What is model drift?
Model drift is a change in the answers an assistant gives to the same question over time, caused by model updates, new search indexes or changed retrieval. Your visibility can move without any change on your side. Weekly tracking over a fixed set of questions shows whether a change is lasting or one-off.
How does it differ from run-to-run noise?
Sampling variance is the spread between two asks of the same question on the same day. Drift is a shift in the average over weeks. You need the noise level to tell the two apart.
Example: your brand is named in 6 of 10 answers for a month, then in 3 of 10 after a provider updates its model, with no change on your site.
How does Nodpick handle it?
Nodpick asks every question once a week, asks again when a brand's mention flips, and counts a weekly change as real only when it exceeds the brand's normal swing. Run the free AI visibility check to see your own numbers; the method shows how they are made, and pricing lists the plans.
Read more: How to track your brand's visibility in AI answers