Stop! Are you ready to remove ratings on your journey to Agile Performance Management?

On your journey to agile performance management you will reach a crossing. This is the point to decide if you will continue to use a rating system for employees or if you will start to manage without ratings.

Expert advice is mixed. The NeuroLeadership Institute suggest that ratings are detrimental to low threat environments and that they drive fixed mindset. Many of the organisations they collaborate with have dropped ratings, however, some of their showcase clients like Time Warner Cable still retain them. In support of ratings, CEB use data to suggest that less than 5% of managers are able to effectively manage employees without ratings, and company performance drops by around 10% when performance ratings are removed.

We believe the challenge of removing performance ratings comes if they are removed prematurely, without first substituting in something better.

In this leg of our journey to agile performance management we look at what you need to do to cross over to a future without performance ratings.

Why do we rate people anyway?

Traditionally, the performance review process meets not just one, but several key purposes.

Performance reviews are commonly used to evaluate bonus eligibility, to determine salary increases, to assess readiness for promotion or lateral job changes, to manage out poor performers, to design individual development plans, to provide input to longer term workforce planning, as a source of HR Analytics, plus a host of other ancillary people management processes.

Decisions are much easier to manage if these processes are supported by numbers rather than words.

Numbers make it far easier to normalise, graph and superimpose some kind of forced distribution (such as a bell curve) over a large workforce. They simplify collecting and communicating the results of performance management processes at scale.

Operationally, ratings have been commonplace as a convenient way to assign people into buckets, and then proportionately allocate limited resources (money, training, promotions, etc) from the most 'deserving' bucket to the least.

So what's the issue?


Much of the problem with ratings stem from their subjectiveness. Subjectivity makes them prone to bias.

Recency bias, spillover bias and halo effect are key examples where ratings are often based on a partial rather than a full account of the contribution during the current review period.

Consider there are roughly 160 other cognitive biases that may influence the rater and it's no surprise the evidence shows ratings reveal more about the person giving them, than the person being rated. In fact ratings are a such a magnet for bias, the collective result is called the idiosyncratic rater effect.

Attempts to counterbalance rater bias tend to make the situation worse not better. To guide raters, review criteria are typically expanded to provide some calibration statements.

Responding to calibrated assessments like these now carries a learning curve for the rater. More cognitive load, and not necessarily any greater reliability. Any improved accuracy also depends heavily on appraisal design to describe behaviours in a way that leaves no doubt which score should apply. This is very difficult to do, and hence is usually done badly.

As if issues with individual ratings were not challenging enough, at the grouped level, forced distribution compounds the problem. 

Forced distributions were introduced to overcome central tendency (where almost all people are rating in the middle of the scale). The approach requires manipulation of the system to make sure the performance review process ends with equal numbers of poor and high performers, and the rest are in the middle. In such a system, ratings are adjusted until they fit the thresholds, regardless of how this compares with reality.

All of this bias means that decisions made using ratings are likely to be driven from faulty data.

Bias is the Achilles heel of ratings and the root cause of employees feeling unfairly treated at appraisal time.


Being ranked against others is psychologically threatening. It compromises our need for Certainty, Options, Reputation and Equity.

Being ranked as 'not the best' is unlikely to motivate improvement. It actually fuels a fixed mindset that our abilities are limited, our rating is unlikely to change, and it's beyond our control to change.

Where reputation is concerned, people carry the reputation of 'not the best' until at least the next review, regardless of how their current performance level has lifted.

The hard work to improve capabilities may not be recognised through ratings until too late, or possibly not at all. This means most people have very little certainty of what they need to do to reach a top ranking, and managers can't commit to that result in the future because ranking is comparative to others at that time.

Interestingly, even top ranking employees are seldom happy with their ranking. In some cases they may feel they've got a target on their back as a person to bring down, where they'd rather be a role model for others to learn from. In other cases they may feel there's really no point improving more, as there's no further status to be gained and they are already next in line for promotion.

The workplace effects of these threats show up as low energy, reluctance to take risks, less collaboration, and suppression of information to still look good when things are going bad.

Backward Looking

Since we cannot change the past, having someone rate it and place a high importance on this will once again trigger a threat response.

In this context, the threat response to a low rating is to justify and defend the past (to preserve reputation) rather than to acknowledge it. For a high rating the response might be to accept the bar is set for what's good enough (to maintain certainty) rather than explore what's actually possible.

What we really want to do is learn from the past and continually improve.

Learning requires both the reviewer and the person under review to reflect and to collaborate on ways to improve. This isn't the main focus when both parties are on the spot to defend their relative stance on rating.

High Performance is also deflated by rating the past. When examples of high performance are mixed in with improvement needs, the brain is still searching for resolution of the threat of needing to improve rather than enjoying the reward response of achievement, praise and recognition. It can't do both at the same time and threat always beats reward.

What could be better?

Fortunately there are several proven techniques to allow for the removal of ratings.

  1.  Separate out performance review from compensation

This is not to say don't pay for performance. It is simply saying keep both processes separate and feed individual and team achievements into compensation decisions.

This becomes more about pay for contribution than pay for performance. Bersin by Deloitte describe it as "rewarding employees who have contributed value to the organization beyond their performance objectives. For example, an organization might reward an employee who landed a big account, implemented a cost-saving measure, prevented a potential loss of business or developed intellectual property."

 2. Acknowledge rather than fight bias by asking questions about the reviewer's opinion and not the subject of the review's competencies

This article from the Washington Post explains how Deloitte are turning the conversation from talking about the ratings to talking about their people by asking managers 4 simple questions about their people

  •  Given what I know of this person’s performance, and if it were my money, I would award this person the highest possible compensation increase and bonus. 
  • Given what I know of this person’s performance, I would always want him or her on my team.
  • This person is at risk for low performance.
  • This person is ready for promotion today.

3. Collect feedback throughout the year, and from multiple sources

Having a stream of feedback that covers the breadth of a review period creates a much clearer picture of actual performance and achievement than a single point in time assessment even if this periodic feedback is still consolidated into an end of year review.

Enabling feedback from multiple contributors (co-workers and customers) in addition to the manager adds to the authenticity

Make sure the process is brain savvy so that feedback can be candid without posing a threat to the person giving it or receiving it.

Capture this as narrative to get as much context for the recipient as possible to maximise improvement potential.

 4. Convert qualitative data to quantitative data (stories to numbers)

Many organisations fail to appreciate that qualitative data (narrative) can be converted to quantitative data relatively easily.

Tools like sentiment analysis will score text to identify the positive sentiment versus the negative and identify the intent of feedback as a number!

This is a powerful way to capture both the story and yet still be able to produce graphs and charts to summarise and communicate the headlines.

5. Hold on to one last rating

With the right preparation organisations have been able to remove ratings from their performance management process all together. However, many have held on to one key rating that they continue to apply. That is to assess if the employee is a cultural fit or not, and it's a yes/no answer which in most cases is transparent to the employee.

An example is Juniper Networks where you are a J-Player or a Non-J-Player. 

If you are a Non-J-Player it's because you don’t seem to believe what those around you believe and what the company believes. Whilst not forced to, and in fact being assisted to fit in, 80% of Non-J-Players ultimately leave knowing that they cannot succeed with misaligned beliefs.

Executing well on this approach requires that the belief system for success is well understood of course.

Getting ready to cross

You will be ready to cross over to a performance management culture that is not reliant on ratings if you separate out performance conversations from pay, increase the frequency with which feedback is shared and train your people (not just your managers) to have candid conversations in the spirit of growth mindset and continuous individual improvement.

To measure performance, develop a strong understanding of the beliefs that define success in your organisation and teams, and measure individuals against their personal contribution and their fit with those beliefs.

If you have these characteristics in place, you are ready to cross. If not, the team at Pay Compliment are able to take your hand and get you safely to the other side of the ratings/ no ratings dilemma.

Pingbacks and trackbacks (2)+

Add comment