Article: Red Hat's global head of DEI on why data and diversity go hand in hand


Red Hat's global head of DEI on why data and diversity go hand in hand

DEI analytics are an under-utilised but highly efficient subset of HR analytics. We hear from Shuchi Sharma about how to leverage data for better DEI outcomes.
Red Hat's global head of DEI on why data and diversity go hand in hand

I think we have to constantly be vigilant about ensuring that we are compassionate and inclusive for all human beings.


HR analytics have been around for a long time, but diversity, equity, and inclusion analytics have taken longer to catch on. Even today, the concept of data-driven diversity is under-utilised, and the general corporate approach to diversity metrics can be hit and miss, partly because the most accurate results can only be shown over a very long period of tracking and are difficult to separate from external factors. However, those companies that seriously leverage data to drive DEI come out at an advantage in the long run.

People Matters met with Shuchi Sharma, Vice President and Chief Diversity, Equity & Inclusion Officer at open source developer Red Hat, to get her perspectives on taking a scientific approach to DEI. Sharma joined Red Hat in 2022 from SAP, where she was VP of global diversity and inclusion, and where she had also previously built a background in business transformation. Trained as a scientist, she brought that experimental and data-focused orientation to her work in DEI. Here's what she told us.

What are your thoughts on how the corporate approach to DEI has evolved over time?

There's been no real guidebook on how to succeed at DEI for a long time. Organisations often start by experimenting, throwing things at the wall to see what if anything will work, but they take a long time to develop cohesive strategies that fit into what the business as a whole is doing. And that's where the opportunity arises to take a more scientific, systems-based approach to DEI and look at it as an enterprise-wide strategic transformation.

When you apply that lens and framework, you can start to ask questions like: What are the root causes we see in the data? How do we address those? How do we use the data to design talent processes that are going to be more equitable in the long term for all people?

And long term here means decades long. Skill-based change happens relatively quickly, process change takes longer, but culture transformation is a 10 to 12 year journey in most organisations, one that is impacted by multiple internal and external factors. Having a more structured systems-based approach, where we experiment and iterate and measure progress and even the lack of progress can potentially have more impact in the long run than the older, programmatic activity-based focus.

But data is often neglected in DEI even though it's already there as part of existing people analytics. Why is that so?

Not everybody has the needed training. This approach does require some proficiency in data and data analytics, to understand how you can take a data set, disaggregate it, look at it based on various different dimensions, and then turn that into a compelling and fact based narrative that you can use to drive your agenda within the organisation. And with DEI, you need two sets of data: the operational data, the metrics that we traditionally have used, and also the experience data from our employees and associates. You need to be able to find the differences between the operational data and the experience data. But today, there's really no sort of training for a DEI leader that touches on this proficiency.

The organisation also has to have a certain maturity, just to be able to collect that data. A lot of data governance needs to be in place, along with the ability to access the data, analyse it, and report on it. Unfortunately, a lot of organisations don't have that.

Let's say an organisation develops the capability to collect and analyse this data. What's the next step?

I personally have found with that leaders in our organisation, especially with Red Hat being an engineering-led organisation, they are more open to discussions built upon data. If I can show them where are the disparities in representation and what might be some potential root causes, they will quickly move towards the question of how they, as leaders, can support initiatives aimed at changing that. And this is really important, because ultimately, it's the leaders who have to drive this change. I can be a partner. I can equip them and enable them. But they are making the people decisions that are going to drive the culture.

Not too long ago I had the opportunity to present to our APAC leadership team and I showed them the data for the region. It was the first time they'd seen the data presented in this way, and they were very open to what it showed, and accepted that the situation was different from what they had previously assumed. There is a lot of power in that, because once people understand that they had a certain bias in their thinking, that is the first step to diffuse that bias and move to a more fact driven discussion around addressing the issue.

What do you usually look for as you go through the data?

Generally, the data will show degrees of inequality and degrees of disparity in representation. But more importantly, it allows us to figure out where we need to focus scarce organisational resources. I can look at the data and determine where the pockets of opportunity are, where the areas of risk are, and how to mitigate those risks.

I can decide how to invest the dollars I have for the greatest impact. I can see that one function is doing well in gender balance but another is not, and focus there instead. I can go even deeper and see that in a certain job category, we have a strong female pipeline but a low promotion rate. And I can dive into that data to find out why; I can talk to more people, understand their experience, and gather data in a different way to conclude that a certain portion of the organisation faces some challenges, and if we apply a laser sharp focus in this area, maybe we can bring up representation, minimise attrition, and also start to address some of the cultural or leadership factors that may be creating challenges for these particular areas.

What role do you think today's advancing AI capabilities will play in a data-driven approach?

There are definitely both risks and opportunities. Risks come from the possibility that the algorithms will have bias programmed into them, and opportunities come in terms of how we look at data, how we operationalise some of the work, and how we amass and incorporate best practice data into what companies are doing individually.

But anything the algorithms do is based on the past. So you still need an element of human intervention, as well as creative thinking processes. One of the things I've been fortunate to do in my career is a lot of design thinking work with customers, and I've tried to introduce that into the DEI space as well, in terms of building personas for use in how we design processes. We're now building a partner ecosystem for the APAC region to really advance the DEI strategy that we have established globally, and we've used design thinking methods and processes to facilitate that work in a very short time.

When you put the rapid pace of AI development together with the long time it takes to see change in DEI – do you think there's a point where we have to pause and say 'We've done everything we can for now, we should pause and let it evolve by itself'?

Based on the data I've seen, organisations that have taken their foot off the gas pedal have seen a slide back in their progress. DEI is proven to be something you really have to keep the momentum and focus on just as you would in any other function of the business. Because if you don't, organisational inertia makes it so easy to slide back into what you were doing before.

Perhaps one day we'll get to a state where we have designed all of our systems and processes to eliminate certain biases for certain groups. But even then, the world is changing so much around us that we always have to take into consideration how those changes are showing up in our own organisations, and then also continue to correct for them if they're creating adverse impact for certain populations.

That impact will keep changing. For example, the Pride community may no longer face challenges in five or ten years' time, but another group or population might instead become adversely impacted for political, social, or other reasons. And so I think we have to constantly be vigilant about ensuring that we are compassionate and inclusive for all human beings. That is a journey humanity is on.

Read full story

Topics: Diversity, HR Analytics, #DEIB, #HRCommunity

Did you find this story helpful?



How do you envision AI transforming your work?