Fair Use Policy

A fair use policy (sometimes an acceptable use policy in the same document) is the clause that gives a vendor the right to limit, throttle, or reprice an account whose consumption is judged excessive, even on a plan advertised as unlimited. It is what makes "unlimited" commercially survivable: without it, a single automated customer could consume an arbitrary share of a service sold at a flat price. For a buyer, the practical meaning is that unlimited is a marketing position and the fair use policy is the actual limit — one that is frequently unpublished, expressed in language like "materially exceeds typical usage by comparable customers", and evaluated after the fact rather than enforced by a counter you can watch. That vagueness is deliberate on the vendor's side and a planning problem on yours. What matters is not whether the policy exists — it always does — but how it is enforced and what warning you get. Good policies name a number, or at least a percentile, and commit to notifying you before action is taken. Weaker ones reserve the right to suspend immediately. Before you build a workflow on an unlimited plan, ask three things: is there a documented threshold, what is the first enforcement step (a warning, a throttle, a forced tier upgrade, or suspension), and does the account dashboard show your consumption against whatever the limit actually is. Also check whether automated or programmatic use is excluded from unlimited altogether, which is common — many unlimited plans are scoped to interactive human use, and running the same volume through an API or a scheduled job falls under metered pricing instead. Treat the answer as a cost input, because discovering the ceiling by hitting it means discovering it in production.