Many years ago when I earned one of my professional qualifications I completed a project on performance appraisal systems. At that time, in the organisation I served performance appraisal had three clear purposes:
- Formal, recorded, quality communication between a manager and one of their direct reports about their performance, development and opportunities.
- Quality information for the Training department to help prioritise development activities across the organisation.
- Agreement of goals and targets for the coming year.
Overall, the performance appraisal system contributed to continuous improvement in performance and business results.
The emphasis – back then – was on the quality of the discussion between the employee and his or her manager and the record that enabled the Training department to respond to the needs of the business. Quality. Quality. Quality.
All discussion related to pay, bonuses and promotions was done at a different time of year.
Some years later performance-related pay was introduced to the organisation. Now we had to give numerical values to performance. Goals and targets had to be expressed in a particular way. Performance appraisals took on a new tone. The focus was on the performance rating and everything else faded in to the background. The quality of records of appraisals dropped dramatically.
At the time I thought that maybe the reason why this new approach seemed less useful must be that we’d failed to implement it in the right way. My career moved on and I didn’t revisit the subject for some time.
The next time I came up close with an organisation’s performance management system I realised that common practice (not necessarily, in my opinion, best practice) had changed dramatically. Most of my client organisations operated a performance management system that relied on a performance rating, given once per year, upon which an employee’s pay, bonus and opportunities depended.
Usually, when the subject arose, it was because people were complaining to me, an outsider, about the unfairness of the system.
On asking questions I discovered several aspects of common practice that continue to amaze me. I simply can’t see why anyone would think these are good ways to foster high performance. Or indeed, any kind of performance.
- The Five-Point Scale
Most rating systems use a five-point scale. A score of 5 indicates a superstar, a score of 1 indicates someone is not delivering. In reality, most managers will want to give scores of 3, 4 or 5. After all, if I score a member of my team as a 1 or 2, it’s a bit like saying I’m not doing a good job of managing that person, isn’t it?
- The Bell Curve
To counteract this tendency to score everyone average or above, lots of organisations have adopted a system of forced ranking. Each manager MUST use all five scores. Pressed as to why, the owners of the system will mutter about ‘the Bell curve’ and ‘normal distribution’. This is usually based on incomplete understanding and lack of consideration of the consequences of their actions.
Firstly, to insist that all managers must use all five ratings, is to presuppose that it’s impossible for every person in a team to be performing brilliantly. If, as an organisation, you want everyone to perform brilliantly, doesn’t it seem somewhat counter-productive to work on the assumption that the thing you want is impossible?
After all, the first rule of goal-setting is that you have to believe it’s possible to get your goal!
Yes, we know that if you were to plot (using some completely fair measurement!) the performance of everyone in the organisation on a chart, it would be likely to show a normal distribution – a Bell Curve. But if you looked at the distribution of individual teams you’d be more likely to find them in clusters. Why? Because one of the most significant factors in the performance of an individual is their boss.
So, to get good performance from all your employees, it makes sense that all the managers must be motivated to do a good job too.
How could forced ranking motivate a manager to get the best out of every person in their team? Knowing that even if everyone did a fantastic job, some of them would have to be given a low performance rating?
- Peer Review
Recognising that the forced ranking system had limitations, the experts than introduced a process of peer review. The principle being that managers may not be totally objective about their own people and that some employees may have taken on more challenging goals than others, so let’s get together all the managers at one level and compare notes. This group of managers can then agree who are the real star performers and who are the low performers – not exceeding the quotas for each rank, of course.
The reality of peer review as I’ve seen it is painful for all concerned. In large organisations most managers have a big enough challenge keeping up to date with the activities of their own team members, without having to know something about everyone else. So the eventual ratings are determined by each manager’s ability to ‘sell’ the achievements of his people to the other managers, all of whom are determined to ‘get a good deal’ for their own people.
If you think about it logically, if you can’t be sure that managers will be objective or fair about performance ratings for their own team, why would you expect them to be able to be objective and fair about ratings for an even bigger group that includes their own team?
So what are the alternatives?
My view of this is quite simple. Let the managers manage.
Why not have a scale that assumes everyone is doing at least an adequate job. If they weren’t you’d already be doing something about it, wouldn’t you?
If, as an organisation you have a budget allocation for salaries, salary increases and bonuses, why not refer that allocation to heads of departments, who can then allocate it to senior managers, who pass on a share to each of their managers and so on.
Decisions about how that money is distributed between individual members of a team would then be a joint decision between the line manager, their manager and a representative of HR.
Would it be fair? Maybe not entirely, but as far as I can tell the present systems used in most organisations aren’t fair. Would it motivate the managers? Possibly, and at least it wouldn’t de-motivate them the way a lot of systems do. There is little more disheartening for a manager than to have to tell an employee they have received a performance rating lower than they deserve and hence a smaller bonus. Or worse still, to have to invent a justification for a low performance rating.
The way the pay increases and bonuses were distributed would, ultimately, reflect the leadership styles of the managers concerned. Since people have a choice to work for a particular manager or not, it’s likely that they would feel comfortable with their own manager’s way of working it out.
The way I see this, rating people on a five point scale, demanding that all managers use all ratings and then trying to achieve a balance across a department by having the managers fight it out is leaving out of the equation a very important factor. Personal judgement.
If you don’t trust your managers to know their own people and judge who deserves what in terms of pay, bonus and opportunities, then no performance management system can plug that gap. And if you do trust your managers to manage, why not let them get on with it?