Putting Tech Diversity Puzzle Pieces Together

I am going to write about an issue with political ramifications while doing my best not to be political, so please accept these thoughts in that light.

One thread going through the proposed ban on legal immigration to the U.S. is the effect it will have on the tech industry.  Those companies are concerned that some of the talent they need from other countries will be unable to either enter the U.S. on an H-1 visa or be allowed to immigrate here.

Another issue that the tech industry has struggled with is hiring a diverse workforce in the U.S.  Much has been made of the lack of women and (domestic) minorities in the tech field.

However, there are more programs to teach tech skills to minorities and girls than you can shake a stick at.  A Google search of “minority tech training” garnered almost 18 million hits and “girls tech training” got 198 million. So, there is not a shortage of opportunities to obtain coding or other tech skills in the U.S. and these programs have created a pipeline of talent.  Likewise, there are specific tech incubator programs for women and minorities who want to start their own companies.

Why is all of this important?  Primarily because for a company to be innovative it needs to look at the world through a window and not a straw.  There are more tech users outside of the U.S. than inside, so to be successful internationally companies need foreign talent.  Shutting our borders and wanting our companies to compete overseas is a difficult problem to solve.  Women and minorities outnumber white males in the U.S., so the companies that harness those perspectives are likely to be the most successful ones.

So, what might be the barriers to connecting talent to opportunity?

  • Hiring like us. I’ve written before about the built in bias we all have of wanting to be with others who have a similar background.  This is very prevalent when it comes to which schools the person attended, which sports s/he played, etc.  This can be alleviated by:
    1. Removing names from resumes.
    2. Removing schools and extra-curricular activities from resumes, unless you have data supporting their use (and the literature on the validity of training and experience measures is not encouraging).
    3. Recruiting where you have never recruited before. For instance, go to schools that are not currently represented in your workforce.
  • Candidates being unfamiliar with the hiring process. In our bubbles we think that every step of the hiring process is normal.  But, to use a tech example, if I come from an area without many tech companies, I might not be familiar with “whiteboarding” code problems as part of an interview.  Being transparent about the steps and letting people know what to expect removes a potential barrier between you and a qualified candidate.
  • Hiring processes that invite bias. Whether it is how you score your interviews or evaluate resumes, having an evaluation rubric will reduce bias.
  • Echo chambers on interview panels. Have diverse points of view (e.g., people from different departments) on your interview panel.  This is likely to encourage meaningful follow-up questions, even within a structured interview.

I do not think that lowering standards is an appropriate response to creating a more diverse and innovative workforce.  Building walls (metaphorical or otherwise) makes the problem worse.  Rather, companies need to build more bridges to find qualified candidates to bring different perspectives to their organizations.  Yeah, I get that it is more work, but a competitive marketplace demands it.

Filling Diversity Buckets

With great fanfare, Intel announced recently that it is making progress in meeting its diversity goals. I’m not going to pick on their numbers as their current demographics are what they are. There are some good lessons we can learn from how they approached the issue.

  • You have to be holistic. They understand that culture, recruitment, and retention all play a part in attracting, hiring, and keeping diverse talent.
  • Drill down in the data. Intel looks at hiring and retention in different job categories. Saying that you are diverse overall, but not in high paying jobs, is not much of a victory.
  • It takes significant resources to make changes. Developing a pipeline of diverse talent requires money in scholarships, helping schools, etc and finding untapped recruitment pools take time and money.
  • Just like any other business outcome, the goals are reached only if they are measured AND if there are rewards for doing so. Sorry, but you cannot assume that people will strive for noble goals out of the goodness of their hearts.
  • It’s more than hiring numbers. You need to get the compensation and culture right to retain people. Oh, and selecting and developing good managers, as that has a huge influence on turnover.

This article goes into a bit more depth about the challenges Intel are facing. Not surprisingly, there are concerns about balancing multiculturalism (celebrating differences) and integration (making one big happy family). It also points out that if people are spending time on diversity programs, it takes them away from their “real” job (unless they are in the diversity department) and makes it tougher to get promoted and make the higher ranks more diverse.

Just as importantly, this is a case study about what doesn’t work. There is a lot of good science about unconscious bias. However, Intel found that training people about it doesn’t really affect their decisions, or at least as much as tying their compensation to it does.

You can see how Intel treated this as a supply chain as a human resources issue. It’s an interesting approach that probably led to some creative ideas. You’ll note that there is no discussion about lowering standards, which is divisive and bad for the businesses. It is also something that probably is not discussed when they are sourcing equipment. Just something to keep in mind when making important business decisions.

Thanks for coming by!

Please provide this information so we can stay in touch.