Skip to content

Find out who AI names in your category.

We ask up to three AI engines (OpenAI, Perplexity and Claude, via API with web search) 20 buying questions about your category, three times each, and show you who they name: you or your competitors.

Email us for a $99 report

Delivered within 2 business days, or a full refund.

Questions first? Email Jake, the founder: hello@ecomindex.ai

A shopper asks

best paddle board for flat water

Do the answers name your brand?

  1. OpenAI (GPT via API, web search)

    For flat water, Brand A is the easiest first board: wide, stable and light to carry. Brand C costs less if you paddle a few times a summer.

    1reviews.example2forum.example

  2. Perplexity (Agent API, web search)

    Most beginner guides start with Brand A and Brand B. Both are at least 32 inches wide, which matters most on day one.

    1magazine.example2video.example

  3. Claude (API, web search)

    I'd look at Brand B first: it's stable and comes with a pump and paddle. Brand A is the other board reviewers keep recommending.

    1gearsite.example

Your store: not named in these 3 answers

Illustration with made-up brands, answers and sites.

Before you rewrite a single page, see which brands AI engines name in your category, and which sites they cite.

  1. Asked

    20 questions a buyer in your category would ask, put to three engines, three times each.

  2. Counted

    Every brand each answer names, out of every answer that came back, with the sources it cites.

  3. Fixed

    A fix plan for your pages and the sites worth pitching. Or we do the work for you.

  • OpenAI (GPT via API, web search)
  • Perplexity (Agent API, web search)
  • Claude (API, web search)

AI Shelf Report

Your store

OpenAI 59/60, Perplexity 60/60, Claude 59/60 completed answers, 20 buyer questions, 3 runs each

2%AI engines named you in 2% of 178 completed API answers to 20 buyer questions, across 3 engines.

Named instead: Brand A (74), Brand B (58) and Brand C (33)

Brand A was named 18.5x as often as your store

Leaderboard

  1. 1Brand A

    OpenAI 25 / Perplexity 30 / Claude 19

  2. 2Brand B

    OpenAI 19 / Perplexity 25 / Claude 14

  3. 3Brand C

    OpenAI 12 / Perplexity 11 / Claude 10

  4. 4Brand D

    OpenAI 7 / Perplexity 5 / Claude 7

  5. 5YouYour store

    OpenAI 1 / Perplexity 1 / Claude 2

The top of a report. Illustration with made-up brands and counts.

What's in the report

One private web page, in this order. It saves as a PDF from your browser.

  1. The verdict

    How many completed answers named you, out of how many, across how many engines.

  2. The leaderboard

    Every brand the answers named, by share of answers, with each engine's count.

  3. Where AI sends shoppers to buy

    The retailers and marketplaces the answers point to.

  4. Answer cards

    Two answers from each engine, quoted with their numbered citations.

  5. Three ChatGPT app screenshots

    Taken by hand for the questions where you were named least, each one labeled.

  6. The sources AI cites

    The sites in the answers' citation lists, most cited first.

  7. Your fix plan

    Which pages to rewrite and which cited sites to approach, edited by Jake.

  8. The full method

    Engines, model IDs, dates, completed answers, all 20 questions and every answer.

See a full sample report 

See the $750 sample, with the playbook 

Start with a report

AI Shelf Report

$99 once

Who AI names in your category, and the sources the answers cite.

Buyer questions, written for your store
20
Engines, via API with web search
3
Runs of every question
3
Answers read and counted
Up to 180
Labeled ChatGPT app screenshots
3
Fix plan, edited by Jake
5 or 6 changes
Delivered
2 business days

Counts toward the Playbook if you upgrade within 30 days.

See a full sample report

AI Shelf Report and Playbook

$750 once

The report, plus a 90-day plan built on your own Search Console.

Everything in the $99 report
Included
What answers repeat from your top competitors' pages
2 brands
Sources to pitch, with who to approach
Up to 10
Your Search Console, last 28 days
Required
Crawler, product page and catalog checks
Included
The 90-day playbook
30 to 50 changes
Delivered, after Search Console access
2 business days

Counts toward your first month of Get on the shelf if you start within 90 days.

See the $750 sample, with the playbook or download the PDF

If a report is late, or we can't analyze your store, you get a full refund.

Questions first? Email Jake, the founder: hello@ecomindex.ai

How we measure

Brand A: named in 74 of 178 completed API answersIllustration with made-up brands and counts. Each cell is one answer: 20 questions, asked 3 times each of 3 engines. The 2 hatched cells are answers that didn't complete, so they're left out of the count.

Twenty questions a buyer would ask

We read your store and the competitors you list at checkout, then write 20 questions a shopper would put to an AI assistant before buying. Jake reviews every one, and the set is frozen so a later run asks exactly the same questions.

Best for
“best paddle board for a beginner on flat water”
Use case
“what size wakeboard for a 180 lb rider”
Comparison
“merino or synthetic base layer for ski touring”
Price limit
“best espresso grinder under $300”
Alternatives
“brands like your top competitor, but cheaper”

Examples from different categories. Yours are written for your store.

These are answers from each engine's API, not a record of what every shopper sees in consumer apps, and answers change over time. The report says so, and shows every number as a count of real answers.

If you use Shopify's Agentic home or Merchant Center's AI insights, keep using them. The report adds the competitors named across up to three engines, the sources those answers cite, and the method behind every count.

Read the full method 

Done for you

Get on the shelf

$1,750 a month

3-month minimum, then month to month.

We do the work that gives AI engines something to quote: pages that answer buyers' questions, guides for the comparisons they ask about, and disclosed pitches to the sites the answers cite.

Email us about Get on the shelf

Or write to hello@ecomindex.ai. Jake replies with times to talk.

Each month, up to

Product or collection pages rewritten so AI can quote them
10
Comparison or buying-guide pages
2
Disclosed pitches to the sources AI cites, logged
10
Re‑runs of your 20 questions
Weekly
Search Console summary of the pages we changed
Monthly

In month 1 we agree your 20 questions and competitors, run a baseline report, and fix structured data and catalog feeds. Bought the $750 Playbook? It counts toward your first month if you start within 90 days.

How we work

  • A limited staff account: products, pages and the online store. Never orders, payments or customer data.
  • Nothing goes live on your store until you approve it, and a change log keeps every earlier version.
  • No posting as customers, no fake accounts and no vote manipulation.

What we never promise

We don't promise any AI mention, ranking, share or placement, coverage on another site, or traffic and sales. AI engines are run by other companies and change without notice. What we promise is the work listed here.

What happens next

  1. Today

    You check out

    Type your store URL and up to three competitors. A confirmation email follows, and you can reply within 24 hours to fix anything.

  2. Within 2 business days

    We run it

    We write your 20 questions, ask each one three times on each engine, and Jake checks every brand count and the fix plan.

  3. After delivery

    You ask

    Reply with any question about your report and Jake answers. Think something in it is wrong? We check it and correct it.

For the Playbook, you also add hello@ecomindex.ai to your Search Console as a restricted user, which reads your data and can't change it. The 2 business days start then.

Jake Ballard

Who does the work

Jake Ballard

Founder

Questions

Is this what shoppers see in ChatGPT?

Not exactly. The counts come from each engine's API with web search. Consumer apps can answer differently from one shopper to the next, so we don't claim to know what every shopper sees. The three ChatGPT app screenshots in each report show how the app answered on the day, labeled with the settings used.

Can you guarantee AI will name my brand?

No. Nobody controls what AI engines say. The retainer promises the work listed above, and every report shows the counts as they came back.

What does the $750 Playbook add?

Everything in the $99 report, plus what the answers repeat from your top two competitors' pages, the sites worth pitching and who to approach, your own Search Console data, checks on what AI crawlers, product pages and catalog feeds show, and a 90-day plan of 30 to 50 changes. Each change says where it goes, how long it takes and whether it's meant to move AI answers, Google or both.

Why does the Playbook need Search Console?

It's the only place your own Google data lives: the searches where your pages sit just off the top, and the pages shown often but rarely clicked. Restricted access lets us read it and never change anything, and you can remove it any time.

Why isn't Gemini included?

Google's terms for Gemini answers grounded in Google Search don't allow analyzing or reselling them, so we leave Gemini out. For Google, the retainer's Search Console summary includes AI Overviews and AI Mode impressions where Search Console reports them.

Which stores is it for?

Shopify brands that sell things people research before they buy. The store has to be live and public, so a password-protected or prelaunch store can't be checked. If we can't analyze your store, you get a full refund.

What do you need from me?

For the report, your store URL and up to three competitors, typed at checkout. No login and no app to install.

Who can see my report?

Anyone with its private link, and no one else. The page isn't listed anywhere, search engines are told not to index it, and the link keeps working after a refund.

Do the answers change?

Yes. Engines update their models and the web changes. Every report is dated and names the model and settings each engine used, so a later run can be compared with it.

Start with one question

When a buyer asks AI engines , do the answers name ?

Email us this question

Opens your email app with the question filled in.