Free snapshot

AI shopping recommendations, measured monthly

Which brands do AI assistants recommend to shoppers?

Shoppers now ask AI assistants what to buy. A few brands get named and most do not. Dopbot measures which ones, every month and by category, and shows brands where they stand.

Run a free snapshot Free · No card needed

From the indexMattressesWeek of Oct 19, 2026

One of the questions we asked AI assistants

“What is the best mattress to buy right now?”

Most recommended across all 10 sampled answers

  • 1Saatva5 of 10
  • 2Helix4 of 10
  • 3Brooklyn Bedding3 of 10
  • 3Leesa3 of 10
  • 3Nectar3 of 10

OpenAIPerplexity All 13 brands

  • Mattresses Saatva
  • Bed sheets L.L.Bean
  • Bath towels Frontgate, Pottery Barn and Threshold
  • Scented candles Bath & Body Works and Diptyque
  • Office chairs Steelcase
  • Cookware sets All-Clad and Tramontina
  • Chef knives Victorinox
  • Coffee grinders Baratza and Fellow
  • Mechanical keyboards Keychron
  • Carry-on luggage Briggs & Riley, Rimowa and Travelpro
  • Backpacks Aer, Bellroy, Eastpak, Everlane, Fjällräven, JanSport, MATEIN, ORTLIEB, Osprey, Peak Design, The North Face and Timbuk2
  • Water bottles Owala and YETI
  • Men's underwear SAXX
  • Sunglasses Ray-Ban

01 The Index Week of Oct 19, 2026

Latest results

Browse the index

A free, public ranking of the brands AI assistants recommend, by category. Each category page shows the full ranking, the sources the assistants retrieved and the questions we asked.

Mattresses

Memory foam, hybrid and latex mattresses sold online and in stores for adults.

Most recommended: Saatva50%
Recommended in 5 of 10 sampled answers

Cookware sets

Pots and pans sold as sets: stainless steel, nonstick, ceramic and cast iron.

Most recommended, tied: All-Clad and Tramontina60%
Each recommended in 6 of 10 sampled answers

Mechanical keyboards

Mechanical keyboards for work, gaming and typing enthusiasts.

Most recommended: Keychron80%
Recommended in 8 of 10 sampled answers

Carry-on luggage

Hardside and softside carry-on suitcases for air travel.

Most recommended, tied: Briggs & Riley, Rimowa and Travelpro40%
Each recommended in 4 of 10 sampled answers

Men's underwear

Boxer briefs, trunks and boxers for men.

Most recommended: SAXX60%
Recommended in 6 of 10 sampled answers

Dog beds

Orthopedic, bolster and chew-resistant beds for dogs.

Most recommended: Big Barker40%
Recommended in 4 of 10 sampled answers

Three markets

The same questions and method, asked for shoppers in each country: the United States, Mexico (questions in Spanish, prices in pesos) and Canada (prices in Canadian dollars).

United States (100) Índice México (30) Canada (30)

40 of 100 US shopping categories are updated monthly, on the first Monday; the other 60 are the October 2026 edition and are not updated. The index also covers México (in Spanish) and Canada, kept as the October 2026 edition and no longer updated.

Category pages across 3 markets
160
Answers in the latest US update
993
Shopper questions per category
5

02 How it works

We ask what shoppers ask, then count the answers

  1. Shopper questions

    Each category has five prompts, from best overall to a plain question about a buying situation that names no brand, written the way people type into an AI assistant.

  2. Sampled on a schedule

    Every prompt goes to AI engines through their developer APIs, with web search on and the location of the market's country: once a month for the 40 US categories that are still updated, on the first Monday. The other categories are the October 2026 edition.

  3. Brands counted

    We read each answer, extract the brands it recommends and the sources it retrieved, and publish how many answers recommend each brand. Read the methodology.

AI is the new shelf.

Shoppers ask, the assistant names a few brands, and the rest stay in the dark.

03 Free snapshot

Free snapshot

Enter your store domain. We ask AI assistants a set of shopper questions about your category and show how often they recommend your brand, who they recommend instead, and which sources they cite. Free, no card needed.

  • Free, no card needed
  • Results in about 1 to 2 minutes
  • A private results page that updates while it runs
What shoppers would call what you sell. Leave it empty and we work it out from your site.

Results take about 1 to 2 minutes. You get a private results page that updates while it runs. By submitting you agree to the terms; see the privacy policy.

04 The paid tracker

See where you stand, and what to change

The tracker runs about 25 prompts built from your own catalog every week and turns the answers into a plan.

  1. 01

    Shortlist score

    The share of sampled answers that recommend your brand, tracked week by week.

  2. 02

    Per-prompt winners

    For each question, which brands the assistants recommend instead of you.

  3. 03

    Cited sources

    The sites and pages the assistants cite when they answer shopper questions.

  4. 04

    Citation gaps

    Sources that back your competitors in answers where your brand is missing.

  5. 05

    Competitor comparison

    Up to 5 named competitors scored the same way as your brand.

  6. 06

    A ranked fix list

    Concrete changes in priority order, each tied to what the data shows.

Pricing

Brand$79/ month1 brand Agency$199/ month5 brands

See pricingPrices in USD · Cancel anytime

05 FAQ

Common questions

The methodology has the full detail: the questions, the engines, how answers are read and what the numbers can and cannot say.

Is this exactly what people see in consumer AI apps?
No. We sample the engines through their developer APIs with web search on and the location of the index's country (the United States, Mexico or Canada). Those answers approximate what people see in consumer apps but are not identical to them. The methodology explains the differences.
What does the free snapshot cost?
Nothing, and it needs no card. You enter a brand name, a store domain and an email address, and you get a private results page.
How often does the index update?
Once a month for the 40 US categories that are still updated, on the first Monday of the month. Runs start on Monday (UTC), and an update is labelled by the Monday of its week. The other US categories, México and Canada are the October 2026 edition: they stay published and are not updated.
How do you decide that a brand is recommended?
An AI model reads each answer and marks the brands the answer presents as picks, separately from brands mentioned only as a comparison. Only picks count, and a brand counts once per answer.
Can a brand pay to rank higher in the index?
No. The index is built from sampled answers only, and the paid tracker does not change it.
Are you affiliated with OpenAI, Perplexity or any AI provider?
No. Dopbot is independent and is not affiliated with, endorsed by or sponsored by any AI provider or any brand in the index.

Is your brand on the shortlist?

Run my free snapshot