How to navigate the 3 challenges inherent in rebooting performance management

Most organizations have a performance management system that is some version of the following flow:
Peer feedback + Self Assessment → Manager feedback & rating → Annual % increase (and/or bonus) tied to rating.

According to the recent Bersin Deloitte High-Impact Performance Management Study this process is broadly despised (net promoter score of -60%) and pervasive (86% use ratings in some form).

As someone who has implemented, wrestled with, and rebooted this process in several organizations, I can tell you that any variant we design must solve for 3 things:

1. How do we facilitate the flow of timely feedback that people are actually receptive to?
2. How do we equip our managers to do the hard, often untaught, job of managing?
3. What are rewards and salaries based on?

Before we get to these questions, let’s unpack current performance management approaches, starting with the performance rating.

What does research tell us about the efficacy of performance ratings?

The simple answer is: human beings are not very good at assessing someone else’s skill-set, and ratings tell us more about the person giving the ratings than the one receiving them.

Job performance accounts for only 21% of the variance in performance ratings.

In 2000 Scullen, Mount and Goff examined the performance of a large number of employees. Their data set consisted of 2,142 managerial employees, each of whom received developmental ratings on 3 dimensions from 7 raters (2 bosses, 2 peers, and 2 subordinates, and the target themselves.) They found that 62% of the variance in the performance ratings were attributable to the individual rater’s 'peculiarities of perception' (a term used by the team at Deloitte who got rid of ratings in their review process).

Job performance itself accounted for only 21% of the variance in ratings (Scullen et al. 2000).

This is without the added pressure or calculus around rating consequences (compensation, promotion, job changes) added to the mix.

Here are some factors that might explain how this plays out:

1. Tweezing out individual performance is hard.
   Individual performance is often a combination of talent + grit + luck/context + and finally the system within which the person is operating (Groysberg et al. 2004).
   

2. When our mind needs to make a judgement about something hard, it somewhat automatically substitutes it for an easier problem – the answer to which is easier to assess.
   Behavioral economists call this phenomenon ‘Attribute Substitution’. A variety of studies illustrate this in the context of performance.
   
   Chia-jung Tsay conducted a now famous series of experiments called ‘Sight over Sound’. She took clips of the top 3 performers in music competitions like The Voice and American Idol and showed them to novices and experts. Participants in the study were given thin-slices of data with either just sound, just visual, or sound and visual. Here is what she found: “novices are able to approximate expert judgments, originally made after hours of live performances, with brief, silent video recordings” (Tsay 2013).
   
   In short, experts who had years of training and credentials to assess musical performance were essentially using visual cues as a substitute to assess talent. This phenomenon has been replicated in other areas: everything from picking which political candidate will win, to chief executive officers’ appearances predicting company profits.
   

3. Extraneous error is real.
   One study found that favorable decisions in judicial rulings are tied to how long after a snack judges are making a ruling (Danziger et al. 2011).
   

4. Finally, calibrations don’t help.
   Studies on how group calibrations affect assessments of performance (Palmer and Loveland 2008) indicate that:
   

• Group discussions reduce the accuracy of the assessment.
  
• They polarize existing impressions of good and bad performers: in other words, an above average performer comes out looking like a rockstar, and a below average performer moves to grade F.

In addition, calibrations add factors outside of performance to the final assessment: managers trying to impress leaders; over-emphasis on one-off leader perceptions; and a manager’s ability to communicate, influence, and advocate for their employees.

So why haven’t we ditched performance ratings and assessments?

Daniel Kahneman, the man who pioneered the study of cognitive biases recounts how awareness has little impact on limiting the effects of these cognitive biases. This is not a lack of knowledge, it is in our very wiring:

“The statistical evidence of our failure should have shaken our confidence in our judgments of particular candidates, but it did not. It should also have caused us to moderate our predictions, but it did not.

We knew as a general fact that our predictions were little better than random guesses, but we continued to feel and act as if each particular prediction was valid. I was reminded of visual illusions, which remain compelling even when you know that what you see is false. I was so struck by the analogy that I coined a term for our experience: the illusion of validity.”

Daniel Kahneman
(Psychologist, Behavioral Economist, Nobel Laureate)

Right here you have the first reason why it's hard to get rid of ratings and calibrations: while it is easy to convince us that others are terrible at assessing performance, the confidence we have in our own ability to do so is infallible.

This brings us to the second part of the performance management equation: compensation and if/how it is tied to performance ratings.

What does research tell us about tying annual compensation increases to performance ratings?

The question we are asking is not if compensation is an important part of the employee value-proposition. It is. Competitive market compensation and the ability to realize progressive compensation gains within the same organization are important drivers for attracting and retaining employees. Setting absolute compensation for jobs and people is tied to the overall compensation philosophy and strategy an organization takes.

The question we are asking here is: Are merit based compensation increases (and by default differentiation) effective?

There are two fundamental assumptions underlying current pay-for-performance mechanisms:

• Human beings are rational economic actors.
  
• Extrinsic motivators, like higher pay increases and monetary incentives, are effective at driving behavior changes.
  

Both these assumptions have been widely debunked by research.

“Most merit-pay systems share two attributes: they absorb vast amounts of management time and make everybody unhappy.”

Jeffrey Pfeffer, Six Dangerous Myths About Pay: Harvard Business Review

Merit based compensation differentiation quite literally causes employee distress.

Loss aversion is by now a well established behavioral economics principle: the distress we feel at losing a $100 is more than the satisfaction we get from gaining $100.

Now factor in two things:

• Most people consider themselves above average. In one study more than 40% put themselves in the top 10% and only 2% put themselves below the 50th percentile (Meyer 1980).
• Happiness from monetary increases is momentary. People quickly return to their set-point of happiness shortly after a gain (Mancini et al. 2011).
  

So most of your employees are heading into performance evaluations expecting high increments. When differentiation rolls out, a majority experience it as a loss and feel distressed, and a small minority feel happiness that doesn’t last and has no long term impact on their motivation. While well placed in their intention to reward top performers and motivate others to meet high standards, experts in the field are nearly unanimous that this approach is not a good return on investment.

People work for money, but they are more motivated by purpose, meaning, and belonging.

Daniel Pink’s now famous book Drive: The Surprising Truth About What Motivates Us makes the empirical case that beyond a certain economic threshold money does not motivate performance. He popularized the widespread understanding of purpose, mastery, and autonomy as fundamental intrinsic drivers. Alan Colquitt, in his book Next Generation Performance Management, points out that Edward Deci’s original research also included ‘belonging’ in this list.

Here is a summary of the key intrinsic drivers:

• Mastery: people naturally strive to acquire, master, and excel at skills relevant to their area of interest and work. They need a level of challenge in their roles, and perform better when their work aligns with their strengths.
  
• Relatedness: a sense of belonging, connectedness, and the ability to bring their whole selves to work contributes to employee well-being.
  
• Autonomy: the ability to have control over the choices and decisions that drive one’s work.
  
• Purpose: the experience of their work contributing to something larger than themselves; combined with an overall alignment between management decisions and organizational purpose.
  

I want to emphasize that meeting that basic economic threshold is important and ties back to the first compensation issue we identified above: providing meaningful compensation growth within a role.

But when it comes to the onerous annual comp increases, management time is better spent:

1. Creating jobs that are interesting, challenging, and meaningful; and
2. Creating cultures that foster trust, belonging, and align to a higher purpose.
   

Extrinsic rewards, and how they are administered, can have a negative effect on intrinsic motivation.

Here’s how leading researchers in the field summarize the effect of extrinsic rewards on intrinsic motivation:

“Rewards given for low-interest tasks enhance free-choice intrinsic motivation. On high-interest tasks, verbal rewards produce positive effects on free-choice motivation and self-reported task interest. Negative effects are found on high-interest tasks when the rewards are tangible, expected (offered beforehand), and loosely tied to level of performance.”
(Cameron et al. 2001)

Some experts make the case that the hyper-focus on monetary rewards and compensation, often a natural outcome of merit-pay systems, can harm collaboration and inhibit creativity and innovation (Colquitt 2017).

The reality is that monetary incentives do drive behavior but often with counterproductive effects: case in point is the recent sales account fraud at Wells-Fargo. The amount of constant evaluation, tinkering, and perfecting it would take to administer monetary incentives in a way that enhances rather than takes away from healthy work and motivation are unaffordable for many small organizations (and big ones apparently).

Confronting mindsets: free-riding, and a culture of mediocrity

Despite overwhelming evidence, it can be hard for organizations to move away from key elements of traditional performance management: performance evaluations, differentiated merit based pay, and calibrations.

The two biggest fears that get in the way of changing performance management systems are:

1. The fear that people who lack motivation and desire to contribute (shorthand: lazy) will take advantage of the system.

Alan Colquitt, previously mentioned under intrinsic drivers for his book: Next Generation Performance Management, summarizes several recent behavioral economic studies on this topic:
“There have been several reviews of this research concluding free riding is not a significant problem. These effects are much less of a problem in normal work settings with more stable, permanent groups and interdependent work activities and where individual effort is publicly visible and peers can influence behavior.”

To learn more you can read a study by Gerald Marwell and Ruth Ames, of the University of Wisconsin, aptly titled research paper called Economists Free Ride, Does Anyone Else? (Marwell and Ames 1981) that debunks ‘free riding’ as more of a conceptual exercise rather than an empirical reality.

To be clear: if the organization has employees that are actively disengaged, managers need to find a way to address that directly and quickly; and not by designing systems that punish everyone for this behavior.

2. The fear that it will result in a culture of mediocrity.

Here is where we will diverge from empirical evidence. ‘Our performance evaluations are why we are such a driven, innovative company’ – said no one ever.

The reality is that many cultures thrive despite their performance evaluation processes, not because of them, and the practices that foster an innovative culture are much harder to put in place and sustain than simply focusing on increases tied to ratings.

So what works?

If you are reading this, you have likely read the many articles advising you to ditch ratings and move to lighter learning-focused dialogues.

While that might be true, most organizations looking to revamp performance management come to realize that they need to replace one broken process (a performance management pandora’s box if you will) with several practices that work in tandem to support their organizational performance goals.

Depending on the culture, context, and strategy of your organization the practices you design will be unique to you. But they do need to solve for the 3 challenges we outlined at the start.

Challenge #1: How do we facilitate the flow of timely feedback that people are actually receptive to?

Let me just start by saying this: we are fighting human nature with this one. It cannot be solved by a process alone, and needs to be intentionally woven into the very cultural fiber of an organization.

When designing a feedback practice consider two equally valid and contradictory facts:

a. A comparative, ratings driven process, inhibits feedback and its effectiveness.

Giving and receiving honest feedback is hard. Period. Tying it to ratings and compensation raises the stakes. Now I am not just telling you what you could do better, I am linking it to your identity and status in the organization.

It's no surprise then that most managers would like to stay in the ‘safe-zone’ when it comes to giving ratings, sticking to the 3’s and 4’s on a 5-point rating scale.

“Though most of us no longer have to fend off predators, our brains are still exquisitely attuned to threats — both physical and social. It’s a vestige of how survival has largely depended on appeasing group members. Among our ancestors, eviction from the group led to a dangerous, isolated existence in the wild…

Feedback conversations, as they exist today, activate this social threat response. In West and Thorson’s study, participants’ heart rates jumped as much as 50 percent during feedback conversations. (Equivalent spikes have been found during some of the most anxiety-producing tasks, such as public speaking.)”

Using Neuroscience to Make Feedback Work and Feel Better by by David Rock, Beth Jones, and Chris Weller

b. Simply removing ratings will actually diminish the frequency and quality of feedback.

Given how uncomfortable the process of giving and receiving feedback is, and how ill equipped most people are at the art of it, taking away the forcing mechanism will likely result in reduced feedback.

In fact, according to a 2016 study by CEB (now Gartner) employee performance can drop when ratings are removed because managers struggle to effectively manage employees without them (which is our challenge #2).

While the solutions will differ in each organization, an effective feedback practice should:

1. Be multi-directional.
2. Start a habit of people asking for feedback, input, and perspective (instead of just getting it).
3. Include, at least at the start, a forcing function that helps people build a feedback habit.
   

Challenge #2: How do we equip our managers to do the hard, often untaught, job of managing?

“What looks like laziness is often exhaustion.”
‍
Switch by Dan and Chip Heath

While it's easy to beat on managers here, the reality is that most of us (managers included) live in a world of constant overwhelm. Any approach we take to helping our managers – master and deliver on the challenge of providing clarity, alignment, empathy, tough-love, connection, and growth in the context of rapid change – needs to take into account the exhaustion they often carry with them.

I would recommend three levels of interventions to play with here:

1. Learning solutions that focus on raising their self-awareness, emotional intelligence, and help them acquire and match management styles to situations.
2. Giving people the permission to narrow focus and move from frenetic activity to intentional progress.
3. And finally talent solutions that help managers with the fundamentals of providing clarity, connection, and career growth.

Challenge #3: What are rewards and salaries based on?

I often like to say ‘compensation is what is wrong with performance management’. Organizations live with painful performance review processes for this simple reason: we need a way to give raises and make people decisions and ratings help us do that.

Except, ratings are idiosyncratic to the rater and turns out rating-based compensation decisions can be demoralizing to a majority of your organization.

Researchers and practitioners tend to agree that new solutions need to:

• Feel fair and adequate.
• Focus on the team and not just the individual.
• Provide a sense of compensation progression to an individual as they grow within the organization.

But despite there being evidence to show that using money as the primary lever of reward and motivation can backfire, there are strong disagreements on the use of differentiated compensation.

At the highest level, we see organizations using one or more of the following approaches as they tackle how to decouple compensation from feedback and learning:

• Job based salaries and team based bonuses/reward tied to differentiated outcomes.
• Moving away from annual adjustments and using target salaries (based on market and transparent internal considerations) to evaluate and adjust salaries as needed.
• Providing across-the-board increases that provide for cost of living and are based on organizational budgets.
• Empowering managers to provide spot bonuses with allocated pools, clear guidance, and back-end-analytics that ensure decisions are being made equitably.
• Colleague and client driven rewards that are meaningful, personalized, and gratitude focused.

More than any other practice you design, the compensation approach you take depends on the values and beliefs of your organization. Your leaders need to truly buy into the approach for it to work.

This is perhaps why many organizations and HR professionals consider changing their performance management process but settle for iterative changes instead of a sweeping overhaul: because it can feel like replacing an old rundown car with a fleet of new luxury cars with no extra cash.

I am here to tell you that we have successfully helped our clients replace one broken performance review process with practices that facilitate feedback, recognition and rewards, and create more transparent and equitable compensation approaches. This is doable.