You already added an llms.txt file. You already dropped in FAQPage schema. You still open Perplexity, type the exact question your product answers, and your name is nowhere in the citations. That gap — the work is done but the citations never showed up — is the moment most site owners start wondering whether they need an AEO audit or whether the DIY checklist just needs one more pass.
An AEO audit is a structured review of the technical and content signals AI agents use to decide which source to cite, covering schema markup, crawl directives, heading structure, and content freshness. AEO Excellence runs this kind of review for agencies and SaaS teams who want to know exactly which signal is missing before they spend another week guessing, and its full methodology is laid out on the about page.
The honest answer is that part of this work is genuinely DIY-safe, and part of it is not. This article walks through exactly where that line sits, so you spend money on an audit only for the part DIY cannot reliably catch.
What Can You Fix Yourself Without an Audit?
You can fix the structural basics yourself, and you should do this before paying for anything. Most of the highest-leverage AEO signals live in files and markup you already control.
- Publish a plain-text
llms.txtfile at your domain root that lists your most important pages in one sentence each. - Add an
AGENTS.mdfile describing what your site does and how an agent should navigate it. - Add Organization, FAQPage, and BreadcrumbList JSON-LD schema to your homepage and top landing pages.
- Rewrite your top three H2 headings as literal questions and answer each one in the first two sentences underneath it.
- Allowlist GPTBot, ClaudeBot, PerplexityBot, Google-Extended, and anthropic-ai in
robots.txt.
Do all five and you will close most of the gap between being invisible to AI agents and being readable by them. The site's blog covers each of these five mechanics in more depth if you want the how-to before the audit conversation. What DIY checklists rarely catch is whether those five signals stay internally consistent once they are all live at once.
Why llms.txt and robots.txt Size Limits Matter
Two structural details trip up more sites than any content mistake does. The llms.txt convention was proposed in September 2024 by Answer.AI's Jeremy Howard as a plain-Markdown file listing a site's key pages in language a model can read without parsing HTML — the specification page documents the format directly. Separately, the Robots Exclusion Protocol, formalized as RFC 9309 in 2022, caps the portion of robots.txt a crawler is required to parse at 500 kibibytes. Anything appended past that limit is silently ignored, which matters if your AI-crawler allowlist sits at the bottom of a long file.
Neither limit is hard to work around once you know it exists. The problem is that most DIY checklists list "add llms.txt" and "update robots.txt" as one-line tasks and stop there, without checking file size or confirming where the AI-crawler allowlist actually lands in the file.
What Does a Professional Audit Catch That DIY Misses?
A professional audit catches errors that only surface when someone checks your markup against what the crawler actually receives, not what your CMS renders in a browser. Content injected by JavaScript after the page loads is invisible to crawlers such as GPTBot and PerplexityBot, which do not execute client-side scripts before reading a page. A site can carry a technically valid FAQPage schema block and still lose every citation because the schema sits behind a script tag the crawler never runs.
An audit also catches contradictions a DIY pass misses: an llms.txt file that lists a pricing page your robots.txt blocks, or an AGENTS.md that describes features your schema markup never confirms. Agents weigh consistency across these signals, so one contradiction can cost a citation even when every file looks correct in isolation.
At the point where checking every file against every other file by hand becomes the bottleneck, a paid audit earns its price: AEO Excellence runs 30 discrete checks across schema, llms.txt, AGENTS.md, and crawl directives, then returns one prioritized report instead of five separate self-checks that never get cross-referenced. If you want to see the exact scope and turnaround before committing to one, the FAQ page covers it in detail.
A concrete version of this problem: a site publishes an AGENTS.md that describes a pricing calculator as a key feature, ships FAQPage schema that never mentions pricing, and leaves the calculator's markup rendered entirely by client-side JavaScript. Each file passes validation on its own. No agent ever cites the calculator, because none of the three signals reinforces the other two, and the one signal that would settle the question — the actual rendered content — is the one the crawler never sees. A DIY checklist run three separate times, once per file, will not surface that pattern; a single pass that checks all three together will.
DIY AEO vs a Professional Audit at a Glance
Here is how the two approaches compare on the tasks that actually move citations.
| Task | DIY | Professional audit |
|---|---|---|
| Publish llms.txt and AGENTS.md | Yes, in under an hour | Included, cross-checked against robots.txt |
| Add FAQPage and Organization schema | Yes, with a schema generator | Included, validated against rendered output |
| Catch JavaScript-hidden content | Rarely caught without crawler-simulation tools | Checked directly against crawler-fetched HTML |
| Cross-file consistency check | Manual, easy to miss | Systematic, 30-point checklist |
| Prioritized fix order | Guesswork | Delivered as a single PDF report |
The pattern in that table is not that DIY is worthless. DIY handles publishing new signals well; an audit handles verifying that the old ones actually work together.
How Long Does Each Approach Take?
DIY implementation of the five basics above typically takes a few hours to a day for someone who already knows their CMS. A professional audit takes longer to deliver but shortens the time to a working result: AEO Excellence's audits arrive as a PDF report, and sites that implement the prioritized fixes typically see citation lift within 14 to 30 days, because AI crawlers refresh their indexes more often than Google refreshes its core ranking signals.
The time cost DIY checklists hide is debugging time. Finding out why a citation never appeared after shipping five separate fixes can take longer than the audit itself, because there is no single report telling you which of the five is the actual blocker.
Is a Paid Audit Worth It If Your Site Already Ranks Well in Google?
Yes, and this is the most common reason sites delay getting one. A top Google ranking and an AI citation are earned by different mechanisms: Google ranks documents using backlinks, keywords, and click-through data, while AI agents pick a citation based on schema clarity, structural consistency, and how easily a passage can be lifted out of context with no surrounding text. A site can hold position one in Google and never appear once in an AI-generated answer for the same query, because none of the signals that earned that ranking are signals an agent reads.
The realistic objection is cost, not doubt. A paid audit is a real expense for a site that has not yet confirmed AI search sends it any traffic. The honest counter to that is narrower than a sales pitch: an audit is the fastest way to find out whether the gap is worth closing before you spend more implementation time guessing at it yourself.
Get a Clear Picture of Your AEO Gaps
If you have already shipped the five DIY basics above and still are not seeing citations, the next step is not a sixth guess. Submit your site through the contact page and ask for an AEO audit — you get back a prioritized list of exactly which signal is breaking the chain, instead of another checklist to work through blind.