Promotions and pay: separating level, scope, and market

A promotion answers three different questions at once. Blurring them is why promotion increases feel arbitrary.

Promotion pay conversations are unusually hard because three separate questions get compressed into one number. Has this person moved to a higher level? Has their scope grown, and by how much? And what does the market pay for the job they are now doing? Each has a different answer source, and treating them as one question is why promotion increases so often feel improvised.

Level is a threshold, not a reward

The first question is binary: does this person meet the definition of the next level? Levelling frameworks exist to make that determination on evidence — scope, autonomy, impact, complexity — rather than tenure or advocacy.

When level determinations drift into being rewards for good work, everything downstream breaks. You end up with people at a level whose pay band assumes responsibilities they do not hold, which creates compression against people who do hold them, and undermines the meaning of the level for everyone.

Scope is where the range placement comes from

Once the level is settled, scope determines where in the new band the person should sit. Someone newly meeting the bar for a level typically lands in the lower portion of that band. Someone who has been operating at that level for a year before the title caught up may legitimately land near the middle.

This is where the "promotion increase percentage" convention causes trouble. Applying a flat percentage to everyone's current salary produces different range placements for people at the same new level, purely because of where they happened to start. The result is that two people promoted to the same level on the same day end up materially apart, for no reason connected to the job.

Promote to a position in the new range, not to a percentage above the old salary.

Targeting a range position rather than a percentage bump fixes this, though it needs a stated policy for the cases where the gap is large: a phased path over two cycles, with the target and the timeline written down, is usually better than either a shock increase or an unexplained shortfall.

Market is the sanity check

The third question — what the market pays for the new job — should already be embedded in the band. If it is not, meaning the band for the new level looks unrelated to what you would pay to hire that role externally, the promotion has surfaced a structure problem.

A quick check worth running on every promotion: if this person left and you had to hire their replacement at the new level today, what would you pay? If the answer is meaningfully above what you are about to pay them, you have identified a retention risk in the moment of celebrating them.

The equity check nobody runs

Promotions create compression as reliably as external hires do, and the check is skipped more often because the decision feels internally sourced and therefore safe.

Before finalising, compare the new number against everyone already at that level in that job family, and against the level above. A promotion that places someone above longer-tenured peers at the same level, or within touching distance of their own manager, is a decision you can still make — but it should be made knowingly, with the reasoning recorded, rather than discovered at the next review cycle.