Blog
AI shopping visibility: research and guides
Research from the Dopbot Index on which brands AI assistants recommend, and guides for brands that want to be on the shortlist.
Articles are drafted with AI tools from Dopbot Index data and public sources, and every number is checked against its source before it is published. How we write.
Latest
What 1,587 AI shopping answers show: the October 2026 Dopbot Index
Headline numbers from 1,587 AI shopping answers in 160 categories: sources, brands per answer, engine overlap and what the data cannot tell you.
Websites in AI shopping answers: OpenAI's links vs Perplexity's sources
The domains most often in OpenAI's links and Perplexity's retrieved sources for 1,587 shopping answers, in separate tables, plus source types per engine.
Wirecutter in AI shopping answers: Perplexity's sources vs OpenAI's links
A Wirecutter page was in 333 of 787 Perplexity answers' retrieved sources and in 0 of 800 OpenAI answers' links. What the two numbers mean and do not mean.
Research
What 1,587 AI shopping answers show: the October 2026 Dopbot Index
Headline numbers from 1,587 AI shopping answers in 160 categories: sources, brands per answer, engine overlap and what the data cannot tell you.
Websites in AI shopping answers: OpenAI's links vs Perplexity's sources
The domains most often in OpenAI's links and Perplexity's retrieved sources for 1,587 shopping answers, in separate tables, plus source types per engine.
Wirecutter in AI shopping answers: Perplexity's sources vs OpenAI's links
A Wirecutter page was in 333 of 787 Perplexity answers' retrieved sources and in 0 of 800 OpenAI answers' links. What the two numbers mean and do not mean.
OpenAI vs Perplexity: how much their shopping recommendations overlap
In 160 categories, OpenAI and Perplexity brand lists overlapped by 47.0% on average (October 2026). How it is measured, how it varies, and what to track.
Budget vs premium questions: how little the recommended brands overlap
In 160 categories, AI answers to a budget question and a premium question shared few brands: mean overlap 8.2%, and none at all in 70% of categories.
How many brands does an AI shopping answer recommend?
OpenAI and Perplexity API answers recommended about two brands each in the October 2026 edition, excluding one question type. The spread by engine and market.
Asking AI for alternatives to a brand: what the answers recommend
In October 2026, answers to "alternatives to [brand]" averaged 5.0 brands, and the named brand was recommended in 21 of 296. What that means for tracking.
How often Reddit and YouTube appear in AI shopping answers
Reddit was in 6 of 800 OpenAI answers' links and 26 of 787 Perplexity answers' retrieved sources; YouTube in 4 and 0. How that squares with public studies.
How many brands share a category in AI shopping answers?
A category's AI answers repeat a few brands and name many once. October 2026 numbers for 160 categories, a small-sample check, and what it means for a brand.
AI shopping answers in Mexico: what the Spanish-language sample shows
297 answers in Spanish across 30 Mexico categories: brands per answer, the share that recommended no brand, and the retailers in the source lists.
AI shopping answers in Canada: retailers and sources in the sample
Do AI answers to Canadian shopping questions point to Canadian stores? Source-list domains from 297 answers, and 24 same-name categories compared with the US.
When AI shopping answers recommend no brand at all
43 of 1,271 October 2026 AI shopping answers recommended no brand. Rates by engine, market and question type, what the data cannot explain, and what to do.
Explainers
What is AI shopping visibility? A definition and how to measure it
AI shopping visibility is how often assistants recommend a brand when asked what to buy. A precise definition, the measurement method and October 2026 numbers.
Glossary of AI search and shopping terms
Plain definitions of the terms behind AI search and shopping visibility, from AEO and GEO to grounding, RAG, source lists and the Wilson interval.
GEO vs AEO vs SEO for e-commerce brands: what actually differs
GEO, AEO and SEO overlap heavily. See what differs for an online store, a table of tactics by channel, and what October 2026 source data suggests.
How AI assistants choose which products to recommend
The documented steps from a shopping question to a named brand, what October 2026 source data shows for two engines at retrieval, and what it cannot explain.
Share of answer: how to compute it, and where it misleads
Share of answer is the percent of sampled AI answers that recommend a brand. Compute it with an interval, see what one answer is worth, and where it misleads.
How ChatGPT shopping works for brands
How products reach ChatGPT shopping results as of October 2026: feeds, Shopify Catalog, crawling and ads, and what API research can and cannot tell a brand.
AI crawlers and robots.txt: what each bot does and what to allow
Each AI crawler's documented role as of October 2026 (OpenAI, Perplexity, Google, Anthropic, Apple), plus CDN controls, trade-offs and how to test your store.
How Perplexity shopping works for brands
What Perplexity documents about product cards, Instant Buy and merchants as of October 2026, and what 787 of its API answers show about retrieved sources.
Google AI Mode and AI Overviews for product searches
What Google documents about AI Overviews, AI Mode and the Shopping Graph for products as of October 2026, how Merchant Center fits, and how to measure it.
Sources retrieved vs sources cited: reading AI citations correctly
What retrieved, linked and used sources mean, why a pooled citation rate describes no engine, and a checklist for reading AI visibility reports.
Why AI gives different answers to the same shopping question
Six sources of variation behind different brands for the same question, what October 2026 Dopbot data shows about short answer lists, and how to measure.
Playbooks
How to get your brand recommended by ChatGPT and other AI assistants
A prioritized playbook for AI shopping visibility, tied to October 2026 sample data and each assistant's documentation, with an evidence level on every step.
Holiday 2026: preparing for AI shopping questions before Black Friday
A dated checklist from Oct 30 to the December shipping cut-offs: make prices, stock, shipping and feeds agree before assistants answer holiday questions.
AI visibility for Shopify stores: a 30-day plan
A day-by-day Shopify plan with hours per task: baseline, bot access, product facts, markup and feeds, reviews, then a re-measure with honest limits.
Product pages AI assistants can use: an audit
A 12-check audit for product pages, mapped to the questions assistants get asked, with tests for any page and a minimum fact set from OpenAI's feed spec.
Structured data for AI shopping: which markup to add on Shopify
Which schema.org types Google documents for products, what Shopify's Liquid filter outputs, how to test it, and what is unproven about markup and AI answers.
Product feeds for AI shopping: Merchant Center, ChatGPT and Shopify
Which product feeds reach AI shopping as of October 2026: Google Merchant Center, OpenAI's feed, Shopify Catalog, and a field checklist to keep them aligned.
Buying guides and FAQs that AI assistants can quote
Write buying guides for "what should I look for" questions, using 6 of the 40 buying-advice prompts Dopbot asks, a passage template and a test.
When AI assistants get your prices or stock wrong
Where assistants get product facts, how to trace a wrong price or out-of-stock claim to its source, fix it, and confirm the fix with repeated sampling.
Can AI assistants read your Shopify store? A crawlability audit
A seven-step audit of what stands between AI search bots and a Shopify store, from robots.txt to logs, and what each test cannot prove.
Retail listings in AI shopping answers: Amazon, Target, Best Buy and Rufus
How often Amazon, Target and Best Buy pages sit in OpenAI's links and Perplexity's retrieved sources, what Rufus is now called, and a listing checklist.
Do reviews and ratings affect AI shopping recommendations?
Where review sites and retailer pages sit in each engine's sources, what platforms say about reviews, FTC rules for earning them, and how to test the effect.
Does your own website show up in AI shopping answers?
How often a recommended brand's own site was in OpenAI's links (46.6%) and Perplexity's retrieved sources (62.0%) in October 2026, and what the rule misses.
Comparison and alternatives pages: do they help in AI answers?
Should you publish "X vs Y" and "alternatives to X" pages? What the alternatives-question data shows, how to write a fair comparison, and the legal limits.
Getting into the review sites that appear in AI shopping answers
Which review domains show up in each engine's sources, how five say they pick products, and a step-by-step plan for pitching within disclosure rules.
When AI recommends your competitor instead: a diagnosis checklist
Check whether the gap is real, then locate it by question type, engine, sources, naming, retail presence and reviews, with October 2026 numbers for each check.
AI visibility for new brands with few reviews
Can a new brand with few reviews be recommended by AI assistants? What the long tail in our data shows, where to start, and how many answers to read.
When AI assistants mix up brand names, and how to fix it
How name variants split a brand's counts in AI answers, how the Dopbot index merges them, and how to audit and tighten your own brand name signals.
Measurement
How to measure your brand's AI visibility: a free manual method
A step-by-step manual method to measure how often AI assistants recommend your brand: question types, clean sessions, a spreadsheet layout and an interval.
How many AI answers you need before trusting a change
Is a rise from 3 to 5 recommendations real? Intervals, exact tests and sample sizes for 10, 20, 40 and 100 AI answers, computed by script, with examples.
API answers vs app answers: what sampling can and cannot tell you
What Perplexity and OpenAI document about API versus app behavior, what Dopbot's setup controls and misses, and a spot check to test any tracker.
Tracking traffic and sales from AI assistants: a GA4 and Shopify set-up
How to see visits and orders from AI assistants in GA4 and Shopify, what Google and Shopify document about AI sources, and what attribution cannot see.
Writing tracking prompts for AI visibility: five question types and the rules
Five question types with real index prompts, what each returned in October 2026, what changes in November and rules for neutral tracking prompts.
Running AI visibility experiments: change one thing, measure enough
How to test whether a change improved your AI visibility: baseline, one change, controls, timing, model updates, exact tests and a worked example.
Reading an AI visibility report: share, position, sources, intervals
A line-by-line walkthrough of a real Dopbot category page: what each count, rank, position and source column means, and the intervals the page leaves out.
Competitor benchmarking in AI answers: a method for small samples
How to compare your brand with competitors in AI answers: same prompts, each engine apart, paired counts, source gaps, and how lopsided a gap must be to count.
AI visibility reporting for agencies: structure, intervals, promises
A client report template for AI visibility: what to show, intervals for small samples, a monthly rhythm, what not to promise, and how to educate clients.
Choosing an AI visibility tracker, and whether you need one
When a manual check is enough, what it costs in hours, and 10 questions to ask any AI visibility tool before you pay, answered for Dopbot's own tracker too.
How we write
Who. The blog is written and published by Dopbot, the operator of dopbot.com. Dopbot is not affiliated with OpenAI, Perplexity, Google, Shopify or any company it writes about.
How. Articles are drafted with AI tools from the Dopbot Index data and from public sources. Every number and every outside fact in an article is recorded with its source: a figure in the index aggregates, a page on dopbot.com, or a public page with the passage that supports it. Software checks that each number matches its source and that each outside page holds its passage; a second AI model then reviews each article for accuracy and substance before it is published.
The data. Index numbers come from the providers' developer APIs, which approximate but are not the same as what people see in the apps. A data article names its engine, its edition and its denominator next to each number; the methodology says how answers are collected and where the numbers stop being reliable.
Interests. Dopbot sells an AI visibility tracker. Articles about measuring or choosing tools say so.
Corrections. If something is wrong, write to hello@dopbot.com. Corrected articles show the date of the correction.