Can AI predict employees leaving?

What if you could predict when people were going to leave their jobs before they even knew they were leaving? Tyler Hochman is the CEO of FORE Enterprise, an AI workforce analytics platform proving just how valuable effective data collection can be.

?

We’ve spent a lot of time recently exploring how to clean up and organize customers’ data, so I welcomed Hochman onto a bonus episode of The Business of Tech to walk us through his approach to data analytics and how MSPs can use it to extract advanced, real-world insights.

?

Hochman’s founded multiple consumer and B2B apps and companies, including TH Analytics, Vela, and Modern OneMore, so he knows a thing or two about helping businesses succeed with the power of data.

?

From tackling data cleanup to fascinating turnover patterns he’s uncovered, here’s a rundown of our conversation.

?

FORE’s predictive data

?

The essence of FORE’s mission is quite impressive: it takes into account a variety of external and internal data sources to create an amalgamation of data that can then be used to predict whether or not people will stay or leave a business on an aggregate and individual basis.

?

What kind of data does Hochman use to make such ambitious predictions? He explained the process as a two-part analysis.

?

First, they pull from about 17 external data sources, including census information and macro and microeconomic trends. With that information alone, they can make some external predictions for companies.

?

But to really strengthen their insights, they then couple the external data with internal data. They only use information employees already know could be collected, like schedule adherence, job performance indicators, and utilization.

?

The data cleanup issue

?

Have you tried to clean up a customer’s data to implement AI? Many AI folks I’ve talked to say data organization is an essential step, but it’s rarely an easy task.

?

I asked Hochman how much data cleanup is needed to use FORE, and he confirmed it’s a widespread issue – so much so that his team has actually built products to help their users structure their data automatically. They call it an AI data pipeline, and it pulls from everything from Asana to JIRA to Salesforce to ADP to a company’s unique employee performance KPIs.

?

I’m a firm believer that even without AI, there’s real value in helping businesses organize their data to pull basic analytics. Hochman agrees and added his belief that for businesses under 50 people, you can glean a lot of value from manual data collection – no fancy tech solutions needed.

?

For MSPs, there’s certainly an opportunity there.

?

Maintaining employee privacy

?

Another concern folks might have with feeding AIs their company data is employee privacy. I asked Hochman for his process and solution for protecting privacy, and he confirmed that they only leverage data from KPIs employees already know are being tracked.

?

“As an employee, you understand that your performance, as an example, is always going to be tracked. They're going to know from a utilization perspective, did you do or not do your job? How much did you do your job? How much did you not do your job? They're going to know from a schedule perspective, what time did you clock in? What time did you clock out? Those are understood parts of data collection,” he said.

?

He also mentioned that this understanding varies greatly company to company. While massive hedge funds might track employees’ moods and read texts and emails, smaller businesses rely on simpler indicators.

?

Turnover takeaways

?

Curious what FORE has discovered about turnover so far? Hochman shared a fascinating finding on utilization.

?

He’s found a relatively consistent U-shaped curve for it – meaning underutilized employees are likely to turnover, middle-utilized employees don’t leave very often (about the middle 40-70% of people), and over-utilized employees are at a very high risk of leaving, especially past the 90% utilization range.

?

FORE also studies the ‘why’ behind these metrics, so I asked for some common reasons people leave. According to Hochman’s analyses, over-utilized workers leave due to burnout, competitor risk, and upward mobility within the organization. Under-utilized people leave for a wide range of reasons, like team composition, communication, skills and dialogue, a lack of skills matching up to the job description, an inability to onboard effectively, and an inability to train effectively.

?

However, he caveated that turnover isn’t always a bad thing:

?

“You want a natural, organic turn in an organization. You just want to optimize it. You want to be able to forecast and predict around what that turnover is going to be. And so if you're losing some low-utilized employees because the job wasn't right for them, it's actually probably a good thing. You don't want to keep those guys anyway. But you just want to make sure you can forecast and optimize for it,” he said.

?

Addressing turnover with data

?

Let’s zero in on FORE’s solutions offering for a moment – how do they help users with retention issues beyond pointing out the problem?

?

A major gap Hochman identified in this market was tracking the efficacy of retention solutions. So, FORE now helps users understand how to best keep people around based on a worker’s individual needs:

?

“So not only do we predict whether or not employees are going to stay or leave in a business, we actually also created an algorithmic way to track the effectiveness of intervention solutions and then prescribe intervention solutions, particularly to the right type of person,” he said.

?

Instead of just tossing the retention playbook at people and overwhelming them, this offering helps employers identify interventions that actually work.

?

Actionable advice for improving data

?

To wrap up our conversation, I asked Hochman what someone can do to make their data more useful.

?

He recommends envisioning this process in three parts: data acquisition, data structuring and streamlining, and data analytics. The most important area to focus on, he said, is streamlining. Because often, you don’t need a sophisticated model for analytics – simply putting all data sources in one place can reveal major insights and clear breakthroughs. And, if you have too much data to process manually, look for AI tools to help with the organizing process.

?

?

Feeling inspired to get your customers’ data squared away? Hopefully Hochman’s insights have pointed you in the right direction. If you’re more interested in predictive workforce analytics, head to www.foreenterprise.com to learn more about Hochman’s work.