Why does purity vary between lots?

This is the reason a certificate belongs to a lot and not to a product. Two vials bearing the same compound name from the same supplier are not equivalent unless both have been characterised.

It is also why we do not carry a previous lot's result forward. When a new lot is released it is tested again, and the certificate published against it is that lot's certificate. A supplier displaying one unchanging certificate is either not making new material or not testing it.

Variation within specification is normal and expected. A result outside specification is a different matter, and that lot is not released.

What varies more than purity does

Purity is actually among the more stable parameters between lots, because purification is run until the target meets specification — the process is steered toward a fixed endpoint. Net peptide content varies considerably more, since it depends on how much residual water and counter-ion the material retained through the final drying step, and nothing forces that toward a particular number.

This matters if you are working across two lots. Two lots at the same stated purity can differ in net peptide content by several percent, which means the same fill weight represents a different amount of peptide. Any comparison made across lots without accounting for that has a variable in it that has not been controlled.

The practical habit that follows: treat each lot as its own material and read its certificate rather than the one you read last time. Lot pages are permanent precisely so that a certificate you consulted six months ago is still retrievable when you need to check what the previous batch actually was.