By: Jeff Higgins, CEO & Founder of HCMI
To find ROI with HR Analytics, companies often try one of the following analyses:
Statistical correlation analysis
While these types of analyses do provide insights into the workforce, rarely do they deliver the actionable interventions, ROI and productivity improvement that management is hoping for. Essentially these analysis techniques are necessary but insufficient for success in analytics.
For example, you may find that the average turnover rate has risen for the past six months and that it correlates with the recent headcount growth. But does it mean that the quality of hires has gone down recently, or is it because there aren’t enough opportunities for internal promotion and mobility? If recent hires are indeed the culprit, then which sources of hires are responsible for the majority of turnover? Which recruiting sources provide better quality of new hire and therefore you should hire more from?
Most business leaders understand that human capital is important for their success. They also get that hiring, training and retaining high performers is key to a productive workforce. They don’t know which courses of action would help generate the ROI or results they’re hoping for. This information historically has not been available to management. When you can’t see the impact or ROI, it’s then an easy place to go to next and start to cut headcount or freeze hiring when the budget is tight.
The key to impactful HR analytics is to integrate HR and other business data sources such as operational data, financial data or even better customer related data, to obtain a better picture of what drives talent performance. By sourcing HR data from various talent management areas such are recruiting, learning, core HR and more, connected and integrated with finance and operations data, analytics teams can see how HR interventions impact the business’s bottom line and key Operations KPIs.
(NOTE: It’s also just as equally important to focus your analysis on questions that are relevant and important for management. Download the Top 25 Key Human Capital Questions to Boost Your Workforce Analytics document)
Why Bother Integrating Multiple Data Sources
Why bother integrating multiple data sources? There is a lot behind this question but one thing we know for sure is that there is no one magic metric that can tell the whole picture. Let’s consider this true story:
Company A is a regional financial services company with its operations spanning across the western coast of the United States. Having done their homework, resulting in a lot of their senior managers being hired via employee referral into the company, company A believed that employee referral was a great source of hires and decided to offer incentive bonuses to the referrers when a new hire employee referral took place. After running the program for a year, management saw a lot more employee referral activity, but didn’t know how or what that translated into for the company. In fact, employee turnover was increasing even as they hired a larger share of employees via employee referral. Ultimately, they wanted to know if their money was being well spent.
Given the situation, the Chief Human Resources Officer (CHRO) asked both the Head of Recruiting and the Head of Performance Management to examine new hire turnover and performance data. Each team went on their separate way to collect whatever data they needed, set up a database and built their own statistical models. Everything went as planned but there was one thing they did not expect - the results!
The Recruiting team found that employee referral was the best hiring sources in terms of percentage of high performers and cultural fit
The Performance Management team found that employee referral had the highest average overall turnover rate
With this result, both teams couldn’t help scratching their heads and wondering whether or not referral was their best or worst hiring source. Do you think referral is the best or worst hiring source for this company?
This is when HCMI was brought in to conduct a detailed analysis on their data. This time, all the available data from model 1 and model 2 were combined to verify the previous conclusions and figure out whether the company should continue their referral incentive bonus program. The results we found surprised everyone.
It turned out that both conclusions were correct! Referrals were both the best and the worst sources of hire for Company A. The key deciding factor was the employee giving the referral or aka the referrer!
When the referrer was a high performer, it was three times more likely that the employee referral would turn out to be a high performer with high potential as well. Vice versa, the referrals made by below-average performers were more likely to be unsuccessful hires. The reason was that the employees giving the referral were more interested in the bonus incentive and brought more referees in, even though they weren’t a good match.
While the program had been implemented with good intention, the company didn’t get the outcomes that they were looking for. The key deciding factor was the employee giving the referral or aka the referrer!
With the results in mind, the HR team implemented a few interventions to discourage system abuse and motivate high performers to refer more often.
Here are the steps they took:
Target high performers for referrals – While it would be difficult to limit who can refer, it’s much easier to target known high performers and encourage them to refer more often. This was exactly what they did.
Delay giving the referral incentive – They delayed giving the bonus incentive right after a new referee was hired, waiting from 90 days to one year.
Add key milestones before referral incentive is given – Nothing is worse than hiring the wrong talent. Implementing “quality control” mechanisms helped the organization avoid making wrong hires and lowered the first 90-day and 1-year turnover metrics.
In conclusion, either insight would still technically be correct and yet was the wrong thing to do for the business! Without integrating all the right data sources, it wouldn’t be hard to see how management would make evidence-based decisions and still get it wrong. The key to all HR Analytics success is to look at data holistically, by integrating all the data to provide a full picture into the talent management lifecycle.
More Workforce Analytics examples can be found here: