How do you handle battery ageing data when test cells differ between labs?

We are comparing cycle-life results across two labs, and the test cells and fixtures are not identical.

I want to know how others account for equipment differences before drawing conclusions.

What do you document or normalise so the comparison still holds up in review?