# How to use EarnCitations: the first hour, the first week, then every week · EarnCitations

> The free tools in the order to use them: fix what stops AI reading you, learn what buyers ask, then a short weekly routine to see whether answers name you, and why not.

Source: https://earncitations.com/routine.html

> Every page on this site: https://earncitations.com/llms.txt (the whole site as one file: https://earncitations.com/llms-full.txt)

Guide updated 2026-11-22

## How to use it, in order

EarnCitations is a set of free tools, and the order matters. First, fix what stops assistants reading you; until that's fixed, an answer that leaves you out may only mean they couldn't read you. Then learn which questions your buyers ask. Then keep a short weekly routine that shows whether answers name you, and why they don't.

### The first hour: can AI read you?

- [Run the audit](https://earncitations.com/#audit): Your homepage and key pages, read the way AI crawlers read them. Start with "Fix first". Download the backlog (CSV) or copy it as an issue, so each fix has an owner.

- [Draft my fixes](https://earncitations.com/#audit): Under the result: an llms.txt, robots.txt rules for AI crawlers, Organization JSON-LD and a Fact Ledger, all drafted from your own pages. Every gap is a TODO for you; nothing is guessed. Or copy the prompt for your coding agent.

- [What AI sees](https://earncitations.com/view.html): For your pricing, product and FAQ pages: the text a crawler that doesn't run JavaScript gets, the facts it can take, and what it misses (widgets, blocked agents, dates that disagree).

- [Delivery check](https://earncitations.com/delivery.html): After you deploy, check that each change reached the crawlers: live, in the HTML (and in the Markdown copy, if the page advertises one), allowed, indexable, in the sitemap, linked. For several pages at once, from the command line: node bin/earncitations.mjs delivery --from pages.txt.

- [Docs check](https://earncitations.com/docs.html): Only if you sell to developers: could a coding agent set your product up from your public docs?

### The first week: what do buyers ask?

- [Read your analytics export](https://earncitations.com/logs.html#export): How many visits AI assistants already send you, and to which pages. It tells you how much tracking is worth doing: if the share is small, a short question set is plenty.

- [Read your server logs](https://earncitations.com/logs.html): Which AI agents actually came, what they read, and what failed. Nothing is uploaded.

- [Question map](https://earncitations.com/map.html): The questions buyers are likely to ask AI about you, proposed from your own pages, each checked: answered, weak page, missing page or crawl problem. Download them as a question set.

- [Or write your own questions](https://earncitations.com/track.html#own): If you already know what buyers ask, type the questions, one a line. Keep them to your main buying decisions, and put a line like ## Pricing above each group to see your rates per decision.

- [Answer check](https://earncitations.com/claims.html): Ask each question in ChatGPT, Perplexity, Gemini or Claude yourself, then paste the answer and its links. Each claim is checked against the pages it cites. Save the answer to your answer log.

### Every week

- [Answer log: what to ask next](https://earncitations.com/track.html#due): The questions that haven't been asked for each engine lately. When an engine doesn't answer, log that too, so your rates show what's missing. A line above the list says what your routine gives you: answers a month, and how big a change is still noise.

- [Where a competitor is the answer](https://earncitations.com/track.html): Questions where a competitor is named in more answers than you are. Start with the widest gap. Set the rates to the last 30 days so old answers don't hide this month's change.

- [Question doctor](https://earncitations.com/question.html): For one of those questions: is your answer blocked, missing, or there but weak? You get the page to fix, the one change to make (or a prompt to hand it to your coding agent), and a Delivery check to confirm it shipped.

- [Run the audit again](https://earncitations.com/#audit): Your browser keeps the last result, so the new one starts with what changed. To be told when something breaks, run node bin/earncitations.mjs watch yourdomain.com --webhook URL on a schedule (daily, from cron or CI): each run compares with the last and posts only when something changed for the worse ([how](https://earncitations.com/developers.html#watch)).

### Every month

- [Before and after a fix](https://earncitations.com/track.html#change): Once a fix has been live for a couple of weeks and you have kept asking its questions, compare the answers on either side of the day it went live: improved, unchanged, worse or inconclusive, per question and engine. A change is a sequence, not proof.

- [Sources to pitch](https://earncitations.com/track.html): The pages that answers cite where a competitor was named and you weren't. If you have a true reason to be there, ask to be included.

- [Head-to-head](https://earncitations.com/vs.html): The same audit on your site and a competitor's, showing only what differs.

- [AI price check](https://earncitations.com/prices.html): After a price change: do the prices crawlers read in your HTML agree with your markup?

Every tool on this page is free and needs no EarnCitations account (asking the assistants themselves needs your own accounts with them). The command-line versions run on your own machine: [developer quickstart](https://earncitations.com/developers.html#quickstart). Terms you meet along the way are in the [glossary](https://earncitations.com/glossary.html).
