Track How Often ChatGPT Recommends Your Products
ChatGPT now influences more purchase decisions than display advertising. If you do not know whether it recommends your products — and why it recommends competitors instead — you are flying blind on your fastest-growing acquisition channel.
ChatGPT has become a shopping engine — without calling itself one
Google Search built an entire ecosystem of shopping ads, comparison pages, and product carousels to capture purchase intent. ChatGPT did none of that — and yet it now influences more e-commerce purchase decisions than Google Shopping among the 18–35 demographic. The reason is conversational discovery.
When a shopper types “what's the best yoga mat for hot yoga if you sweat a lot,” they are not looking for ten blue links. They are asking for a recommendation from something that seems to understand context. ChatGPT answers with specific product names, specific reasons for each recommendation, and increasingly, direct links to purchase. The entire decision arc — from need identification to specific product selection — happens in a single conversation, before the shopper has visited any e-commerce site.
This shift fundamentally changes the economics of discovery. In traditional search, you can buy your way into visibility with ads. In ChatGPT, recommendations are based on what the model believes to be the best answer to the question — informed by training data, review site coverage, and structured product information. There is no sponsored placement. There is no bidding. You are either recommended because your product data supports a confident recommendation, or you are not.
Meridian Skincare, a DTC brand selling clean beauty products, spent $8,400/month on Google Shopping with a 2.1x ROAS. When they connected AISeen, they discovered that ChatGPT was recommending their top competitor for every relevant query about clean moisturizers for sensitive skin — while mentioning Meridian Skincare in fewer than 5% of those same queries. The competitor spent nothing on ads. They had better- structured ingredient lists, more specific clinical claims, and articles on five dermatology-adjacent sites that ChatGPT cited as sources. No ad budget could fix that. Product data improvements could.
Up from 34% in 2024. ChatGPT is the most-used AI shopping assistant, cited by 67% of AI-assisted shoppers.
Shoppers who arrive from a ChatGPT recommendation convert at 3.4× the rate of cold search traffic.
Unlike Google or Amazon, there is no way to pay for visibility in ChatGPT. Recommendations are purely data-driven.
Stores implementing AISeen's top recommendations typically see measurable ChatGPT mention rate increase in 2–4 weeks.
How AISeen tracks your ChatGPT visibility
Accurate ChatGPT tracking requires running real queries with real product context — not searching for your brand name, but sending the exact buying questions your customers are asking. Here is the precise methodology:
Query generation from your catalog
AISeen analyzes your product catalog and generates 100–2,000 natural-language shopping queries across nine categories: comparison, gift, problem-solving, feature-specific, budget, use-case, sustainability, alternative-seeking, and demographic. These are the specific questions your potential customers are asking ChatGPT today.
Daily query execution
Each query is sent to the OpenAI API daily using a consistent shopping assistant system prompt. We use gpt-4o and lock to a specific model version for comparability. Every response is stored with full text, timestamp, and the model version used.
Mention detection and position extraction
AISeen parses each ChatGPT response for brand and product mentions, extracts the position (1st recommended, 2nd recommended, etc.), the specific claims made, and any links or citations included. Sentiment is scored on a 3-point scale: recommend, neutral mention, or recommend-against.
Daily visibility score update
Your ChatGPT visibility score updates every morning: mention rate (% of tracked queries where you appear), average position, share of voice versus competitors, and sentiment ratio. The score is comparable over time and across AI platforms.
What ChatGPT visibility monitoring reveals about your store
Which queries trigger your brand — and which do not
AISeen shows you the exact distribution of your mentions across query types. Most stores discover they are mentioned frequently for one query type (often feature- specific queries like “yoga mat with alignment lines”) and nearly invisible for others (comparison queries like “best yoga mat vs. Manduka PRO”). This tells you exactly what content and schema work to do next.
How ChatGPT describes your brand versus competitors
When ChatGPT recommends your brand, AISeen captures the exact language used. When it recommends a competitor, it captures that too. Side-by-side, the comparison is often revealing: a competitor is described with five specific attribute claims (“6mm natural rubber, 72-inch, non-slip base, machine washable, 4.8 stars across 4,200 reviews”) while your brand is described vaguely (“a well-reviewed option for serious practitioners”). The gap tells you exactly what to add to your product data.
Which competitor queries you could win
AISeen identifies queries where a competitor appears in ChatGPT responses but no product in the query is clearly dominant — meaning the recommendation is competitive and winnable. For these queries, the attribute gap analysis shows the specific claims your competitor is getting credit for and you are not. These are your highest-leverage optimization targets.
Which sources ChatGPT cites about your category
When ChatGPT includes citations (in browsing-enabled mode), AISeen logs which third-party sites are referenced. If Wirecutter, The Strategist, or a specialist publication appears in citations for your category and does not mention your brand, that is a coverage gap. Getting featured on those specific sites is worth more for ChatGPT visibility than almost any other single action.
How to get your products recommended by ChatGPT more often
Make your product descriptions machine-readable
ChatGPT generates recommendations by extracting specific, citable attributes from product data. A description that says 'premium quality' cannot be cited. A description that says '6mm natural rubber base, 72 × 24 inch, 4.8 lbs, machine washable, closed-cell foam top surface that resists moisture absorption' gives ChatGPT six specific claims to use in a recommendation. AISeen identifies which attributes are missing from your descriptions by analyzing what ChatGPT cites for competitor products in your category.
Get coverage on the sites ChatGPT trusts
In browsing mode, ChatGPT cites specific review sites and editorial publications. For kitchen tools, it frequently cites Wirecutter, Serious Eats, and America's Test Kitchen. For outdoor gear, it cites REI's guides, Switchback Travel, and OutdoorGearLab. For skincare, it cites Allure, Byrdie, and Healthline. AISeen surfaces which sites appear in ChatGPT citations for your category and flags which ones do not mention your brand — the exact list of outreach targets.
Add complete Product schema
ChatGPT's training data includes structured data from web crawls. A product page with complete Schema.org/Product markup — including brand, material, dimensions, reviewCount, and aggregateRating — is parsed more reliably than one where attributes are buried in flowing paragraph text. AISeen audits your schema coverage and generates the exact JSON-LD blocks to add.
Create comparison content
ChatGPT frequently sources recommendation content from comparison articles — pages that explicitly compare two or more products on specific attributes. Publishing a detailed comparison of your product versus the category leader, with specific attribute tables and an honest assessment of trade-offs, gives ChatGPT exactly the kind of content it needs to cite you in comparison queries — the highest-intent query type in most categories.
How Meridian Skincare went from 5% to 58% ChatGPT mention rate in 8 weeks
Meridian Skincare connected AISeen after noticing that their competitors were being cited in ChatGPT conversations while their brand was invisible despite comparable product quality and a strong review profile on their own site.
AISeen identified four issues. First, their product descriptions used qualitative language (“clean, gentle, effective”) without specific claims (no ingredient percentages, no clinical testing references, no dermatologist testing status). Second, they had no Product schema on any product page. Third, none of the editorial skincare sites that ChatGPT cited in their category (Byrdie, Healthline, Allure) mentioned their brand. Fourth, their best-selling moisturizer had a 4.9-star rating on their own site but no indexed reviews on third-party sites — so ChatGPT had no external validation to cite.
The fix plan: rewrite 12 product descriptions with specific claims (active ingredient percentages, hypoallergenic certification, dermatologist-tested status, fragrance-free certification); add Product schema to all 34 SKUs; and pitch three editorial contacts identified by AISeen as high-citation sources in the clean beauty category.
Eight weeks after implementing the fixes, Meridian Skincare's ChatGPT mention rate rose from 5% to 58% on their 150 tracked queries. Two of the editorial pitches resulted in inclusion in published roundups. Revenue from AI-referred traffic increased by $6,800 in the following month.
Find out your ChatGPT visibility score today
Run a free audit with your store URL. Get your ChatGPT mention rate, your top missed query, and your single highest-impact fix in 90 seconds. No account required.