In part three of The Antiracist Leader, I covered some of the ways workplace policies can unintentionally and disproportionately exclude people, even when we intend to make them as objective as possible. When creating policies, such as determining who can work remotely, who can switch teams, or who can have access to professional development opportunities, I laid out why it’s so problematic to rely on seniority as the main criteria for who qualifies. It’s far better to grant privileges and opportunities to a diverse set of folks who have earned and will benefit from those investments. But if your aim is to target employees who have been deemed as ‘high-potential’ or ‘high-performing’, you might guess that this is also problematic because these are most likely to be completely subjective determinations rife with bias, and they are often a major contributor to why our senior ranks are so ethnically homogenous in the first place!

If you’ll recall part one of this series, I talked about how Black women report needing to prove themselves over and over again, and how Black employees are penalized more harshly for mistakes. The Kapor Center’s 2017 Tech Leavers Study found that underrepresented women of color were significantly more likely to cite being passed over for promotion as a major factor in their decision to leave tech. It’s critical that we confront these statistics head-on and examine how these same biases might creep into our own organizational processes.

The opaque process

In this article, I’m speaking directly to leaders who have the power to create and change processes – especially processes that pertain to an individual’s professional development, performance, and advancement. One thing we often forget as we rise through the leadership ranks is what it was like to have no clue what happens behind closed doors – to not understand how promotion decisions get made, or why someone got tapped to launch a new product line, or how it was determined that a new hire was given a particular title. At some point in our careers, these things were likely a mystery to us. And importantly, if we’re part of a group that enjoys majority representation in a company and it’s easy for us to build networks with senior leaders, maybe these things weren’t so opaque because we had mentors helping us navigate them. Or perhaps it simply didn’t matter that they were opaque because they always worked for us, and we never learned to question their intent or outcome.

As managers, we participate in these types of hiring and performance processes frequently, and we may have convinced ourselves that they are mostly fair and sufficient. We build and iterate on these processes and agonize over their details. We build safeguards to promote objectivity. We convince ourselves that we’ve considered all the angles and built something that maybe isn’t perfect, but is at least leaps and bounds better than the process at our last company. And then we forget that the bar should be higher than that, and that those who are beholden to the outcomes of a process deserve to at least know how it works so they can navigate it and hopefully convince themselves of its fairness.

An approach to process iteration

At a previous job, we stumbled on an approach to process iteration that was completely eye-opening for me. It began with a working group. Dissatisfied with the inertia on various other diversity initiatives, a small group of engineering managers and directors teamed up with an HR partner and recruited a few underrepresented engineers to form this new diversity working group. The rules were simple:

1. Pick one thing at a time to address and focus only on that.
2. The only action items we could give to the engineers were to give us feedback on our problems and proposed solutions. They were encouraged to solicit this feedback from members of our Employee Resource Groups (ERGs), but the rest of the implementation and execution was otherwise on HR and engineering leadership, as it should be.

The first thing the engineers asked us to focus on was the promotion process. There was no specific request, but with so much D&I focus on hiring, they wanted us to focus on retention and advancement. It was immediately clear that many of our engineers didn’t fully understand how promotion decisions were made, and as a result, they were concerned that it might be unfair.

Now, we had built all sorts of bias checks into our promotion process: every cycle we offered office hours, where folks from the ERG could come to get their self-reflections and promotion packets reviewed by an experienced manager. We ensured our calibration meetings had the most diverse representation possible. We reviewed a common list of biases that were handed out in every calibration meeting and we made it someone’s job to call these out when they heard them. We collected and analyzed data after every review cycle. But how often did we actually talk about these things?

Naively, our working group thought there was a simple solution. We’d simply explain how the process works in two info sessions for underrepresented engineers (anyone could join, but we sent targeted invites to the relevant ERG). I think we expected that doing so would put people at ease.

Instead, I ran into one of our directors on his way out of leading the first info session, and he was visibly shaken. The session was useful overall, but the questions from a nontrivial number of engineers revealed a concern we hadn’t yet heard: many people felt that their managers weren’t advocating for them and that they had no recourse. This problem took several different shapes:

1. Their manager was too new and had asked their engineer to wait until the next cycle when they would be better calibrated (we were growing quickly at this point so this happened more often than you’d think).
2. Their manager disagreed that they were ready for promotion, and was unwilling to submit them for consideration (there was no concept of ‘self-nominating’ for promo at this point).
3. Their manager was willing to submit them for promo consideration but expressed feeling uncertain about their chances, which led the engineer to believe they wouldn’t be properly advocated for in the calibration meetings where promo decisions were made.

Within an hour, we jumped in a room with HR to discuss how we might urgently address these concerns. Educating new managers on the process and allowing people to self-nominate would only partially solve this issue (we did some of this, but we considered it insufficient as a solution). The common thread between these three concerns was all about the manager involved. And we couldn’t deny that some folks just got a rough shake – some people might have lost and gained a manager in the weeks leading up to a performance review, others may have found themselves with a manager who just wasn’t skilled at giving them clear feedback. How could we address this holistically?

Within 30 minutes, we’d devised the Promo Sponsor program, and within a week we had rolled it out. The idea was that if you wanted to be considered for promotion, and for whatever reason you didn’t feel like you had the full support of your manager, you could request a ‘sponsor’. Sponsors were experienced managers across the org who were known for giving good feedback, and who were deeply familiar with the company promotion process. Engineering directors and HR business partners could easily identify a small set of managers who fit these criteria.

Should an engineer request a sponsor ahead of a promotion cycle, HR and the working group lead would confidentially assign them a sponsor who was outside of their reporting chain. The sponsor, having undergone a short training session on the program, would then reach out to the engineer to learn more. If the sponsor uncovered a seriously toxic conflict between the engineer and their manager, they were instructed to inform the engineer they would need to report that back to HR. But that was very uncommon – in virtually all cases we found the situation to be much more nuanced: a manager who failed to recognize that their engineer was in dire need of positive feedback, a manager who was supportive of a promotion but lacked confidence in their pitch and felt fearful of getting their engineer’s hopes up, or an engineer who just refused to accept the feedback that had been delivered to them. These situations were all addressable through the sponsor program.

Once an engineer met with their sponsor, they could decide to proceed to the next step, which involved their sponsor and manager speaking 1:1 about the situation. The program was not designed to villainize or circumvent one’s manager – quite the opposite, in fact. One of the largest benefits of this program was that it provided an opportunity for managers to connect with these highly respected sponsors, who could then work in partnership with them to provide the best possible support for the engineer seeking sponsorship. The sponsor might offer a different perspective on why feedback had not been landing properly for the engineer. They might help a manager calibrate themselves to the company levels and processes. They might actually reinforce the feedback that a manager had been giving. They might also attend that team’s calibration session to help provide additional color or support. There were many possible ways to approach a given problem, and it was the job of the sponsor to work with the manager to figure out what that was.

This program was widely praised by the set of folks who’d originally expressed concern with our promotion processes. Even people who did not participate in the program said they felt comforted by its existence; it would be there if they needed it.

What we learned

The point of this example is not to convince you that you need a Promo Sponsor program. It’s not even a claim that the program was successful, although that depends on how you measure success. After two iterations of the program, 15 out of 16 participating employees were either not put up for promotion, or they had their promotion denied. However, in a follow-up survey we ran, ⅔ of participants surveyed said that having a sponsor helped them feel better about the decision. In multiple instances, engineers who requested a sponsor or simply considered requesting one wound up pulling out of the program because it led to a clearer conversation with their manager, and they no longer felt they needed outside sponsorship. Having this program also gave us additional insight into how many folks were feeling like they lacked advocacy from their manager and what those demographic breakdowns were. (Interestingly, once this program was rolled out to the wider organization, more than half of the engineers requesting sponsors were from majority groups, proving once again that improving processes for your marginalized employees improves those processes for everyone). But really, the anecdotal praise was enough to invest in further iterations.

The point of this example is instead about the actual approach to building and fixing a process. We learned so many important lessons through this working group. We learned that it didn’t matter that we thought our process was fair, nor did it matter that the data from this program seemed to support that assertion (had we instead found that this program resulted in a bunch of overturned promotion decisions, that would have been a very sad win indeed). What mattered was that folks initially perceived the process to be unfair. Why stick around if you think the deck is stacked against you? This was a way to show we were listening, and to make an honest effort to root out their concerns.

We also learned how important it was to give managers ways to connect with and learn from each other. Sometimes we get so busy at work and so mired in our team’s fires that we forget to invest in building relationships with our peers, or we don’t want to admit to our peers that we are struggling with a tricky issue on our team. Of course, we have HR to help us with these situations, but although I’ve worked with some incredible HR partners over the years, no one can really check me quite like an experienced peer. We have to be able to set our egos aside and request help from each other. I’ve incorporated this concept into so many aspects of my work since then by establishing a cross-functional manager community at my current job and finding different opportunities for managers to review each other’s feedback or roleplay tough conversations.

You can adapt this process in your own organizations. Here are just a few ideas:

• Create a forum for managers in disparate parts of the company to meet regularly and discuss the issues they are facing. Ensure that each forum represents a broad range of experience. 
• During performance review time, make it easy for managers to request a peer review of a performance assessment that was particularly difficult to write or deliver.
• One of my personal favorites: insist that any manager delivering feedback to an underrepresented employee about their demeanor, communication, or confidence seek out a review from a set of trusted managers. Trust me, managers will want this help if you offer it.
• Invite a senior engineer to your calibration and ask them to explain the process from their own perspective to your team at an All Hands. Invite people to give feedback directly to you or via your senior engineer.
• Explain to your underrepresented employees how they can go about switching teams, or how they would find out if there’s a new leadership opportunity elsewhere in the organization. Listen when they undoubtedly tell you it’s not as simple as you make it sound. Develop a solution with them in mind and seek their feedback throughout the process. Measure the perception of the process in addition to the metrics you define as success criteria.

There are so many directions you can take this in. If you don’t know where to start, start with your Black employees. Ask them what about the org they find unfair, problematic, or just plain mysterious. Have a trusted leader explain how the process works in a forum where it’s safe for people to share candid feedback (meaning, don’t do this in a room where they will be in the minority, or where their own management chain is present). Work with other organizational leaders to propose solutions, and get buy-in from the group. Iterate, iterate, iterate.

Finally, if you want to see the conference talk version of this article, check out my talk, ‘Re-engineering Inclusion’ from GOTO Chicago in 2019.

The end(less process)

Antiracism, just like allyship, is an ongoing process. It’s fighting the paralysis we experience when faced with intractable problems that we might not feel responsible for creating. It involves serious introspection of our own motives and our complicity in systems that, regardless of the intent behind them, uphold and reinforce inequality. By no means do I have the answers, and you’re much better off listening to the voices of experts – namely folks of color – as you embark on this work. However, my hope is that I have been able to reach leaders who, like me, might be low-key terrified to find these issues lurking in their own organizations, and who may struggle to understand the actions we can take to make positive change.

Thank you for coming on this journey with me. I hope we’ll continue it together.