Internal equity: what to check before every offer

Market data tells you what an outsider costs. Internal equity tells you what the offer will do to the people already doing the job.

Internal equity is the check most often skipped and most expensive to skip. Market benchmarking answers a question about the outside world. Internal equity answers a question about your own building: if this person joins at this number, what does that mean for everyone already doing comparable work?

The skip is understandable. The market check is easy — pull the band, look at the percentile. The equity check requires knowing who the comparable people are, what they are paid, why they are paid that, and whether the differences are explainable. That is real work, and under offer-deadline pressure it is the first thing to go.

Defining the comparison set

Most equity analysis goes wrong at the first step: choosing who counts as comparable. Job title is a weak proxy. Two people with the same title can have materially different scope, and two people with different titles can be doing effectively the same job.

• Start with level and job family, not title.
• Adjust for scope: team size, budget ownership, breadth of surface area, or whatever your levelling framework actually measures.
• Account for location if your structure is geo-differentiated — and be consistent about which geo tier applies.
• Include tenure as context, not as a justification. Tenure explains a gap; it does not automatically excuse one.

Once the set is defined, the useful comparison is not "is this offer higher than theirs?" It is "is the difference between this offer and theirs explainable by something we would be comfortable saying out loud?"

The four questions

For any incoming offer, four questions cover most of the risk.

First: where does this land relative to peers at the same level and scope? A new hire above the median of their internal peer group is not automatically wrong, but it should be a deliberate choice with a reason attached.

Second: does this create compression against anyone above? If the offer lands within a few percent of someone one level up, you have created a problem that will surface at the next review cycle, if not sooner.

Third: does this create a pattern? One placement above peers is a decision. Five in the same function over two quarters is a structural signal that your band is wrong, and it should be treated as a band problem rather than a series of exceptions.

Fourth: if a current employee asked directly why the new person is paid more, is there an answer you would be willing to give them? This is the plain-language version of the whole analysis, and it catches things the percentiles miss.

Every offer is also a statement to the people who already work here. They will eventually find out what it said.

When the answer is "adjust the incumbents"

Sometimes the equity check says the offer is fine and the existing team is underpaid. This is a genuinely difficult outcome because it converts a hiring decision into a budget conversation, often mid-quarter.

The wrong response is to lower the offer to preserve an internal relativity that is itself out of market — you will lose the candidate and still have underpaid incumbents. The better response is to make the offer, log the equity finding explicitly, and put the adjustment into a planned cycle with a stated timeline. What you must not do is note the gap and let it disappear, because the same gap will resurface as a resignation.

Making the check routine

The reason equity checks get skipped is friction, so the fix is to remove friction rather than to add a mandate. The check should run automatically when a number is proposed, name the specific comparable employees and the specific gaps, and state whether the difference is explainable by level, scope, location, or performance.

When it takes seconds instead of an afternoon, it stops being the step that gets dropped under deadline pressure — and the finding, along with what you decided to do about it, becomes part of the record for the next similar hire.