A context hint is a short, structured statement that tells an answer engine which situation your ad group is for. It is not a headline and not a keyword. It is a targeting instruction written in natural language, and the engine reads it the way a thoughtful account manager would read a brief.
Every good hint carries four elements. The persona says who this is for. The intent says what they are trying to do right now. The scope says the boundary conditions — size, budget, timeline, region. The disqualifier says who this is explicitly not for. The disqualifier is the element teams omit and the one that most improves precision, because it lets the engine rule you out of situations you would have lost anyway.
Keep hints under 280 characters. Longer hints do not add nuance; they dilute the signal and make it harder for the engine to extract the constraints cleanly. Write in plain declarative sentences, use concrete numbers, and avoid brand-voice adjectives entirely — save those for the promise, not the targeting.
Write one hint per situation, not one hint per product. If the same product serves three distinct situations, that is three hints in three ad groups, each with different scope and different disqualifiers. Merging them produces a hint that describes everyone and therefore matches no one strongly.
Check each hint against your evidence before shipping it. If a hint claims you suit teams under twenty-five seats, something in your public material must say so. A hint that asserts a constraint your evidence cannot support does not just fail — it teaches the engine that your statements need verification, which suppresses future mentions.
Finally, write hints your client would recognise. A hint set is the clearest artefact you can bring to a review meeting: it shows exactly which situations you are buying into and which you are declining, in language a marketing director can argue with. That legibility is a commercial asset as much as a technical one.