 45% received an offer, 100% of whom accepted.
The results suggest that, despite our efforts at innovating, the pilot was not an effective means of increasing the number of URM candidates in the top of our funnel. That a monetary bonus didn’t drive a measurable increase in the number of underrepresented referrals or hires suggests that it’s not that Squares lack motivation to refer great people (which of course was great news), but that our employee networks are the limiting factor. One might think this is a successful outcome — everyone’s leveraging their networks the best they can! — and while that’s true, it means we need to keep pushing for tactics that will successfully add to our pipeline and effect positive change.
“The results suggest that, despite our efforts at innovating, the pilot was not an effective means of increasing the number of URM candidates in the top of our funnel.”
One unanticipated benefit of the pilot was that in forcing our hand at a more detailed referral tracking process, we were able to track underrepresented referrals (as compared to our broader pool of candidates) throughout the entire hiring process. Prior to the pilot, we had a limited view into how different talent populations were performing in each stage. With a better understanding of the pass-through rates for each group in the pilot, we had an additional opportunity to assess the health of our overall hiring process and our ongoing efforts to mitigate for bias.
What we found was that for every phase of the process, the pass-through rates were virtually the same for the general pool and for the underrepresented referrals coming through the pilot. This also was true for pilot referrals as compared to non-pilot referrals. This promising finding seemed to demonstrate that our continuing initiatives to strengthen and remove bias from our hiring process are working.
Ultimately, because the program didn’t drive change, we discontinued it. We’re still focusing on other opportunities to increase top-of-funnel diversity, including investing in partnerships with professional communities for underrepresented talent and focusing outbound sourcing efforts in a more targeted way. We’re going to take the learnings from this pilot and apply them specifically in markets where we need to move the needle. We’ll continue to report what we see.
For this pilot, we defined underrepresented minorities as candidates who identify with one or more groups that are underrepresented at Square, based on data we collect for required EEO-1 reporting purposes: Black, Hispanic or Latino, American Indian or Alaska Native, and Native Hawaiian or other Pacific Islander.