For quick DEI wins, focus on employee data
Inequalities in technology and the wider STEM field persist, despite a great deal of attention being focused on trying to level up. How can companies use employee data to shift the dial?
Tech employers of all shapes and sizes are finding it hard to build workforces that reflect the wider population.
The most obvious imbalance is one of gender. The top line figure usually cited for women in all tech related toles is about 26%. That figure masks variations between certain types of tech jobs. For example, Statista reports that globally, women constitute 5% of software developers. Black women hold only 0.7% of tech related jobs in the UK despite making up 1.8% pf the overall workforce. Women tend to be better represented in areas such as project management, sales, data science and UX/UI. They're also better represented in larger firms than smaller ones.
Tech employers almost always claim that diversity of thought is crucial, not just because it's fair but because it's profitable. The global diversity, equity and inclusion (DEI) industry is worth billions. Millions of tech employees worldwide will have undertaken training in areas such as unconscious bias in an attempt to stop the same patterns repeating themselves for eternity. Yet substantive change remains elusive. Why?
There are multiple complex, overlapping social issues at play, and likely more than a little bias. But at least one of the reasons for the apparent ineffectiveness of diversity initiatives might be more straightforward and easier for employers to control - the failure to collate, analyse and draw insight from employee data.
Data driven diversity
It stands to reason, that if a company is trying to solve a problem - in this case the homogeneity of its workforce - that the logical place to start is assesment of the extent of the challenge. Yet it would seem that a majority of businesses don't collect anywhere near sufficient data to enable them to start identifying areas for improvement.
Gender pay gap data is a good example. Every company already has this data, although in the majority of cases it is unlikely to be centrally collated and reviewed. SMBs (defined as those employing fewer than 250 people) employ three fifths of the working population and are not legally obliged to report pay gap data. The Gender pay gap service informs us that only 79 SMB employers in the categories of "Information & communication," and "Professional scientific and technical activities," did so for the year 2021-2022.
If a company chooses not to collect and report gender pay gap data, how many of these companies are likely to collect data on the ethnicity, religion, sexuality or socioeconomic backgrounds of its employees?
One person who has successful used a data driven approach to recruiting a team is Sherrie Fernandes, Vice President of Product Management and User Experience at G-P.
Fernandes has managed to build a 60 strong team with a 50/50 gender split. Her approach illustrates the importance of data to informed decision making about employees, both current and future.
"I think the first step is just having the data and being knowledgeable about where you're at on this journey," she says. "It takes a cross functional initiative to focus on improving that and it takes continual effort day after day.
"What I mean by that is where I've seen success is in companies that start a programme all the way from bringing individuals in from an internship programme, having a ratio and saying, ‘we are committing to making sure if we hire 10 interns for example, that 50% of them are female, and we're not going to move forward unless we have that.' Bringing that class in together and growing them into full time roles is a really a great method that I've seen work in terms of getting more diverse talent into a technology organisation. That class grows together, and women feel more comfortable coming in and working collaboratively with this group and growing."
Of course, by bringing women into a culture where they don't constitute a small minority, it also becomes more likely that those women will stick around and not be driven out by bro culture and everyday sexism. Fernandes goes on to explain that G-P is developing its own SaaS platform with a view to enabling their customers to build this sort of data into their talent acquisitions.
"We are starting to think about how we bring this concept of talent intelligence to our customers. There's lots of data sources out there. We're working on building out a data infrastructure that can use artificial intelligence to make some of these suggestions, recommendation to our customers, as they're going through and using our platform whether that be talent supply data, salary benchmarking data, all sorts of different data sources as part of our mission."
You may also like
/feature/4334521/tech-isnt-meritocratic
Leadership
Tech isn't as meritocratic as you think
And relying on graduates to fill vacancies isn’t working
/news/4322678/tech-launches-manifesto-generational-change
Skills
Tech She Can launches Manifesto for Generational Change
Launch showcased an impressive community of sponsors, educators and policy makers
/news-analysis/4212923/long-reads-scarlett-johansson-claims-openai-copied-voice
Artificial Intelligence
Long Reads: Scarlett Johansson claims OpenAI copied her voice
But first, the company ignored it
For quick DEI wins, focus on employee data
Inequalities in technology and the wider STEM field persist, despite a great deal of attention being focused on trying to level up. How can companies use employee data to shift the dial?
More to diversity than protected characteristics
Mark Holt, Chairman of employee survey platform Divrsity suggests other examples of data useful for employers trying to create a diverse workforce.
"Have you got diversity amongst the types of organisations that people have worked for? Has everyone always worked in the same industry? Has everyone always worked for companies that are the same size? Or have you got an interesting blend of experiences?"
Who gets in through the door of an organisation is one part of the diversity discussion. Who gets on? Well, that's another matter entirely. Holt explains how data generated by employee surveys can be used to begin unpicking biases working against certain groups.
"One question is to ask whether someone has received an individual award in the last year such as Employee of the Month," he says. "Or have they been promoted? Now you can start to see patterns like fewer women receiving awards or being promoted or Black employees not being promoted at the same rate as we would expect given the overall population. Or perhaps the neuro diverse or the religious population are being treated differently. There are lots of interesting ways you can start to cut data to identify bias."
A critical part of retaining a diverse workforce is building a culture of inclusion. Whose voices are being heard? Holt explains how a company can find this out.
"This is where you start to ask questions such as ‘have you presented a monthly all hands?' Or ‘how many times have you been quoted in the press?' ‘How many interviews have you conducted in the last month?' Once you've got that data you can start to make targeted interventions to keep improving inclusion."
The third aspect is equity.
"Questions can identify whether women are less likely to cultivate mentors for example, or less likely to be a mentor for someone else. If you can identify the fact that men are great at cultivating mentors and women aren't you could look at creating a mentorship scheme for the female population."
"But pipeline"
When technology executives make statements about DEI in their businesses, they almost invariably restate their commitment to doing better, then frequently follow up with a lament that the pipeline of diverse talent isn't sufficient to enable them to recruit as they would like. Schools are often blamed for failing to inspire girls in particular, and there is some truth here. For example, only 19% of UK computer science graduates are female. It is also true that tech employers are not going to find all the skills they need until more people from all backgrounds decide that they can see themselves pursuing a career there. There is a pipeline problem. Nonetheless, what the focus on pipeline and the future often serves to do is divert energy and attention away from what companies can do now to build a culture of equity and inclusion and allow employees from all walks of life to develop themselves and their careers within their business.
"It's so easy to get dragged into that ‘but pipeline' stance," says Holt. "Companies will say they try really hard but the pipeline just isn't there. But I don't think many would argue with improving inclusion inside their organisations. I think that's a very sensible approach.
"Also, a lot of companies don't have the data to be able to talk about some of the great things they are doing from a DEI perspective. Being able to report success is crucial because ESG reporting is getting more important and we're seeing diversity information being requested more and more when RFPs are put out."
Hilda Davies, Senior Manager, Data Science, Data Engineering and Performance, Field CTO at cloud data platform Snowflake agrees:
"By collecting employee feedback and demographic data, businesses can measure employees' level of satisfaction, and identify potential biases or holes in hiring and/or promotion practices. Any disparity seen within the data can equip organisations to tackle any unconscious biases that could favour certain demographics, resulting in a more diverse and inclusive work culture. This in turn can generate better retention rates and less employee churn.
These insights must also be powered by qualitative data by soliciting feedback and input from all employees to ensure the company is representative of the workforce and ensure their voices are heard, and that they are invested in a company's policies. This data must be gathered anonymously in order to produce honest and candid feedback, for which technology can play an important role. For example, dynamic data masking features as part of modern cloud data platforms, enable organisations to define how they want to protect specific data without the risk of re-identification or diminishing data quality."
These are important points. Some employers cite privacy or legal concerns when asked why they don't collect more employee data. Some employees are also reluctant to respond honestly to these types of survey, fearful of the repercussions. Sometimes people just don't want their employer to know about their backgrounds or private lives which isn't unreasonable. Not everybody will want to respond to surveys but by building a culture of inclusion where employees feel safe to speak candidly, the chances of persuading others to do so is increased.
Davies also echoes Tim Holt's point that data gathering and analysis should not be considered a one- off exercise. The process of gathering data on the views and experiences of employees should be as dynamic as the employees are.
"After implementing this strategy," Davies continues, "businesses then need to implement continuous data collection processes for every part of the employee lifecycle, from recruitment and retention to pay and promotion. Each of these areas should be tracked and continually monitored for instances of bias or unfair treatment. Modern cloud data platforms can help in this journey by seamlessly bringing data from various different departments in near real-time, as well as making data accessible for those that need it such as HR or the leadership team. This data can be tracked over time to understand if new procedures are having the desired effect, or whether further improvements are still required."
Data is frequently used for insight into customers and markets for goods and services. It's time companies assigned equal value to their employee data and found a way to gain actionable insight from it. Instead of lamenting a talent pipeline that they can't control, by building inclusion and equity in the here and now, employers may find that the pipeline of diverse talent starts flowing in their direction.
You may also like
/feature/4334521/tech-isnt-meritocratic
Leadership
Tech isn't as meritocratic as you think
And relying on graduates to fill vacancies isn’t working
/news/4322678/tech-launches-manifesto-generational-change
Skills
Tech She Can launches Manifesto for Generational Change
Launch showcased an impressive community of sponsors, educators and policy makers
/news-analysis/4212923/long-reads-scarlett-johansson-claims-openai-copied-voice
Artificial Intelligence
Long Reads: Scarlett Johansson claims OpenAI copied her voice
But first, the company ignored it