June 1, 2018

Did We Throw the Baby Out with the Bathwater on Rank & Yank?

Don't worry. This post is not a recommendation to bring back Rank & Yank carte blanche.

I am really excited about the paper my colleagues and I just published in the Journal of Consulting Psychology entitled The ROI of Rank and Yank in a Simulated Call Center Environment which suggests that R&Y may prove to be a useful approach for increasing organizational performance in certain contexts. 

Here is a link to the original research: http://psycnet.apa.org/record/2018-04372-001. 

I hope after reading this summary you walk away with a more nuanced view of this controversial topic.

Background

CEOs went student body left when Jack Welch started hyping Rank and Yank back in the 80s and 90s and many blindly implemented it in their own organizations.  I used to like to say that when Jack Welch was CEO of GE, if he had started wearing a red dress, every male CEO in America would have gone shopping at Talbots.

Then those same organizations, including GE, went student body right and abandoned R&Y as the theoretical and empirical data came back indicating that the unfair comparisons, subjective ratings, and churn were crushing morale, undermining teamwork, and not only not improving organization performance, actually hurting it.

Let me say, again, that my colleagues and I are not recommending bringing back Rank and Yank. It is, unequivocally, not a management approach that can be widely or indiscriminately applied.

But we wondered if we might have thrown the baby out with the bath water and whether there are some jobs for which Rank & Yank can deliver performance improvements without the deleterious downsides?

If there were an environment where R&Y could work, we think it is call centers and other "call center like” environments (house keepers in hotels, etc). 

Why? Because these environments are characterized by 1) a large number of people, doing the exact same job where an actual distribution of performance is possible, 2) worker performance that is relatively independent with little need for teamwork (the performance of a housekeeper on the 29th Floor has little impact on how a housekeeper on the 12th Floor performs), and 3) objective measures of performance exist (like batting averages, or foul shooting %s in sports, or first call resolution %, or compliance % in a call center) that obviate the need to stack rank workers based on subjective management ratings across jobs that are completely different.

But we wondered if we might have thrown the baby out with the bath water and whether there are some jobs for which Rank & Yank can deliver performance improvements without the deleterious downsides?

Research Questions 

A) If you implemented rank and yank in a simulated call center environment, would you observe an improvement in performance especially when you took in account the high voluntary turnover that call centers are plagued with? (Turnover in call centers eats all the gains that accrue from attempts to make each agent better because those agents quit and walk out the door with your improvement investment.)

B) And if yes, would there be significant ROI for that increased performance when the costs of the turnover due to rank and yank were factored in? 

Since a model is no better than its assumptions, here were our key parameters: 1) we assumed a hypothetical 1,000 agent call center, with agents making $31,000 per year, 2) From Hunter and Schmidt’s work in the early 1980’s we estimated the benefit of a one standard deviation shift in mean performance would be worth 40% of agents’ salary, 3) in the realistic call center simulation (high turnover), we assumed 30% voluntary agent turnover…the 10% R&Y turnover was on top of that, 4) finally, the costs for firing through the rank and yank program were estimated to be approximately 46% of an agent’s salary based on published estimates that are cited in the paper.

Results

C) The answer to A) is unequivocal. Rank and yank had dramatically positive effects on performance, but that effect was dampened considerably over time by the voluntary turnover. Call center turnover "eats" any investment in trying to improve performance agent by agent. This is consistent with other research I have published on how turnover eats "by agent" performance gains: https://www.isixsigma.com/operations/call-centers/futility-call-center-coaching/

D) The answer to B) is that the R&Y program had an immediate, positive ROI in year one of $469,295 (1.4% overall performance increase) and a cumulative ROI after year five of $7,562,620 (6.1% performance improvement). 

A +1% YOY performance improvement for five years from a organizational intervention is nothing to sneeze at.

Of course, “your returns may vary." If the cost parameters in your center are dramatically different, your results will be too. For example:

* If you pay your agents more than $31K, your gains could be higher. 

* If your costs to replace workers were higher than the $15K we modeled, your returns would be lower. 

* If your voluntary turnover is lower than 30% your returns would be much, much higher. 

* But if your voluntary turnover is so high that you are already having trouble replacing workers, a Rank and Yank program in that environment would do more harm than good.

So What?

Organizations are constantly looking for ways to lift productivity. A +1% YOY performance improvement for five years from a organizational intervention is nothing to sneeze at. Of course, if you have interventions that would deliver more improvements at lower costs and with less disruptions, by all means, start there. 

It is interesting that Jack Welch indicates he didn’t learn about Rank & Yank from another CEO or another company.  He learned it on the playground:

“When we were making a baseball team, the best players always got picked first, the fair players were put in the easy positions, usually second base or right field, and the least athletic ones had to watch from the sidelines. Everyone knew where he stood. The top kids wanted desperately to stay there, and got the reward of respect and the thrill of winning. The kids in the middle worked their tails off to get better, and sometimes they did, bringing up the quality of play for everyone. And the kids who couldn’t make the cut usually found other pursuits, sports and otherwise, that they enjoyed and excelled at.”  https://frrl.wordpress.com/2009/10/31/jack-welch-on-differentiation-or-making-winners-out-of-everyone/

And speaking of baseball, it was Michael Lewis' Moneyball, that taught the baseball world and then the entire sports world about the value of building teams based, not on the fact that a guy "looked like a baseball guy" as the scouting reports used to state, but based on reams of performance statistics. In Moneyball, some of these stats were brand new ways of looking at players that GMs had never considered before, like "% of at-bats where he didn't make an out" now known as On Base Percentage, and one of the single most important statistics in evaluating baseball players.

Now all teams in all sports use, not wordy scouting reports, not "looking the part," not coach assessments of being a "team guy," but performance data that is hoovered up from every minute of every game and used to stack rank players on all teams across the US, outside the US, major and minor leagues. The players know that this is the world they now live in. It is a fact of life and no one complains: produce at a higher level than those on your team and on other teams or when your contract is up you may be looking elsewhere.

Now it is true that it may only be a very small set of jobs have these objective performance measures and lots of people doing the same job to compare performance against. But if you find yourself leading in a situation that involves jobs like that, the argument here is that Rank & Yank should not be overlooked as a way to cost effectively boost performance.

More posts

Get In Touch

Add your email below and we'll reach out as soon as possible. We look forward to working together!
Thank you for your inquiry - we'll be in touch shortly!
Oops! Something went wrong while submitting the form.