Struggling to Build a Diverse Engineering Team? This Might Be Why: Part 1
Last month, I interviewed over 10 engineering leaders who graciously shared their struggles and successes when it came to building a diverse and inclusive engineering team. Some leaders have had great success building out a team with many intersections of diversity; some struggled to get candidates to respond to their reachouts or through to even the first interview stage. After sifting through interview notes, I’ve pinpointed standout themes related to diversity, equity, and inclusion (DEI) which I’ve collected for you below. In each post in this series, we’ll walk through one theme in more detail.
First up: Demographic data isn’t collected during the interview process.
Many leaders I spoke to said their companies have skipped over collecting demographic data during the interview process. A report from Lever, a popular Applicant Tracking System (ATS) found that only 39% of companies were using any kind of data to control for bias in their interview processes.
Instead, their diversity initiatives are measured based on demographic data from employees who have already been hired on. While assessing diversity of your current team is important, this approach provides you with a lagging indicator that means it’s far too late to make an intervention if your hiring processes are not reaching the communities you’ve prioritized.
There are also those engineering leaders who feel that tracking these characteristics is inherently racist or sexist (it is not), or that it causes “reverse racism” (which is a myth), or “lowers the bar” (which it does not).
The bottom line here is that you can’t hire candidates from different backgrounds if they are not in your candidate pool, or if your interview processes have inherent biases that may prevent certain groups of people from getting a fair shot.
It’s difficult to debug a system without metrics to reference — whether it’s a distributed system or one made up of people.
When talking about collecting data about demographics, it’s important to note that this data should generally be viewed in aggregate. It’s not being collected for the hiring managers to make decisions about individuals, but rather for those responsible for hiring processes to make decisions about the systems they’ve designed.
Wait, I thought I can’t ask about this stuff?
There may be a sound reason for not collecting this data. Namely, demographic data generally references protected characteristics, and in most countries, it is illegal to require the candidate to answer questions about their gender, race, religion, parental status, etc. This leaves recruiting teams relying on “observables” — if I’m interviewing someone who presents as a woman and she refers to herself with she/her pronouns, I can pretty confidently record that. But characteristics like race, religion, neurodiversity, or sexual orientation aren’t so obvious. With those characteristics as well as gender, the best approach is to allow your candidates to self-identify in a survey.
Data collection should not be done in an interview — that is illegal in many countries. This data isn’t for a hiring manager to make a decision about a candidate, but rather about allowing the hiring team to hold themselves to account when it comes to designing an inclusive and equitable process. And as mentioned before, this data should be viewed in aggregate, but in reality, the interviewing panel will have access to at least some of it simply by meeting the person — like the “observables” mentioned above.
In the USA, employers over 100 people are legally obligated to report demographic information about their workforce, and it’s become common practice to collect data from applicants for that reason. The UK explicitly states it’s okay to ask about protected characteristics during the application process in order to report on discrimination, but that people making hiring decisions shouldn’t have access to the data.
- If you’ve tried this: Relying only on observable characteristics and making an assumption about a candidate’s race, gender, or other characteristic
- Try instead: Asking the candidate to answer a survey about demographic information, after you’ve explained your data privacy policies and that the data is used in aggregate in order to control for bias in the interview process.
If your company uses an ATS like Lever or Greenhouse, there are voluntary demographic surveys available in the product that can be tweaked for your needs.
What about anonymized applicant screening?
Anonymized applicant screenings, or blind screenings, are a common way that teams try to control for bias in hiring, especially at early stages of the interview process. Hiring teams may do an application review where all personal information is stripped away before they are reviewed. (I’ll cover this in depth in the next post in this series.) This technique is not incompatible with data collection from your applicants, because the demographic information is not meant for the hiring manager to make a decision about any specific individual.
- If you’ve tried this: Using anonymized application review as a way to control for bias in your hiring process
- Try this: Use demographic data to demonstrate that the anonymised screening process is indeed unbiased
I don’t care about demographics. I just want the most qualified person.
Another myth here is that hiring is or can be a meritocracy.
So often I’ve seen interview processes designed by white men in their 20s, none of which have caretaking responsibilities, consistently result in hiring other white men in their 20s without without children or dependents. In these cases, the application and interview processes have inherit bias that disadvantages people from different backgrounds. This may be a job ad with overstated requirements, a technical challenge that requires ample time on nights and weekends to complete, or a poorly planned interview stage without evaluation criteria or trained interviewers.
It’s difficult to see these trends without meaningful reporting of the candidate funnel. No need to boil the ocean here or get fancy with data pipelines. A Google sheet might be just enough.
- If you’ve tried this: Measuring the success of your diverse hiring initiatives at the offer stage, or after the candidate has been hired
- Try instead: Create a lightweight reporting framework to measure candidate advancement throughout your interview stages, so that you have time to make interventions early
You can start with questions like: how does the demographic breakdown of my applicant pool compare to those that made it to the first interview stage? What about first stage candidates and candidates at the offer stage? This can lead you to targeted interventions, such as better sourcing methods, or reevaluating an interview stage that seems to disproportionally favor candidates of a certain background.
Beyond diversity interventions, this reporting can have a positive impact on the candidate experience for everyone in your hiring process. If you notice that there are high number of candidates getting through to a certain stage and then a massive dropoff, it can be that your assessments in a previous stage aren’t working as intended. You’ve ended up wasting the candidate’s time and your interviewers’ time, but it’s hard to see without reporting.
If this sounds like your experience, but you’re not sure where to start, get in touch. I can help get you started.
Up next in this series:
Assuming that anonymized interview stages create an inclusive interview process. Taking away personal data from your interview process doesn’t remove biases. In fact, it can strengthen the ones you’ve build in to the system by design.
Focusing too much on diversity without investing in equity and inclusion. What good is hiring talent from many different backgrounds if your company is set up to favor those already in the dominant culture?
Looking only for “senior” talent, or inflating job requirements in the job ad to cover every nice-to-have quality. Do you really need a senior engineer with 10+ years of Javascript experience to fix every UI bug?
Using the wrong reachout methods and messaging for the communities they want to see represented. How can you authentically let a candidate know that DEI is important to your team and company without it coming across as the only reason you’re approaching them?
