EN
FR
ES
DE
Model drift refers to how an AI's behaviour changes over time: updates, retraining and version changes alter its answers. As a result, a brand can see its visibility move without having changed anything on its side.
What is model drift?
Language models are not fixed: providers update them, retrain them and switch versions. As a result, the same question can receive different answers from one month to the next. This is model drift.
Why does it matter for measurement?
Without precautions, this drift would be mistaken for a genuine variation in a brand's visibility. A score going up or down might simply reflect an update to ChatGPT or Gemini, rather than any action by the brand.
How we keep it in check
The AGS isolates this drift using an anchor set of control brands. A single measurement is misleading: that is why we favour tracking the trend over time. See our acknowledged limits.
Every question asked to ChatGPT without your name in the answer is a competitor recommended instead of you — measured across 6,820 real AI answers.