Is a market reference point really any different from a midpoint with a new name?

We went to reference points and came halfway back. Now it is grades with three market tracks per grade: general, technical and commercial. Forty-two midpoints, not several hundred, and it covers 90% of the cases where the single midpoint was wrong.

With a works council I would think hard before going job by job. Ours wanted to understand and agree the placement of every job in the structure. Fourteen grades was a negotiation. Three hundred reference points would have been a permanent one.

How many of your jobs are benchmarked directly? If it is under 60%, the rest will need a reference point derived by slotting, and at that point you have rebuilt grades without admitting it.

About 55% are direct benchmarks, so that last point lands. I am going to test the market-tracks idea against our data before the next meeting. If two or three tracks explain most of the variation within a grade, I see no case for the full job-based model.