Negotiation
Rating scales, distributions and the arithmetic behind your score
Where ratings drive pay, the shape of the distribution decides more than the description of each rating does.
This is less a set of instructions about performance ratings than an argument, and it is worth saying so at the start.
The argument in brief
- Guided or forced distributions limit how many top ratings can be awarded.
- A pay matrix converts rating and band position into a percentage.
- Position in the band can affect an increase as much as the rating does.
The distribution is the constraint
Many rating systems expect or require a rough spread across the scale, so a limited number of top ratings exist in any group. That means an excellent year can still produce a middling rating if the comparison group had several excellent years. Employers vary in how rigidly they apply this, and some have abandoned forced distributions entirely.
Asking whether a guided distribution applies tells you what kind of contest you are in.
A matrix does the conversion
Where a matrix exists, the increase is a function of two inputs: your rating and where you sit within your pay band. Someone rated highly but sitting near the band maximum can receive a smaller percentage than someone rated adequately near the minimum.
This is deliberate, on the logic that the band represents the full value of the role. It also means the rating alone does not predict the money.
Middle ratings are the design, not a criticism
Most scales are built so that the middle rating means the job is being done fully and well, and the top rating means something exceptional and rare. People frequently read the middle as a warning, which is a communication failure rather than an accurate reading.
Where it helps most, asking what specifically distinguishes the top rating in practice, with examples, is the way to find out what is actually being measured. If nobody can describe it concretely, that is informative in itself.
Timing within the year affects the rating
Recent work is weighted more heavily than earlier work simply because it is easier to recall. A strong first quarter can be forgotten by the time ratings are set, unless it is in writing somewhere the manager will look. Keeping a short running record and sending a summary before the rating window counteracts this directly.
It is a well-documented feature of how people assess periods of time rather than a failing of any particular manager.
New starters and leavers are treated differently
Partial-year employees are often rated on a shorter period or excluded from the distribution, and increases may be prorated. A joiner arriving late in the cycle can receive a reduced or nil increase for reasons unconnected to performance. Asking at offer stage how the first review will work avoids an unpleasant surprise.
Where this is the case, negotiating the starting salary matters more, because the first increase may be small.
Where ratings go wrong
Systems that rate people against each other are sensitive to who is in the group, who is visible and who has an advocate. Where you believe a rating reflects a protected characteristic rather than the work, that is a legal question rather than a negotiation one. Rules, remedies and time limits differ entirely by country, and your labour authority or union is the right first stop.
Where it helps most, contemporaneous notes and the written record are what make any such case possible.
The takeaway
Ask whether a distribution and a pay matrix are in use. Both decide more than the rating description does.
Pick the one that costs you least, and let the rest wait.
Questions readers ask
Why did a good rating produce a small increase?
Because most matrices combine rating with position in the pay band. Sitting high in the band usually reduces the percentage awarded.
Are forced distributions still common?
They vary widely and some employers have dropped them. Ask whether a guided distribution applies, since it changes what a rating means.
