What gets measured gets done to make organisations more diverse and inclusive. A large financial company in America’s mid-west was concerned about harassment of women within the organisation, so it used Diversio, the world’s first AI-based DE&I platform, to determine the extent of the problem.
It came up with the surprising result that, while harassment was an issue, it was mainly experienced by LGBTQ+ employees.
“We found that people who identified as LGBTQ+ were about half as likely to say that their opinions were valued by their team and three times as likely to report harassment,” says Diversio Founder and CEO Laura McGee.
“That was a lightbulb moment for the company, and then they were able to develop, with our support, an action plan, against which they achieved meaningful results over the next 12 months.”
This is one of McGee’s favourite examples of how what gets measured gets done and the importance of focusing on the right metrics to support DE&I. She launched Diversio in 2018 to fill a gap in the market for a software solution that would “transform data collected in a way that provides actionable insights.”
Her early career was in management consulting, helping organisations to develop diversity and inclusion strategies. “On the flip side, I did a lot of work with governments figuring out how to engage more people in the economy,” McGee recalls.
“I discovered that number one, the commitment to diversity, was there. But where organisations fell was in not knowing how to do it. I kept hearing, ‘we don’t have the data, we don’t have metrics, we don’t have KPIs, there’s no way to measure what programmes are having an impact.’
“When I took a step back and looked at how companies were typically doing diversity, there wasn’t a platform helping companies execute against the strategy, despite how critical it was at both the company and economic level. We said, ‘we want to build the world’s first platform focused on diversity and inclusion, and we have some interesting technology that will help us move the needle.'”
Inclusion metrics
She believes that although events over the last two years, such as the Black Lives Matter movement, had raised awareness of the importance of creating working environments that accepted and empowered everyone, organisations generally didn’t understand which data they needed to collect and what was important for their business.
They tended to only look at leading indicators, such as how many women are on boards of directors. Diversio’s software goes beyond those surface-level metrics by measuring inclusion metrics, which allows them to identify deeper underlying problems that might not be apparent through diversity data alone.
Inclusion metrics cover areas such as whether employees feel that their opinion counts with their teams, experiences of harassment and if they have access to networks, mentors and sponsors. The data is then broken down by demographic group – gender, race, ethnicity, sexual orientation, disability and socio-economic background.
“From that, you can see where people from a non-dominant group are scoring lower than people from a dominant group,” McGee explains. “We often see, for example, that almost every white male in an organisation has a mentor, but Black women are much less likely. Sponsorship is consistently the lowest scored across almost every sector and business size. Creating a culture where senior leaders invest in junior staff is a consistent problem. Change has to be on an individual level, with the senior leadership team making an effort to create opportunities for underrepresented individuals.”
One of the biggest surprises for many organisations was the least included employees were those with a disability.
The metrics are used to look deeper into employee experience and better understand the biggest challenges faced by which types of employees and in which part of the business and then, most importantly, used to create solutions.
Measuring the pain points
Machine learning is an important part of the process. Firstly, it collects data on employees’ experiences. Says McGee: “We noticed early on that employees don’t want to fill out a 200-question survey and don’t always know how to classify their experience as an inclusion problem. So, we use natural language processing to analyse free-text data from employees. Our algorithm can parse through that data and identify, what we call, inclusion pain points.
“For example, someone might say that a manager came on to them at an office party and, when they reported it, nothing was done. Our algorithm can recognise that’s a recourse problem. If enough people say there’s a recourse problem, we can recommend anti-harassment training with a reporting tool.
“The second use of machine learning is that we are now tracking the impact of different pilot programmes: the policies and interventions that our clients are implementing. As we get that longitudinal data, we can improve the quality of recommendations based on all of our clients’ experiences.”
Failing to invest in data collection and tracking inclusion and engagement leads to companies not getting the most out of their employees in productivity and innovation. Also, MeToo and Black Lives Matter encouraged employees to go public, posting harassment and other negative experiences on social media, which often go viral.
“We are seeing that, socially, people are rallying around the cause,” McGee reveals. “And that can impact the share price. Of course, the scandal breaks and investors and the board ask what’s happening. We’re at the point where doing nothing is no longer an option.”
Investment challenge
Surprisingly, given Diversio’s success over the last four years, there were initial difficulties persuading investors that there was a market for a DEI platform. The perception is that organisations wouldn’t want to spend money on that solution.
“We had to build a sustainable business model that was profitable and generated cash that we could invest in development,” McGee says candidly. “It was a challenge that turned into a blessing. We won some awards, had unbelievable logos, 100% retention, and now we can’t get rid of investors.”
Diversio has been working with Diversity VC to understand the sector better and create a standard for VCs to tackle bias and other DEI issues. This standard is being extended to other industries, including private equity firms, hedge funds and IT companies.
A three-step approach to data
Returning to data collection, McGee offers the following advice to those companies that have made a commitment to DEI but are unsure what to do: “Collect data, give employees a safe space to self-identify and share their experiences. You’ve got to do both. It’s not enough to get people through the door; you’ve got to make sure that they’re adequately equipped, then use the data to guide your programming.
“We find that a lot of people go the other way, where they’ll develop a strategy, make all kinds of public statements, then they go back to the data, and it’s not aligned. So, step one: make a commitment and ensure buy-in; step two: collect data, both demographic and experience; and step three: create a measurable strategy and map back to the data, so you’re solving the right problems.”
