What is AEO, and what does it have to do with ads?

AEO, or answer engine optimization, is the work of getting your brand named when someone asks an AI assistant like ChatGPT, Gemini, or Claude for a recommendation. The same check that shows where AI assistants name you, ignore you, or credit a rival is a ready-made map of where to spend your ad budget and where not to.

Does Adacity write my ads?

No. Adacity surfaces the real buyer questions, where your brand is named or absent, and which questions you can win. You or your team turn each one into an angle. It is the brief, not the creative.

Which questions should I run ads against first?

Attack first: the early questions where you're absent and no brand dominates. Defend the strengths AI assistants already credit you for. Avoid the head-to-heads a stronger competitor already wins, and co-occupy in the roundups instead.

Defend, attack, or avoid: how to turn your AI visibility report into a marketing strategy

Q: How is this different from keyword research?

Keyword research gives you short phrases people type into a search box. This gives you full questions people actually ask AI assistants, across ChatGPT, Gemini, and Claude, scored only when AI assistants name you on their own, and sorted by whether you should defend, attack, or avoid.

The short version

A good AI visibility (AEO) report hands you the exact questions buyers ask AI assistants, plus the answer and citations each one returns. Read those and sort each question into a bucket, and that list becomes your strategy. To make it concrete, we ran the numbers on Quince, a real DTC apparel brand, across ChatGPT, Gemini, and Claude.

Defend the ground AI assistants already give you. The models credit Quince for affordability 298 times and its factory-direct model 206 times. That is the ground to hold before you spend anywhere else.
Attack the questions nobody has claimed. Only 3 of Quince's 78 category questions name the brand, and no competitor dominates the category. The 75 open questions are ad briefs nobody has written.
Avoid uphill battles. Quince wins just 37% of its head-to-head comparisons, and the losses cluster against the category's biggest names. Show up beside them in the roundups instead of spending to go head to head.

Step 1: Defend the claims AI assistants already credit you for

These are the strengths AI assistants repeat about you without being asked. Holding a claim you already own is cheaper than building one you don't, so the plan starts here.

When we audited Quince, AI assistants named it for affordability 298 times, and its factory-direct model 206 times. They bring these up most, so lead your ads with them. You're amplifying a message they already agree with.

Also look for credit you're getting but don't feature prominently. AI assistants tie Quince to ethical manufacturing and price transparency, two strengths its own site barely mentions. If the models already credit you for it, feature it in your ads.

Step 2: Attack the questions nobody has claimed

Attack the early questions where you're absent and no competitor owns the answer yet. This is the cheapest ground to take, because you're arriving before a competitor, not unseating one.

Out of 78 category questions we asked across ChatGPT, Gemini, and Claude, only 3 named Quince (the full read is here). Read as a loss, that's discouraging. Read as an ad strategist, it's 75 open questions where a buyer is deciding and your competitor is getting the mention by default. And no one dominates this category: the leader, Everlane, holds only 21.1% of the share of voice, so the top of the funnel is genuinely open.

So a question like "what is an affordable alternative to designer cashmere" becomes a hook built on the affordability AI assistants already tie to you, aimed at the moment a buyer is first forming a shortlist. You are not guessing the angle. You are answering a question you can see, with a claim you know holds. Win those early questions and you're in the consideration set before the comparison even starts.

Step 3: Avoid uphill battles

The head-to-heads where a stronger competitor already wins are the ones to skip. Spending there sets you up for an expensive fight.

Quince wins 37% of its head-to-head comparisons, 165 of 442, and the losses concentrate against two brands already ahead of it in share of voice, Madewell and Everlane. Ad spend won't close that fast, and a buyer comparing Quince to Everlane directly is deep enough in the funnel that one ad rarely flips them.

So don't pay to fight this one, co-occupy it. 93% of what AI assistants cite about Quince is third-party, so get into the roundups and review pages your competitor already sits in, and buyers find you beside them, not beaten by them.

Now build your own plan

You have the method: pull the buyer questions in your category, sort each answer into defend, attack, or avoid, and point your budget at the buckets. Do it by hand and you have what you need to start.

If you would rather have it automated, packaged, and backed by real data on your own brand, run an Adacity report. It gathers the questions buyers actually ask AI assistants across ChatGPT, Gemini, and Claude, scores where you are named or missing, and hands back which questions to spend on and which to skip. Free to start, or $39 for the full-depth run, with no account and no sales call, in your inbox in under an hour. See how it works, or run yours now.

The short version

Step 1: Defend the claims AI assistants already credit you for

Step 2: Attack the questions nobody has claimed

Step 3: Avoid uphill battles

Now build your own plan

✦ Find out if AI assistants recommend your brand to buyers

Find out if AI assistants recommend your brand to buyers