Dear Tech Companies: What are your “diversity stats” really saying?

Each story in our Dear Tech Companies series focuses on issues in the tech space and provides strategies and solutions to companies looking to invest in meaning solutions that will drive impactful industry change and make the industry more accessible to Black, Latina, and Native women.

After much public pressure, Alphabet, Apple, Facebook, Microsoft and Twitter released their first diversity reports in 2014. In the past ~7 years more companies are releasing diversity reports, but the energy supporting diversity in tech is dwarfed by the homogenous reality of the sector. Shouldn’t collecting and releasing these numbers inspire more urgent solutions? How many more years of bleak numbers can we stomach before we see change? Failure of this length and magnitude would not be tolerated for any business product, so where’s the disconnect and what can we do about it?

One big issue: diversity stats rarely tell the full story.

In tech, data talks. For better or worse, how diversity data is collected will determine if you get the story you want to hear versus the one you need to hear. For example, a large company might report “increased diversity” broadly — without collecting data that would illuminate that the majority of Black and brown employees are in positions with little to no power or support to advance. Or, most commonly, diversity reports might aggregate data into overly broad categories — for example, broadly treating all employees who identify as Black as a monolith, ignoring the nuances of the different experiences within racial groups, such as African American versus African versus Caribbean versus Afro-Latino. This can happen in other ways too — many diversity reports reveal the total number of Latinx employees, or the total number of male and female employees, but fail to provide the total number of Latina employees. These approaches are lazy and flawed, and they prevent companies from understanding and addressing the real and specific barriers that so many experience.

But there’s more: tech companies aren’t being held internally accountable for the diversity data that they are reporting.

Many companies fail to quantify, track, and scale corporate diversity, equity, & inclusion (DEI) goals like they do other parts of the business — and there’s often little structure (metrics, funding, etc.) to ensure companies are achieving what they say they want to achieve — if they’ve even declared a goal in the first place. No business unit would be considered successful if return on investment (ROI) increased by only a few percentage points (or fewer). But, sadly, that’s the rate of DEI growth. Everyone knows that diverse teams are more productive and overall contribute to a higher ROI, but, from a business standpoint, diversity is not set up to meet or exceed expectations.

So, we know that releasing diversity reports is not changing the landscape. How can we positively impact the numbers and create business goals that acknowledge the necessity of diversity? First, know your baseline. Next, know your target. Finally, build measurable, achievable strategies that will get you from point A to point B. When it comes to DEI, most companies need to start at the very beginning: knowing your baseline.

It’s time to delve into point A so that you can get to point B. Here are three things to consider measuring to ensure your actions are improving diversity statistics and informing strategies that move the needle:

1. Disaggregate employee data by race, gender, and employee level. Treating employees as a monolith helps some and harms many. It’s a reductive (but widely used) approach that makes it inevitable for you to miss critical insights about the different experiences and challenges particular communities face — this, in turn, perpetuates a cycle of inequitable and broad solutions that don’t fit employees’ unique needs. Diversity statistics often incorrectly group people under a broad umbrella, like how all non-white and non-Asian people are frequently categorized as “minorities.” This has to stop. The experience of a Latinx person is different from that of a Black person, just as the experience of a Black man is different from that of a Black woman — categorizing different races and genders under one group is both inaccurate and has the potential to harm those who face unique barriers at the intersection of both their gender and race. On top of race and gender, understanding where power is concentrated is critical to addressing power imbalances in your company. For example, if a majority of the Black women in your company are in entry level positions, then you know you need to design unique programs (mentoring, sponsorships, internal and external trainings, etc.) to help retain and promote them into leadership positions. You won’t know where to start until you know what the baseline is.
2. Break down employee data by raw numbers and consider them in proportion to all employees. Some companies tout big increases in women’s representation — and, at first glance, you may agree. But, have they really increased the overall percentage of women employees at the company, or just the raw numbers? Increases in raw numbers are a step in the right direction, but it’s important to keep an eye on the change in proportion, as well. Big increases may indicate an overall busy hiring season that also saw the hiring of many men, which doesn’t significantly alter the company’s gender ratio or improve the overall gender parity of the company. Increasing the number of one group of people in tech is progress, but the same inequity problems persist if the proportion remains the same.
3. Measure employees’ feelings of inclusion and disaggregate the results. Employees’ feelings of inclusion are key to predicting retention, and are just as important for diversity efforts as hiring — they are two sides of the same coin. You can measure employee satisfaction through employee and employee resource group (ERG) surveys. And I may sound like a broken record, but it’s not enough to look at how “women” are responding — the data must be disaggregated by race and gender to be helpful and to promote equity. How are Black women reporting their satisfaction levels? What about Latinas? Or Native American women? Their responses will be illuminating. If employees are exiting out the back door, bringing more people through the front door won’t change long-term representation and equity. Interrupt the cycle and change your workplace so people aren’t looking for the back door.

It’s no secret that the tech sector lacks diversity, and it’s past time to hold companies accountable — in both how they collect employee data and how they turn data into results. If you aren’t looking at the intersectionality of your employees and actively addressing specific aspects of their identity, then you aren’t being inclusive. And, if you aren’t being inclusive through both diverse hiring methods and programs that support equity and inclusion, and thus retention, then any change will be short-lived. Diversity is easier to measure, so it’s the start, but not the end — equity and inclusion are harder to measure, but are what help retain diverse employees. The ultimate goal of a more representative industry is a feedback loop that will positively impact every company and, ultimately, our entire society. But you can’t get where you’re going if you don’t know where you’re starting.