Note: This is the second of three articles dealing with managing performance and compensation in the post-ratings organization. Monday’s article discussed the broad issues raised when ratings are eliminated. Wednesday, the final article discusses ratings distribution.
————
Often, when bonuses or long-term incentives are distributed, complaints about “budget bottlenecks” can be observed. In reality, these complaints are just a symptom. The actual problem — the root cause — goes back further to expectations that are awakened during the feedback process, whether reviews are conducted hourly, daily, weekly, monthly, or “classically” just once a year.
Here, two paradoxes can be seen in the current discussion:
- For inexplicable reasons, the sensible notion of increasing the frequency of feedback conversations and of better balancing backward and forward looking aspects in those conversations, is usually paired with the suggestion that employee ratings be abolished. There is no logical reason for this, as these are two separate and rather independent questions.
- The feasibility and effectiveness of isolating the feedback process from performance-based remuneration decisions, which are often discussed in this context, seems unreasonable given both are inherently all about employee performance.
Grade inflation and pay
So, what is actually at the heart of the matter? If all members of a team receive “above-target” feedback, and expectations of an above-average salary increase or bonus are awakened, frustration in the workforce is inevitable.
[clickToTweet tweet=”If everyone is rated ‘above average,’ how will they get above average raises from a fixed budget?” quote=”Ultimately, when a manager determines who gets how much of the pie, he or she will find that it’s just not possible to give each employee an above-average amount.”] This is, because in the vast majority of companies the size of the pie is actually determined before individual entitlements are calculated. It is often based on the sum of the contractual 100% bonus components of all team members, as a percentage increase in the sum of all employees’ salaries, or as an amount set at the discretion of the board.
Consequently, the relative distribution of the available budget must be oriented toward this average base line, regardless of the absolute size of the pie, that is, the budget.
Therefore, if a manager allocates too much “above-average” feedback, causing too many employees to have expectations of being “entitled” to an above-average amount, the mathematical equation – unfortunately – no longer holds true. The average of all the allocations cannot be more than the size of the pie divided by the number of employees, that is, the average.
So, how can you maintain an average across an entire team? Either you give everyone the “average” (which is 100% of his/her formal entitlement, e.g. the contractually agreed bonus) or every dollar allocated to an above-average employee must be taken from the allocation of another (below-average) employee.
Ultimately, when the allocation challenges and the mathematic axiom of weighted averages kick-in, even the most inflated performance reviews must come down to earth, and reality can no longer be avoided.
High-performing employees will be disappointed by a bonus or pay raise that is average at best, or lower-performing employees must be allocated a smaller share of the pie, and thus told the unpleasant truth that might have been sidestepped in earlier feedback conversations.
Calibration
Hence, calibration of feedback and performance reviews is pivotal to the whole process, be it as moderated management meetings or systematically enforced distribution curves. Only calibration can ensure an average evaluation result and, in turn, realistic employee expectations regarding any subsequent financial allocation processes. But, calibration demands discrete measures, categories, indicators, or ratings, as a unique currency in this process. It is just pointless trying to calibrate poetic performance descriptions.
Now, calibrated ratings are certainly far from perfect, but they are the best available mechanism to mathematically verify that realistic expectations have been raised during the evaluation and feedback process. It is literally impossible to employ a process of manual screening and case-by-case assessment in an organization of tens of thousands of people and thousands of managers.
Pay and bonus alternatives
Apart from providing unlimited budgets for bonuses or salary increases, there are not too many alternatives to avoid all these causalities and negative consequences:
We could completely abolish all compensation and career development models that are based upon individual performance and replace them with collective models that are linked exclusively to the company’s operating results. But, this would not provide a solution for all the talent management decisions that are strictly related to individual performance, for example promotion decisions.
Or, even more extreme, we could get rid of all differentiation and treat everyone the same in a culture of egalitarianism. But this principle would also hit a wall: It would quickly get pretty cramped in the boardroom – or there would be no board at all.
Or, as a third option, we could migrate to pay-scale models like those in the public sector which are largely free from individual performance criteria, but more seniority and tenure based.
In either case, companies would gradually descend into mediocrity and the best talents would leave to join organizations with an active performance culture and remuneration. The future of these companies would be sacrificed on the altar of all the current hype about performance ratings abolition.
Rating and ranking as the only model
Formally and ostensibly abolishing ratings does not change the simple fact that a company can’t give out more than it has available. In a performance culture context, some tough decisions have to be made in order to appreciate the different performance levels and contributions of the various members of a team. At the tail end of all those decisions – on “pay day” so to speak – managers must adhere to the available budget. Here, at last, differences between employees have to be made, and a rating and ranking process must take place, at least implicitly in the minds of managers.
If we concede that employees must be grouped, classified or rated in some way for all the reasons outlined above, then let’s call a spade a spade, and let’s do it deliberately and transparently. Trying to work around these realities behind managers’ closed doors using any hidden “shadow accounting” simply does not live up to modern people leadership standards.
What’s the alternative to ratings?
At the end of the day, the success of any performance management system requires competent managers who can realistically assess the capabilities and performance levels of their employees, and don’t shy away from difficult dialogs and subsequent people development tasks. The argument for abolishing ratings and instead introducing algorithms or bloomy conversations only diverts attention from the real issue.
Many companies have euphorically jumped on the bandwagon of rating elimination that has engulfed the entire HR world. Unfortunately, the concept had not been fully considered, and the consequences are now beginning to surface as these companies commence their first salary, bonus, and promotion cycles in the post-rating era. These companies are struggling severely to maintain a proper process while employees are starting to view the rating-free era critically.
Here’s one suggestion: How To Link Pay To Performance When You’ve Eliminated Reviews
Those who recommend abolishing ratings must also devise a plan for what should replace them in the subsequent processes where they used to play a role, for example, in the calibration of performance assessments and in the determination of bonus and salary increases. The complete division of these processes and the computer-generated algorithms proposed to supersede them, in the end, fall into the category of improvements made for the worse.