Hi team, I am getting a particular attribute as default hashed attribute, which when missed to explicitly change while creating an experiment inside a feature, is causing a lot of problem across the org. How is default hashed attribute selected? I tried fixing alphabetical order in the naming of the attritbute and all, didnt help.

Hi @delightful-shoe-91325 👋 For an experiment created inside a feature, GrowthBook first filters to active attributes available to that feature’s project and marked as an Identifier. It then selects: 1.

id

, if an eligible attribute with that exact key exists; 2. Otherwise, the first eligible identifier in the organisation’s underlying attribute schema; 3. If none exist, it falls back to

id

. It isn’t based on alphabetical order, so renaming the attributes won’t change the default.

Ah, thank you @flaky-noon-11399

Otherwise, the first eligible identifier in the organisation’s underlying attribute schema;

-- Can you help elaborate this point? Now that I have more than two identifiers whose all names have id in them with some prefixes, how do I make sure I get to choose the default instead of automatic one?