AI can help benefit leaders with the compensation process

Determining salaries is one of the most critical pieces to attracting and retaining talent, and technology may have the potential to improve the decision process significantly for benefit managers and employees.  

Over 70% of U.S. companies use AI in some capacity, including compensation decision making, according to business networking platform Connex Partners. While benefit managers may feel uncomfortable relying on automation for something as complex and sensitive as deciding salaries, it has the potential to drastically improve the process for them, as well as current and prospective employees. 

"Compensation professionals and HR leaders have been using data to figure out how to pay people fairly and accurately since compensation was invented," says Sara Hillenmeyer, the senior director of data science at Payscale. "AI can allow us to better understand all of the factors and features of payroll and make it easier to make good decisions with that data." 

Read more: Don't let PTO go to waste: An AI-powered approach to cashing out time

It's critical to demystify the process for employees, as 68% of employees feel underpaid even if their compensation is at or above market rates, according to a Payscale report. This perception plays a significant role in employee engagement and turnover. While workplace pay transparency and equity efforts have made progress in quelling some of those fears, the disconnect between the workforce and the compensation process persists: In fact, 65% of workers still express a poor perception of their current pay. 

AI can equip benefit leaders with the data analytics necessary to support their compensation decisions, enabling them to communicate those decisions fairly and accurately, Hillenmeyer says. 

Simplifying compensation data

A common example of the consequences of misinterpreted data is "data dominance," which is when organizations in cities with higher salary bands or with a higher budget skew the data and drown out the earnings from smaller organizations. This makes it hard for benefit leaders in every industry to truly understand how much they should be offering to remain competitive in their target demographic. Investing in an AI tool could improve the accuracy of pay disparity detection by up to 65%, according to findings from HR analytics platform Moka, and reduce pay calculation errors by 92%. 

Read more: Monitoring employee productivity with AI? Better benefits can restore trust

"I like to say that AI can either reduce bias at scale or scale bias," Hillenmeyer says. "And if you get that wrong, it's detrimental to the business and generally leads to not being able to attract or retain talent, which can really set an organization's business goals back significantly. It's an expensive mistake to make." 

If organizations are building and introducing AI internally, benefit leaders need to be setting responsible and specific guardrails to ensure that certain groups aren't discriminated against based on external data. And if they're partnering with a third-party company, they should be thorough with their selection, and ask as many questions as possible about where the AI's data is getting pulled from and how it could affect their workplace demographics. 

"I expect compensation, reporting and planning to become much more automated, much more streamlined and generally take less time in the future," Hillenmeyer says. "Being able to forecast out the budget of labor in a much more data-driven way will help businesses flourish and make long-term decisions and improvements around compensation."