Six guides, in
the right order.
Six guides, ordered as a path rather than a list. The first three are the foundations — each takes about a quarter of an hour and none of them require a developer. The middle two compound on that base over a couple of weeks. The last one is for engineers who want to rebuild the measurement layer themselves, and it is deliberately honest about what that costs.
Work down the path in order, and resist the temptation to skip to the interesting-sounding one. The foundations exist because the later guides assume them: the Perplexity guide takes for granted that your content is already chunked into standalone answers, and the AI Score guide opens by telling you to ship llms.txt and FAQ schema, which are two separate guides of their own. Doing them out of order is not harmful, it is just slower — you end up reading the same prerequisite three times from three different angles.
Each guide states a realistic time cost up front and closes with an expectations paragraph that says what actually happens afterwards, including the lag. That second part matters more than it sounds. Almost every one of these changes has a delay between shipping and measurable movement, usually one to two weeks while indexes re-crawl and retrieval signals re-score, and the most common failure mode we see is a team shipping all of it in an afternoon, seeing nothing on day three, and concluding it does not work. Prerequisites are listed on every step below so you can see the chain before you start.
Foundations
Easy · 3 guidesNo developer required and none of them take longer than twenty minutes. Between them these three account for most of the movement teams see in their first month, and every later guide assumes they are already done.
- 1
How to set up llms.txt
The robots.txt for language models, in 15 minutes.
llms.txt is a single Markdown file at your domain root that tells AI crawlers what your site is about and which pages matter. It's a 2024 proposal from Jeremy Howard / Answer.AI that's now honored by Claude, Perplexity, and Bing's grounding layer.
15 minutes6 stepsEasyRead first: How to add FAQ schema to any page
- 2
How to get cited by ChatGPT
A 20-minute setup that moves the needle.
Being cited by ChatGPT is not a single-feature fix — it's a cluster of signals working together. This is the minimum-viable setup that, in our data, produces a measurable citation rate bump within 14 days.
20 minutes5 stepsEasyRead first: How to set up llms.txt, How to add FAQ schema to any page
- 3
How to add FAQ schema to any page
The single highest-leverage structured data change you can ship.
FAQPage schema is the highest-hit-rate schema type for AI citations. It's also trivial to implement — a JSON-LD block in the page head, nothing more. Here's how to do it right.
15 minutes5 stepsEasyRead first: How to set up llms.txt, How to improve your AI Score by 20 points
Compounding
Moderate · 2 guidesA few hours of work spread across a week or two, on top of the foundations. These are the guides that change how you write and publish rather than what sits in your site root.
- 4
How to improve your AI Score by 20 points
A checklist that reliably produces a 20-point swing in 14 days.
The AIRank AI Score is a 47-point rubric, weighted against real citation-rate data. Most sites score 40–60 on first scan. This is the checklist that reliably moves that number to 70+ inside two weeks.
2–4 hours spread over a week6 stepsModerateRead first: How to get cited by ChatGPT, How to set up llms.txt
- 5
How to rank in Perplexity
Perplexity is unusually gameable if you understand its priors.
Perplexity's retrieval layer is the most transparent of the major AI answer engines — it publishes citations inline, it prefers recent content, and it weights domain authority heavily. Here's how to show up in its citation panel reliably.
1–2 hours per target query6 stepsModerateRead first: How to get cited by ChatGPT, How to track AI citations without AIRank
Build it yourself
Advanced · 1 guideAn engineering project. Read this one to understand the shape of the measurement problem before deciding whether to build or buy — the honest cost accounting is in the closing section.
- 6
How to track AI citations without AIRank
A DIY setup you can build in a weekend.
This is a how-to for building a minimum-viable AI citation tracker yourself. It's meant for engineers who want to understand the shape of the problem before deciding whether to build or buy. AIRank does all of this and more, but the DIY version is a useful reference.
4–8 hours to build, ongoing maintenance6 stepsAdvanced
Stuck on vocabulary rather than mechanics? The glossary defines everything these guides assume, and the documentation covers how the product itself works.
Which guide should I start with?
Start with FAQ schema if you want the largest single change first, or with llms.txt if you want the fastest one. Both are marked Easy, both take about fifteen minutes, and both are prerequisites for almost everything else here. Once those two are live, the guide on getting cited by ChatGPT ties them together with the two remaining steps that need a human rather than a code change.
Do I need a developer to follow these?
Not for the three Easy guides. llms.txt is a Markdown file uploaded to your site root, FAQ schema is a JSON-LD block your CMS can usually hold in a custom field, and the ChatGPT guide is mostly editorial work. The Moderate guides involve rewriting headings and surfacing content currently hidden behind JavaScript, which sometimes needs a developer. The Advanced guide is an engineering project.
How long before any of this shows up in AI answers?
Perplexity is fastest — a crawled change can appear in answers within a day or two. ChatGPT lags because it grounds through Bing's index, so expect flat results in week one and movement in week two. Claude and Google's AI Mode sit in between. Across the whole set, ten to fourteen days is the realistic point at which to judge whether something worked.
Do these guides only work if I use AIRank?
No. Every guide here is written to be followed by hand with no account, including the one that explains how to build your own citation tracker and what it costs to run. AIRank automates the measurement and can apply several of the fixes through the WordPress and Shopify integrations, but nothing on these pages is gated behind it.
What is the difference between these and the documentation?
How-to guides are task-shaped and product-neutral: they describe a change you make to your own site. The documentation is product-shaped and explains how AIRank itself works — installation, the AI Score rubric, the API, how citations are detected. If you are asking “how do I do X”, you are in the right place; if you are asking “what does this number mean”, the docs are.
Not sure which of these you actually need? A scan takes about a minute and tells you which are already done.
Run a free scan