Using Replacement Costs to Measure Turnover

On July 18, ERE.net featured “How to Really Calculate the Cost of Employee Turnover,” which highlighted a few key metrics that factor into the real cost of turnover. The opening statement stands out:

Employee turnover costs are often described with generic numbers such as “$X,000.00 per employee” or “X percent of annual salary.”

Turnover cost, specifically “X percent of annual salary” — which can also be translated to $$, is one of the most effective KPIs to use in achieving the real measure. They tell a much deeper story than the “generic” term implies, and they are much easier to use. In 2010, while doing research at Aberdeen Group, we found that most companies use replacement costs to measure the cost of turnover. After taking out some outliers in high-volume/low-skill environments and some very high-level C-suite and management consultants, the analysis showed on average that using 86 percent of starting salary is a very fair estimate of the cost of turnover. NOTE: One of the research notes where this finding was published can be found here (page 2).

Here are four reasons why this metric is effective and not as generic as one might think:

  1. It is easy to measure: Some companies measure it as percentage of salary of the replacement. Others measure it as percentage of salary of the terminated employee. Others take the midpoint of the range for the position. Regardless of how you choose to measure it, stick with one. Do not change it down the road, and when measuring trends, change them into dollar figures.
  1. KPIs are just that: The cost of turnover is a soft benefit for the company. Most HR folks will have a hard time selling a CFO on it as a hard benefit, and I doubt shareholders will buy into it. Even if they do buy in, KPIs by definition are indicators, lagging indicators at that. It really does not matter what the actual number is. What matters is the movement and trend line. If it goes down, that means we are doing the right things — we are recruiting better, we are developing better, we are rewarding well, we are finding the right fit, we are using the right assessment tools, etc. Retention and turnover — its inverse — are symptoms of how well we are doing all of these things.
  1. It tells the key elements of the story: Using percentage of starting salary automatically compensates for the difficulty of recruiting for tough positions. Higher-level skills require additional recruiting efforts and tend to deserve higher salaries.  Secondly, this metric can be built in a way that addresses internal hires. Say a high-skill level position is vacated and the starting salary is $120,000 per year, and the replacement is someone internal who used to make $100,000 per year. The best way to account for that is to assign zero turnover costs for the higher-level position and 86 percent of the $100,000 for the vacancy created by the internal position. Some might argue that just because an internal hire filled the position doesn’t mean the recruiting process wasn’t as arduous or lengthy as filling it with an external candidate. I agree with that, but ramp-up time is usually less, and time to productivity is much shorter, which will discount the replacement costs a little.
  1. It can be used as a metric for cost per hire: The main difference here (compared to turnover cost) is that it applies to new requisitions vs. filling existing vacancies. One of the most difficult things to quantify is cost per new hire. Part of the reason is finding all the costs (job advertising, software, the time spent by recruiters, and hiring managers, travel, etc.). It might not be as high as 86 percent of starting salary, but one can use a portion of that. I don’t think it’s fair to attribute ramp-up time to the effectiveness of our recruiters, but some recruiters are willing to take responsibility for turnover and they are willing to go with the full percentage. Similar to the cost of turnover, it also accounts for the difficulty of recruiting for the position. When a position is filled internally, it discounts some of the costs, which is when we want to ensure our HR team receives credit for making a great hire, developing them, and retaining them. If the position is closed out for that employee who took the open req, then HR deserves zero for the cost of hire, which erases the need to hire for that position and makes us look good. Driving down the cost per hire is a good thing.

NOTE: The alternative here is to use the costs we know and remain consistent, and those include FTE salary, benefits for recruiters, software subscriptions, job advertising, and travel during the recruiting process, fairs, outsourcing vendors (RPO), and search agency fees. The only problem here is it gets murkier because it does not account for the difficulty of hiring for the position, especially if you do not rely on search agencies whose fees are usually dependent on the role and salary.

Article Continues Below

So how do should we track this metric to account for turnover cost?

  • Get your senior leaders and those in the business to agree to the percentage of salary — feel free to use 86 percent for cost of turnover since the research is already there.
  • Add up the cost for the measurement period — one way to do that is to take the average of the mid-points for all the ranges of the vacancies that open up that period (not including requisitions for brand-spanking-new positions) and multiply by the 86 percent. We should only use the ones that open up so we do not double count. Another way is to use the salary of the termed employee who created the vacant position.
  • Measure it consistently every period; do not change the calculation method.

As for cost per hire, using the same percentage is OK as long as we account for new requisitions and give HR credit for internal hires.

For HR’s road to be more strategic is to align its efforts to the goals of the company. Reducing turnover costs falls under that alignment. While this is not a metric that our CFOs might boast on an earnings call, we should make sure we are executing on our recruiting and retention strategies. Our objective here is to keep a positive trend — in this case a descending one for turnover cost.

Jayson Saba is Ceridian’s Vice President of Market Research & Industry Relations. He leads Ceridian’s market research, competitive intelligence and industry analyst relations. Prior to joining Ceridian, Jayson was the lead analyst covering core HR, Workforce Management and Outsourcing at the HCM practice of Boston-based Aberdeen Group. During his tenure, he wrote and published over 100 research papers and benchmark reports.

Topics

3 Comments on “Using Replacement Costs to Measure Turnover

  1. Good point DK. Personally I am not aware of any more recent studies on this. Until there is one, it is OK to start with the .8x number and if you had to discount it a little or a lot to get the buy-in, that’s ok. The main goal should be less about the absolute number and more about decreasing it (%) every period. After all if things are going well then the number should go down. Just be consistent with the multiplier. Thanks for reading!!

  2. Would be curious to see how these numbers play out in the world of sales. Perhaps the biggest cost of turnover in sales forces is the customer’s perception of your company because of revolving door reps. Worse are the lost opportunity costs of a competitor winning business while the failing incumbent is PIP’d out the door. Finally, the time it takes to find and replace a new rep along with new hire ramp up time allows for another several months of a competitor having open field running in new and existing accounts. Probably easy to just say the cost of turnover regardless of function is enormous and warrants putting great rigor in your selection and recruiting process.

    1. This is a great point. I agree 100 pct that one if the hidden almost unquantifiable costs of turnover is the damage to the employer’s brand. Your point is spot on in close knit industries (eg HR technology) where incestuous hiring practices are prevalent, and not to mention that often Reps on the local level know each other very well and they interact a lot.

Leave a Comment

Your email address will not be published. Required fields are marked *