Applying Talent Intelligence Methodology to DE&I

From The Talent Intelligence Collective Grand Jamboree, Director of Insight & Talent Analytics, Elizabeth Birrell discusses how to apply TI methodology to DE&I, particularly at a senior level.

Let's talk about how we apply TI methodology to DE&I or building representative teams, particularly at a senior level. Now I'm not going to make the case for representative teams, for difference, for representation at a leadership level, for a culture of belonging. The data is very much in, as you'll all be well aware, that representative teams make more resilient, more innovative, and ultimately more productive businesses. Our clients, whether internal or external, know it too, which is why this has become an ever more urgent mandate for us for decades now and why it isn’t going away – because while I think the case has been made, we will all still see that there’s a huge amount of equity work left to do, particularly at senior levels. 

We’ve had the pleasure of seeing that work really gather momentum, and I’ve come to think of it as a virtuous cycle: as we hire leaders from underrepresented groups, whatever that means in a particular industry at a particular level in a particular place, we build representation and motivate other to join the team or to excel within the team and step forward for advancement. In other words, a little progress means a lot. Often the sentiment I hear from clients is that there’s pressure on them and their teams to build representation but a real lack of clarity about what that means, what success might look like, and what a feasible objective looks like for their teams. This uncertainty can make it difficult to get started, maintain momentum, and celebrate success or recognise failure. Good TI can help talent acquisition teams, and ultimately HR leaders, solve all four of those problems: to get started, keep going, reward progress and diagnose underperformance. That’s really the point where we come in. 

Leveraging Labour Market Data

We support clients by using external labour market data to do three things: to set targets that are ambitious but achievable (we want to lead our industry, we don’t want to give our teams an impossible task) and also to establish what representation means in each case. That second one, I think, is often overlooked in DE&I work, but that’s the critical measure of success here. Finally, our work is to make recommendations to convert this into actionable insight: here’s what success looks like, how far you have to go, and how you close the gap. Yi Ting from Micron actually did a really interesting session just an hour or so ago on setting goals in particular, and I’d highly recommend nipping back to listen to that if you haven’t, as she made a really astute point that got me thinking: who should be in the room when these decisions get made?

To who we are, our background in this is that for decades now, clients have asked our Search & Talent Pipelining business to help them find candidates from underrepresented groups, usually in scarce and senior roles. Before we kick off, before we make that promise, we’re doing our own feasibility checks – not just to see what’s possible in that talent pool, in that region, at that level (ambitious, but achievable, remember) but also what our goal looks like: what does representation mean here. Sometimes that’s really straightforward: our early inclusion work largely focused on women, particularly women in leadership roles in male-dominated industries – at the time I’m talking about this included FS and tech, where I know huge strides have been made, and also industrial (oil and gas, engineering, transportation, heavy manufacturing) where there’s still a huge amount of work to do. Representation in that case  (at least in regions where there isn’t significant exclusion of women from the workforce, as in parts of the developing world) looked like 50%. A business where 50% of senior leaders were women, where 50% of professionals in technical roles were women, and where women were hired at an equal rate and advanced at an equal rate. If that sounds like old hat – as the work of DE&I has certainly expanded since then – it’s worth recalling that less than 5% of Fortune’s Global 500 companies have female CEOs.

Elsewhere it’s more complex to define representative – which I’d argue is the first step to any meaningful change. When you’re considering culture or ethnicity, demographic data for the region in which you’re hiring is a factor, of course, which is where national statistical data helps us – although even that has its quirks, just look at  Germany – and education data gets us a few steps further – graduate numbers by discipline where we can get them, enrolments where we can’t. Our consumer base's diversity is also important if we’re a consumer-facing business. 

The impact of Location on Diversity, Equity & Inclusion

Knowing how diverse the city you’re hiring in is (or isn’t) used to be a very important first step, but a complicating factor now, which I’m sure everyone in this room is working on in some area, is that remote working has allowed a lot of industries to broaden the search, at least in certain roles, meaning that search region isn’t just about commuting to the site, it’s also about salaries and cost of living and, inevitably, representation. Levels of flex and fully remote working have started to normalise – at least where I am, in the UK, dropping down from mid-COVID’s massive high but probably never now returning to a pre-COVID low, and that’s created chaos, as you’d expect, and a critical mass of think-pieces about the return to the office, and also an opportunity: can we put this role somewhere we could attract more applications from individuals in historically excluded or marginalised groups for this team, for this function, for this business?

I’d go so far as to say that’s a challenge we should be taking to every role where it makes sense: why wouldn’t we be flexible? Why wouldn’t we use this vacancy to open the doors to leaders who can make our executive team more representative either of the region we operate in, or the consumer base we serve? A lot of DE&I work is long-term, tie-ups with universities, increasing exposure of new graduates, working with internal advocates, developing ERGs – not just worthwhile, critical, but you might not see the results during your time in role. Where there are opportunities to do good work at a stroke, good TI gives us an evidence base to make that case. 

DE&I Competitor Benchmarking

Having defined what representation looks like, what our goal is, we then need to understand the starting line: not just how diverse is our current function, leadership team, or organisation, but what’s the state of the industry? How do we perform against our peers? This kind of benchmarking is often data that’s never been available before, and it tells us how far there is to go, as well as what an industry-leading position would look like. It’s possible to be a long way from truly representative, however we’ve defined it, and still lead your peer set in diversity levels across all those measures that count: not just headcount but hiring and early careers and advancement and exco, your leadership layer.

Now just in case that wasn’t enough to make the case for TI driving equity work, I think it’s important to acknowledge where we’re at. In our working lives, we’ve seen a distinct shift I think from female representation to other underrepresented groups: culture and ethnicity is a huge focus area. From a UK perspective, there’s still a massive amount to do. For example, in 2021 there were no black CEOs, Chairs or CFOs in the entire FTSE100, and according to Origins, the classification system designed by Richard Webber, which I’d really encourage you to check out, the line-up was no more diverse overall than when analysis began, nearly ten years ago. In Australia and North America, Aboriginal, Indigenous and First Nations representation in prominent positions is even lower.

What does the future hold for DE&I?

A more recent shift though, is that clients are looking at building diverse teams in even more complex ways: this year alone we’ve seen a huge spike in interest in understanding equity work focused on members of the LGBTQ community, and on people living and working with disabilities. Region by region, role by role, that introduces more complexity again to the question of ‘what does representation look like here?’ Or more simply, what is our goal? We have pretty solid data in most of Western Europe (In the UK, for example, 71% of Gen Z identify as straight, compared to 91% of Baby Boomers), in the US, pockets of good data in Eastern Europe and south-east Asia, but elsewhere as you’ll be well aware there is a huge cultural and political element when it comes to questions of identity. 

In most cases we can build a persona to understand the number of people who present as female, for example, or for people who likely belong to a particular ethnic group by understanding the background, but it’s a very different thing  to try to understand questions of identity, or even community. That’s the frontier I’d say we’re facing now: with an established methodology for defining representation in terms of gender and ethnicity, and testing an emerging methodology for the doing the same work to define representation in terms of identity, sexuality, and disability. 

So how are we doing it? Let’s say we land on a figure for meaningful representation that we’re happy with, how do we start using the tools we have to decide what’s feasible when it comes to identification, approaching, hiring self-identified members of these communities? 

The impact of AI and machine-learning

If like us you’re using AI or machine-learning enabled software, either to search professional communities or to build and encode the personas your teams use to do that, you need some benchmarks. It’s old-fashioned, but the best way to start to understand who’s leading the market here is desk research: it’s awards and events and conference speakerships and press coverage (not just press releases) and ERGs, both their existence and their membership and activity levels. All of that secondary data helps us focus the lens on a small group of businesses which appear to be high-performing at hiring and promoting, for example, people living and working with disabilities within their organisations. That gives us a sample so we can ask the question, what are the characteristics of a high-performing organisation? What clues should we be searching for on a company level? What interesting correlations exist here? That allows you to expand that shortlist: where your high performers are, for example, established software companies with East Coast US headquarters with high start-up exposure, relationships with Stonewall or GLAAD and a strong under-25s intake, the next step is to go away and find a dozen more that fit that pattern. 

Next we’d look at talent: what does a qualifying profile look like, according to self-identification, and what characteristics does it have? This is where we start to see more interesting patterns emerge, and also where we should start to take note of the potential for unintended consequences! For example, people who self-identify as not straight, or as members of the broader LGBTQIA+ community, skew younger, while people who would self-describe as disabled skew older, and across both groups people are more likely to be born in North America or Western Europe. 

Now we’re never going to conduct research on the scale required to challenge the numbers the secondary research gives us, even without the complicating factor of the political climate in which we’re finding our candidates. What we are able to do, however, is compare like with like across the industries or companies of interest, and offer actionable insight, recommendations: at which level to hire, in which companies or industries we’ll have better luck building a representative team, whether that’s to welcome more neurodiverse individuals or properly represent all the communities we serve within our customer or consumer base.

The importance of primary talent research

There’s a second string to all this, however, which I believe is more unique to the work that we do – as a third party – and can’t always be reproduced in-house even by the most sophisticated TI groups, and that’s primary insight. Primary insight capability represents roughly half the headcount of my team today – primary simply because it’s data being gathered for the first time, for this purpose. Because it isn’t enough to understand what representation means for any individual role, or even how we’re performing as a business, as an industry, and within our peer group. Really, what our clients want TI to tell them is, given how far we have to go, how do we move the needle? And we’ve found that secondary data alone isn’t enough to tell that story. 

An ethical and compliant primary data approach identifies individuals in the relevant talent pool for in-depth, structured conversations, treating them effectively as talent research respondents, who as they aren’t in process and aren’t speaking to hiring managers at a named organisation, are more likely to give us a transparent and unbiased look at their experiences. In the case of equity work, our job is to understand the experience of individuals from underrepresented groups, whatever that means in the function or region we’re working in, and how it differs from their majority peers. This might include standard themes like salary and benefits data, EVP and attraction strategy, recruitment approaches, how and where they look for jobs, understanding network strength, but it should also create an action plan for future recruitment. In the past we’ve had great success with actionable insight that’s ready to implement: in one role, female respondents had strong preferences regarding location which didn’t match their male peers. In another project, BAME respondents were uniquely more likely to look for and find jobs within their personal networks, while white respondents were more likely to turn to professional social media or industry job boards first. It’s also powerful – and instructive – to learn about the negative experiences individuals have had with recruitment approaches, if only so we know what not to do in order to create a recruitment experience which is tailored to be attractive and crucially not othering or alienating to exactly the individuals we’re working to attract. 

Ultimately our work isn’t to source or identify, it’s to count what counts: what does success look like, what’s possible, what’s the best way forward? With that in mind, I’d like to close by sharing the recommendation we make most frequently in projects like these, based on the data we’ve discussed today.

Is the job description reflective of your DE&I objectives?

First, and maybe most important, is to go back to the role profile or job description, and check that it only contains what it really needs to contain. It can be effective to sit down with the hiring or line manager and start again, stripping out everything in the template so you can focus on what’s critical: what experience or characteristics or qualifications can’t we live without? We often find these things are freighted with unintentional bias, with exclusions based on assumptions that might be out of date or actively work against our aims when it comes to building representative teams. 

For example, if someone is in a relevant role today, do they REALLY need to be graduates of an Ivy League, red-brick university, or ex-Big Four? As you’ll be well aware, these institutions have their own representation work to do, and by tying ourselves to one particular background, almost as a shorthand for quality, we tie ourselves to the demographic associated with that background and miss out on the opportunity to capture arguably the most exciting candidates of all: those from a non-traditional background for their role, who have successfully broken into the field and excelled. The same applies to educational background – particularly when we’re looking at experienced candidates: if someone is in a relevant role with a competitor today and appears high potential, should we exclude them because they don’t have a Master’s degree? And have we fully considered the consequences of that choice? 

A really powerful use of TI is to show internal stakeholders how, with a simple click of a button to remove one filter, we can grow our potential talent pool from a few to a few hundred.

About the Author

Elizabeth Birrell is the Director of Insight & Talent Analytics at Armstrong Craven. She joined 8 years ago and now runs the TI consulting group, but Armstrong Craven has been in this world for over 30 years, and the team made the transition from a back-office function supporting talent acquisition with research to delivering bespoke insight directly to our end clients.

Along the way, we've worked with many of the biggest companies in the world in tech, financial services, healthcare, consumer and industrial, and gained what I think is still a unique insight into what clients know about TI, what they need from it, and its critical importance. All the challenges talent acquisition faces can be better understood through the lens of good talent intelligence: who and how and where and when to hire, what we should be paying, talent supply and demand, what candidates want and expect, how our competitors are hiring, how well our EVP works in any talent pool, every facet of TA and much of HR can really be supercharged by this data, and there's a huge amount still to do, which I think makes this an exciting industry (maybe a micro-industry?) and an exciting time to be in it.