Kareen Onyeaju spent her time on Reimagining Company Culture talking about a part of DEI strategy that often gets squeezed out of headline metrics. Representation in tech is not just a hiring problem. It is a daily-experience problem, and the two reinforce each other in ways most leaders underestimate. As DEI and People Operations Lead at Ada Support, Kareen has the dual vantage point of someone who builds DEI programs and someone who has lived the experience those programs are trying to change. She is a first-generation Nigerian-Canadian woman whose work centers on supporting diverse teams and pushing DEI conversation forward in technology workspaces.
What stood out across the conversation was her insistence on specificity. Generic DEI frameworks tend to flatten differences that matter. The experience of a Black woman in a senior engineering role is not the same as the experience of a junior product manager who is the only person of color on her team. Kareen wants HR leaders to design for those differences, not around them.
The Representation Numbers in Tech
The data continues to show how far the industry has to go. Pew Research analysis on STEM jobs finds that women hold only about a quarter of computer occupations and roughly 15 percent of architect and engineer roles. The disparities are sharper for women of color, who hold around 4 percent of technical roles in tech companies. McKinsey research on the career ladder notes that in technical roles, only 52 women are promoted to manager for every 100 men, a broken-rung effect that compounds at every level above it.
Kareen pointed out that hiring metrics often look better than they really are because they ignore retention. A company can run a strong outreach campaign, hire a more diverse class, and still end up back where it started two years later if the day-to-day experience pushes those people out. Representation is a stock, not a flow, and stocks only grow when retention exceeds attrition by demographic group.
What Daily Experience Actually Looks Like
How does experience differ for underrepresented engineers compared to their peers?
Kareen described patterns that show up across many tech orgs. Underrepresented employees often spend more time defending the legitimacy of their contributions, navigating coded feedback, or carrying invisible DEI labor on top of their actual jobs. The cumulative cost is real. It shows up in delayed promotions, lower performance ratings, and earlier exits, none of which are captured in glossy diversity reports.
How do you know when DEI programs are doing harm instead of help?
Watch for signs that programs are tokenizing. Asking the same handful of underrepresented employees to speak on every panel. Treating ERGs as free DEI labor. Promoting visible spokespeople into burnout because they are the only people of their identity in the building. Programs that move forward without resourcing the people they rely on can do more damage than no programs at all.
What Actually Works in Tech DEI
Principle 1: Hire deeper into the funnel, not just at the top
Senior hiring is slow and noisy. Junior and mid-level hiring is where companies actually shift representation in five-year horizons. Investing in apprenticeship pipelines, university partnerships, and internal mobility is less glamorous than landing a marquee VP, but it produces durable change. Kareen made the case that pipeline work and culture work need equal investment, not the trade-off many companies make.
Principle 2: Treat inclusion as a leadership behavior, not a training event
Inclusion is built in calibration meetings, in who gets stretch assignments, in how feedback is delivered, and in whose ideas get repeated. None of that is fixed by a single workshop. Kareen wants managers held accountable for inclusive behaviors as a real performance dimension, with feedback from their own teams, not just their bosses.
Principle 3: Listen for what employees actually experience
Surveys are necessary but not sufficient. Underrepresented employees often hesitate to fill out surveys honestly when their identity makes them identifiable. AllVoices' anonymous reporting and DEI hotline give employees a way to share experiences without fearing retaliation, which is the precondition for any honest data about the day-to-day.
Where DEI Strategy Fits in the Tech Stack
The tech industry has built unusually powerful HR systems but does not always use them well for DEI. AllVoices' solutions for tech and DEI connect anonymous reporting, ER case management, and pulse data into a single signal layer leaders can act on. The point is not more dashboards. It is fewer surprises.
How DEI signals interact with ER and engagement
The most important signals are the ones that span systems. A pattern of implicit bias in performance feedback shows up in promotion data, ER cases, and survey comments at the same time. A culture of microaggressions on a team shows up in attrition rates, exit interviews, and engagement scores. Looking at any one of those in isolation misses the picture. Treating them as one connected dataset is how leaders learn what is actually happening.
Frequently Asked Questions About Representation in Tech
What is the single biggest lever for representation in tech?
Manager behavior. Hiring matters, but managers determine whether new hires actually get the assignments, feedback, and sponsorship that translate into careers. Companies that invest in manager development on inclusive practices change their representation curves more than companies that invest only in recruiting.
How do you keep ERG leaders from burning out?
Pay them. Treat ERG leadership as part of compensated work, not a volunteer hobby. Resource the groups properly. Give ERG leaders development paths. Without that, the people doing the most DEI labor are also the people most likely to leave.
How should representation goals be set?
Tied to the talent market and to the lifecycle, not pulled from a hat. Use realistic benchmarks at every level, not just at the top, and track promotion equity alongside hiring equity so the funnel does not leak.
What about backlash to DEI programs?
Backlash usually intensifies when DEI is framed as a zero-sum reallocation rather than as a system upgrade that benefits everyone. Programs that focus on better hiring rubrics, fairer feedback, and clearer promotion criteria tend to face less resistance because they read as fairness improvements that apply to all employees.
Does representation matter if the work is good?
Yes. Representation affects who gets heard, whose ideas get prioritized, and whose customer needs get understood. Tech products built without diverse perspectives miss real markets and real failure modes. Representation is a product quality issue, not just a moral one.
The Bottom Line for HR Leaders
Kareen's argument is that representation in tech is not a single metric to be moved. It is the product of dozens of decisions that companies make every day. Who gets hired into which roles, who gets promoted, who is asked to lead a project, and who is taken seriously in a meeting. Each of those is a small lever, and HR has more influence over them than is sometimes assumed. The work is in pulling those levers consistently and measuring whether they are actually moving the system.
Representation is also a credibility issue. Employees, candidates, and customers can tell the difference between a company whose numbers reflect its values and one whose numbers reflect its slogans. Closing that gap is what real DEI strategy is for, and the companies willing to do the slow, structural work end up with workforces that look more like the markets they serve and stay longer once they arrive.
See how AllVoices helps tech HR teams turn representation strategy into measurable change.
.png)
.avif)
.avif)
.avif)