SEO and AI search

How to appear in ChatGPT and AI answers: a 30‑day plan for online stores

This 30-day plan shows how to appear in ChatGPT, Gemini, Perplexity and Google AI answers. You pick 20 buyer questions and record who the engines name. Then you fix crawler access and product data, rewrite key pages and ask again monthly. AI for Marketing Ops can run it for you with our SEO for ecommerce and B2B websites service.

Book a discovery call

Updated:

Who it is for
Ecommerce marketing teams and store owners on PrestaShop, WooCommerce, Magento or Shopify, with access to the store admin and Search Console.
Time
30 days. Our estimate is 2 to 4 hours a week for one person, plus page writing in weeks 2 and 3. Structured data changes can need a developer.
Outcome
A baseline of 20 questions in 4 engines, clean crawler access and product data, 10 rewritten pages and a monthly re-run. The goal is more answers that name your store. No one can promise that result.
Tools
ChatGPT · Gemini · Perplexity · Google Search · Google Search Console · Rich Results Test · Google Analytics 4 · A spreadsheet

Steps

  1. 01 Day 1 to 2

    Pick 20 buyer questions

    Write 20 questions a buyer asks before they buy what you sell. You use the same 20 questions for the whole plan, so each month compares the same questions.

    Mix three kinds: questions about the category, comparisons between options, and questions about buying from a store like yours. Take the words from customer emails, chat logs, reviews and Search Console queries.

    What to do

    • Export the top queries from the Search Console Performance report for the last 3 months.
    • Read 30 recent customer emails or chat messages and note the questions in the buyer’s own words.
    • Write 20 questions: about 10 on the category, 5 comparisons and 5 on delivery, returns or trust.
    • Mark the 5 questions closest to a sale. You check those first every month.

    When it is done: A sheet with 20 buyer questions, each tagged by type and by the page on your store that should answer it.

  2. 02 Day 3 to 5

    Record a baseline in the AI engines and Google

    Ask each question in ChatGPT, Gemini, Perplexity and Google, and note which stores and brands each answer names. Without this starting point, you cannot tell later if anything changed.

    AI answers change from one run to the next. Use a signed-out or new session, the same country and the same language every time, and save a screenshot of each answer.

    What to do

    • Ask all 20 questions in ChatGPT with search on, Gemini, Perplexity and Google Search. Note when Google shows an AI Overview.
    • For each answer, record whether your store is named, which competitors are named and which pages are cited as sources.
    • Save a screenshot and the date of every answer.
    • Count the questions where your store is named, per engine.

    When it is done: A baseline table of 20 questions by 4 engines, with your mentions, the competitors named and the cited sources.

  3. 03 Week 2

    Fix crawler access, structured data and company facts

    Check that search and AI crawlers can read your store, and that your product and company facts are clear and the same everywhere. An engine cannot cite a page it cannot read.

    Google says AI Overviews and AI Mode have no extra technical requirements. A page must be indexed and eligible to show a snippet in Google Search. ChatGPT search uses the OAI-SearchBot crawler and Perplexity uses PerplexityBot. Both companies say robots.txt controls them, and OpenAI says blocking GPTBot for training is a separate choice.

    What to do

    • Open robots.txt and confirm that Googlebot, OAI-SearchBot and PerplexityBot are allowed. Check bot rules at your CDN or firewall too.
    • Decide on Google-Extended. This robots.txt token lets you choose if Google may use your pages to train Gemini and to give Gemini facts at answer time. It does not affect Google Search.
    • Add Product structured data with name, image, price, currency and stock to product pages, and test it in the Rich Results Test.
    • Add Organization structured data with your legal name, logo, address and profile links.
    • Make your company facts match on the site, Google Business Profile, Merchant Center and marketplaces.

    When it is done: A crawler check, valid Product and Organization structured data, and one list of company facts used everywhere.

  4. 04 Week 2 to 3

    Write answer-first category and product pages

    Rewrite the pages that should answer your 20 questions, so the answer comes in the first two sentences. The buyer gets the answer without scrolling, and the details follow.

    Google asks for important content as text, with structured data that matches the visible page. The GEO research paper (Aggarwal and others, KDD 2024) tested page edits on AI engines. Adding cited sources, quotations and statistics raised a page’s visibility in their tests, and keyword stuffing gave little or no gain.

    What to do

    • Start with the 10 category and product pages that match your 5 questions closest to a sale.
    • Open each page with a 40 to 60 word answer to its question, in plain words.
    • Add the facts a buyer checks: sizes, materials, delivery times, return terms and a comparison table where options differ.
    • Cite a source for every number, and quote a real person only with their permission.
    • Keep key text as HTML text, not inside images.

    When it is done: 10 rewritten pages, each with a direct answer at the top, sourced facts and structured data that matches the text.

  5. 05 Week 3 to 4

    Earn mentions on other sites

    Get your store named on the sites the engines already cite for your questions. Your baseline table lists those sources.

    Earn each mention with something useful: data from your own store, a clear guide or a fair product test. Do not buy links. Google’s spam policies count bought links, and links given in return for free products, as link spam unless they are marked.

    What to do

    • List the 10 sites cited most often in your baseline: review sites, guides, forums and marketplaces.
    • For each site, note how a store can appear there: a product test, an expert answer, a listing or a correction.
    • Offer each site something real: data from your store, an answer from your team or a product to test. If you send a free product, ask that any link to your store carries rel="sponsored".
    • Correct wrong or old facts about your store wherever you find them.

    When it is done: A list of 10 cited sites with one action each, and the first pitches or listings sent.

  6. 06 Day 30, then monthly

    Ask the same questions again every month

    On day 30, ask the same 20 questions in the same 4 engines, in the same way, and compare the answers with your baseline. Then repeat it on the same day every month.

    Change is slow and uneven, and one run can mislead you. Check the trend over 3 monthly runs before you judge a page change.

    What to do

    • Ask the 20 questions with the same settings as in step 2, and save the screenshots.
    • Count the mentions per engine and compare them with the baseline.
    • In Search Console, check clicks and impressions for the rewritten pages. Google counts AI Overviews traffic inside the Web search type.
    • In Google Analytics 4, check sessions from chatgpt.com, perplexity.ai and gemini.google.com.
    • Choose the next 10 pages to rewrite from the questions where no engine names you yet.

    When it is done: A one-page monthly report: mentions per engine, change since the baseline and the next 10 pages.

Mistakes to avoid

  • Changing the questions every month. You lose the comparison with your baseline.
  • Blocking OAI-SearchBot or PerplexityBot by mistake, for example with a CDN bot rule. ChatGPT search and Perplexity then stop showing your pages in their search results.
  • Writing for the engine instead of the buyer. Keyword stuffing gave little or no gain in the GEO study.
  • Structured data that differs from the page, such as a price or stock status the buyer does not see.
  • Judging the plan on one run. AI answers change between runs, so judge the trend.

What to measure

Metrics for the 30 days and after
MetricHow to measureWhen
Store mentions in AI answersNumber of the 20 questions where each engine names your storeBaseline, then monthly
Cited pagesWhich of your URLs the answers link to as sourcesMonthly
Clicks from GoogleSearch Console Performance report, Web search type, filtered to the rewritten pagesMonthly
Visits from AI assistantsGoogle Analytics 4 sessions by source: chatgpt.com, perplexity.ai, gemini.google.comMonthly
Valid structured dataRich Results Test and the Search Console reports for productsAfter each page change

Rather have us run it

We run this playbook with you or for you

Sources

  1. Google Search Central: AI features and your website
  2. Google Search Central: Google common crawlers (Google-Extended)
  3. Google Search Central: Product structured data
  4. Google Search Central: Organization structured data
  5. Google Search Central: Spam policies
  6. Google Rich Results Test
  7. OpenAI: Overview of OpenAI crawlers
  8. Perplexity: Perplexity crawlers
  9. Aggarwal and others: GEO, Generative Engine Optimization (KDD 2024)

Questions

Common questions

How do I appear in ChatGPT?

ChatGPT search finds pages with its OAI-SearchBot crawler, so allow it in robots.txt. Then make sure your pages answer buyer questions in plain text, and get named on the sites ChatGPT already cites. No one can guarantee a mention, so ask the same 20 questions every month and count the results.

What is on a generative engine optimization checklist?

Generative engine optimization (GEO) is the work to get your pages used and named in AI answers. The term comes from a research paper by Aggarwal and others, first published in 2023 and presented at KDD 2024. A GEO checklist covers what this plan does: buyer questions, a baseline, crawler access, structured data, answer-first pages, mentions on other sites and a monthly re-run.

Do I need special markup or an llms.txt file for AI answers?

Google says no. AI Overviews and AI Mode need no special schema.org markup and no new AI files. Product structured data with the required properties can make your pages eligible for product results in Google, so add it. An llms.txt file costs little to add, but no engine in this plan asks for one.

How long until my store appears in AI answers?

There is no fixed time, and no one can promise a result. OpenAI and Perplexity say robots.txt changes take about 24 hours to reach their systems. Page changes and mentions on other sites take longer. Judge the trend over three monthly runs, not one.

Book a call to run this plan with us

On a free 30-minute call, we plan your first steps with you. Before the call, our agents check whether ChatGPT, Gemini and Google name your store for 20 customer questions, and a person reviews the check. Every SEO project we run includes this AI search baseline. You keep the short plan, even if we do not work together.

Book a discovery call

30 minutes · Free · Online