An AEO checklist is a short, repeatable sequence of moves that gets your brand mentioned, ranked, and cited inside AI answers from tools like ChatGPT, Perplexity, Gemini, and Google AI Overviews. This one is built for your first 30 days — not for running a mature program at scale. If you have never measured how AI describes your company, this 30-day AEO checklist gives you the exact order of operations: baseline audit first, seed prompts second, quick wins third. No guesswork, no 200-item spreadsheet.
Most guides hand you a flat list of "best practices" and leave you to sequence it. For a beginner the hard part is not knowing what matters — it is knowing what to do on Monday. This plan fixes the sequence, and every step ends in a number you can defend.
What is an AEO checklist?
An AEO checklist is a defined set of steps that make your content easy for answer engines to select, summarize, and cite. It differs from a classic SEO checklist because the target is not a blue link on a results page — it is a mention inside an AI-generated answer, often with no click at all.
The discipline sits under answer engine optimization: structuring pages so large language models quote your facts when a buyer asks a question. A starter AEO checklist has three jobs — measure where you stand, fix the highest-use gaps, and prove the change with a re-test. Every move in the next 30 days maps to one of those three.
Why 30 days is the right first horizon
Thirty days works because AI answers refresh far faster than organic rankings. Where a new backlink can take months to move a position in classic search, retrieval-based assistants like Perplexity and ChatGPT search re-read the live web and can update their responses within days to weeks of a re-crawl.
That speed cuts both ways. You can see movement inside a month — but only if you measured a baseline first. Without a Day 0 snapshot, any "improvement" is a story, not a result, and marketers defending a budget cannot ship stories.
This is a starter plan, deliberately narrow. Once these 30 days are done, the operating model behind a mature AEO program covers ownership, cadence, and scale. For now, keep scope tight.
Week 1 — Run a baseline AI visibility audit
Start by measuring, not fixing. Your Week 1 job is one number you did not have before: your mention rate — the share of buyer questions where an AI answer names your brand. Everything else this month is judged against it.
Open each assistant in a private/incognito window to strip out personalization, then run your prompt set (built in the next section) across ChatGPT, Perplexity, Gemini, and Google AI Overviews. Run each prompt twice and log five data points every time.
Use this scorecard as your Day 0 template:
| Metric | What it measures | How to log it |
|---|---|---|
| Mention rate | % of prompts where your brand appears | Mentions ÷ total prompts |
| Position | Where you land in the answer | First / middle / last |
| Citation rate | % of answers that link your domain | Note when the source is your site |
| Sentiment | How the AI describes you | Positive / neutral / negative |
| Competitor set | Who appears instead of you | List rival names per prompt |
Log it in a plain spreadsheet. The goal is a repeatable snapshot you can re-run on Day 30 — not a dashboard yet.
How to build your seed prompt set
Your baseline is only as honest as your prompts. The goal is to mirror how buyers actually ask — not how you describe your own product. For a first audit, 15–25 prompts is the sweet spot: enough to reveal patterns, few enough to run by hand in an afternoon.
Weight the set across four intent bands instead of piling on brand-name queries (which flatter you and teach you nothing):
| Prompt type | Share of set | Example |
|---|---|---|
| Category discovery | ~40% | "best [category] tools for [use case]" |
| Use-case / problem | ~30% | "how do I [job your product does]" |
| Comparison | ~20% | "[competitor] alternatives" or "X vs Y" |
| Brand / direct | ~10% | "what is [your brand] and who is it for" |
The 40/30/20/10 split keeps you honest: most AI-driven demand starts with an unbranded question, so most of your prompts should too. For a rigorous view of set size and coverage, see how many AI search prompts you should track. Lock this set as Version 1 — you will re-run the exact same prompts on Day 30, so freezing the wording matters.
Week 2 — Fix your entity core
AI cannot recommend a brand it cannot cleanly identify. Week 2 makes your company legible to a model: a crisp identity, consistent facts, and machine-readable markup. For most beginners this is the highest-use week because it fixes root causes, not symptoms.
Work through these in order:
- Rewrite your About page so the first sentence states plainly what the company is, who it serves, and the category it competes in. Assistants lift this framing almost verbatim.
- Align your brand facts — name, founding, category, key claims — so they match across your site, LinkedIn, Crunchbase, and review platforms like G2. Contradictions make models hedge.
- Add Organization and Article schema that reflects only what is visible on the page. Google is explicit that structured data must match on-page content in its structured data guidelines.
- Confirm AI crawlers can reach you — check that
robots.txtisn't blocking assistant user-agents (GPTBot, PerplexityBot, Google-Extended), and that key pages render without heavy client-side JavaScript.
Do these four and you remove the most common reason a capable product stays invisible: the model isn't sure who you are.
Week 3 — Content quick wins that earn citations
Answer-first formatting is the fastest content lever you have. In Week 3, restructure your highest-intent pages so the answer a buyer wants sits at the very top — before context, before story. Assistants extract self-contained passages; give them clean ones to lift.
Three moves, in priority order:
- Add a 40–60 word direct answer to the top of your five highest-intent pages. Lead with the definition or conclusion, then expand.
- Chunk everything into short paragraphs under descriptive H2/H3 headings phrased as the questions buyers actually ask.
- Publish or refresh one comparison page and one FAQ hub — both are formats assistants quote heavily.
Format choice is not cosmetic. Assistants disproportionately quote structured passages — direct-answer paragraphs, numbered steps, comparison tables, and FAQ blocks — because those lift cleanly without distortion, while long prose usually gets paraphrased or skipped. One well-formatted comparison page often outperforms ten thin blog posts for AI visibility.
Week 4 — Off-site signals and independent sources
AI recommends the brand that independent sources already agree on. Weeks 2 and 3 fixed your own property; Week 4 shapes what the rest of the web says about you, because assistants weight third-party consensus heavily when assembling a shortlist.
You will not build authority in a week — but you can start the flywheel:
- Earn or update 2–3 third-party mentions — an inclusion in a "best [category] tools" roundup, a refreshed review profile, or a mention in a relevant listicle.
- Correct the record anywhere an outdated or wrong description of your brand still lives.
Order matters more than volume here, and why off-site signals move AI recommendations explains the mechanism. Then close the loop: re-run your Version 1 prompt set and compare, metric by metric, against Day 0.
A worked example: 30 days of tracking data
Here is what a first month typically looks like. The figures below are representative of an early-stage B2B SaaS brand running this exact plan — a mid-market tool in a crowded category, tracked across 20 prompts and four assistants. Treat them as a realistic pattern, not a promise; your category and starting authority will change the slope.
| Metric | Day 0 (baseline) | Day 30 | Change |
|---|---|---|---|
| Mention rate (20 prompts) | 15% (3/20) | 45% (9/20) | +30 pts |
| Category-discovery mentions | 1 of 8 | 5 of 8 | +4 |
| Citation rate | 5% | 20% | +15 pts |
| Sentiment (neg/neutral/pos) | 0 / 3 / 0 | 0 / 4 / 5 | more positive |
| Position when mentioned | last | middle | up |
Two lessons repeat across brands that run this. First, the entity-core fixes in Week 2 do most of the heavy lifting — the jump in category-discovery mentions almost always traces to a clearer About page and consistent facts, not to new content. Second, citation rate lags mention rate. Being named comes first; being linked follows as off-site signals accumulate over months. Thirty days moves mentions; citations are typically a Quarter-2 story.
The 30-day AEO checklist (copy-paste)
Here is the whole plan as one runnable list. Print it, assign owners, and check items off by week.
Week 1 — Baseline
- Pick 15–25 buyer prompts across the four intent bands (40/30/20/10).
- Run each in ChatGPT, Perplexity, Gemini, and Google AI Overviews (incognito, twice).
- Log mention, position, sentiment, citation, and competitors for every prompt.
- Calculate your Day 0 mention rate and share of voice. Lock prompts as Version 1.
Week 2 — Entity core
5. Rewrite the About page to define the company in the first sentence.
6. Make brand facts identical across site, LinkedIn, Crunchbase, and review sites.
7. Add Organization + Article schema that matches visible content.
8. Confirm AI crawlers aren't blocked and pages render cleanly.
Week 3 — Content quick wins
9. Add a 40–60 word direct answer to your five highest-intent pages.
10. Publish or refresh one comparison page and one FAQ hub.
11. Chunk content with question-style H2/H3 and short paragraphs.
Week 4 — Off-site + re-test
12. Earn or update 2–3 third-party mentions and correct any wrong descriptions.
13. Re-run the Version 1 prompt set and compare to Day 0.
14. Fix the top three remaining gaps and schedule the next re-test.
That is the complete starter AEO checklist. Fourteen items, four weeks, one measurable outcome.
What NOT to do in your first 30 days
The biggest beginner mistake is optimizing before measuring. Skipping the Week 1 baseline means you can never prove what worked — and you will over-invest in whatever felt productive rather than what moved the number.
Avoid these traps:
- Don't chase all eight assistants at once. Start with the three or four your buyers actually use, then expand.
- Don't add schema you can't back up. Marking up reviews or FAQs that aren't visible on the page violates Google's guidelines and can hurt you.
- Don't publish 20 thin posts. One authoritative comparison page beats a content dump for citations.
- Don't confuse a one-time check with monitoring. A single audit is a photo; visibility is a movie. The point of locking Version 1 prompts is repeatability.
Restraint is a strategy here. A tight, measured 30 days beats a frantic, unmeasured one every time.
What to expect after day 30
Day 30 is a handoff, not a finish line. You now have a baseline, a fixed prompt set, a cleaner entity core, and a first round of results. The job now is to turn a one-off sprint into a routine.
Three directions from here. Increase cadence — daily or weekly checks catch new competitors and answer shifts a monthly glance misses; pick a monitoring frequency that matches your stage. Scale the operating model — assign a clear owner and a repeatable workflow instead of a solo sprint. And set stage-appropriate goals — a seed-stage startup and a category leader need very different targets, as AI search visibility by company stage lays out.
The brands that compound results keep re-running the loop. Baseline, fix, re-test — every month. That is the whole game.
Frequently asked questions
How long until AI search results actually change?
Retrieval-based assistants like Perplexity and ChatGPT search can reflect on-page changes within days to a few weeks once they re-crawl. Knowledge baked into model training updates far more slowly. That is why this plan targets mentions and citations, which move fastest, and treats deep sentiment shifts as a longer arc.
How many prompts should I track to start?
For a first baseline, 15–25 prompts weighted toward unbranded, category-level questions. That is enough to reveal patterns without becoming a manual chore. Grow the set only after you have a repeatable process — precision matters more than volume at the starter stage.
Do I need a tool, or can I do this manually?
You can run the full 30-day AEO checklist by hand — a spreadsheet and incognito windows are enough for Day 0. A tool earns its place once you need daily tracking across many prompts and assistants, historical trends, and reporting you can hand to a stakeholder. Start manual; automate when the manual version breaks.
Is AEO the same as GEO?
Mostly, yes. Answer engine optimization (AEO) and generative engine optimization (GEO) are used almost interchangeably — both aim to get your brand into AI-generated answers. The difference is emphasis: AEO leans toward direct-answer engines and answer-first formatting, while GEO is often used more broadly for any generative surface. For this 30-day plan the tactics are identical, so don't let the label slow you down.
Is AEO replacing SEO?
No — it extends it. Classic SEO earns the click from a results page; AEO earns the mention inside the AI answer that increasingly sits above those results. The foundations overlap (clean markup, crawlability, authoritative content), but AEO adds answer-first formatting and off-site consensus as first-class levers. Run both — the same entity and content work feeds each.
What's the single most important item on the list?
The Week 1 baseline. Without it, nothing else is provable. A clear About page (Week 2) is the highest-use fix, but the baseline is the highest-use action — because it is what turns AEO from opinion into evidence.