How to Use ChatGPT for Amazon Product Research with Live AMZScout Data
20 min read
ChatGPT can research Amazon products with live data once it is connected to AMZScout MCP. In our test, the figures it returned matched AMZScout's own data for the same listings, with differences in sample size that we explain below!
We ran five prompts on one niche, cat water fountains on Amazon US, in ChatGPT Plus on October 23, 2026. Each section shows what ChatGPT returned and what we checked against AMZScout data. If you are new to MCP, start with our explainer on AMZScout Skill + MCP. The same prompt types, tested in Claude, are in our Claude version of this article.
What You Need Before You Start
A paid ChatGPT plan. AMZScout lists ChatGPT Plus and higher as supported.
Access to the AMZScout MCP so that the sign-in at the end of the setup works.
How to Connect AMZScout MCP to ChatGPT
Menu names can differ between ChatGPT accounts and versions; these are the ones we saw in October 2026.
Open Plugins, click Add, and choose Add custom MCP server.
Enter the name AMZScout Skill + MCP, the server URL https://chatbot.amzscout.net/mcp, and choose OAuth as the authentication.
Check the "I understand and want to continue" box, and click Create.
Sign in to AMZScout with the email you used to get access.
Start a new chat and type "How many AMZScout tokens do I have left?" to check the connection.
Start every prompt with a reminder: "Using the AMZScout Skill". AMZScout's guide says ChatGPT does not always apply the Skill by itself. We used the reminder in all five prompts below.
How We Ran the Test
Setting | Value |
|---|---|
Niche | Cat water fountain |
Marketplace | Amazon US |
Account | ChatGPT Plus, regular chat |
Date | October 23, 2026 |
Order | Five prompts in one chat, each starting with "Using the AMZScout Skill" |
Repeats | Prompts 1 and 4 re-run in new chats |
We pulled the same data directly from AMZScout and compared it with ChatGPT's figures. We did not cross-check the ceramic and cordless samples in prompt 5. The ceramic sample is marked where it appears.
Five Prompts, One Niche
1. Get a Market Snapshot
Prompt used in ChatGPT: Using the AMZScout Skill, give me a market snapshot for the cat water fountain niche on Amazon US — top products, revenue split, and who's dominant.
Metric | Value* |
|---|---|
Listings analyzed | 30 |
Est. monthly revenue | $22.83M |
Est. monthly units | 614,252 |
Median price | $37.49 |
Top-5 revenue share | 37.82% |
Veken | 51.08% of revenue, 13 listings |
PETLIBRO | 31.49% of revenue, 8 listings |
Neareal | 9.45% of revenue, 4 listings |
*All figures are accurate at the time of publishing and may change as the Amazon market evolves. This post is for informational purposes only and should not be considered a recommendation, business advice, or investment advice.
Two brands, Veken and PETLIBRO, hold 82.6% of the revenue in this sample.
ChatGPT also warned that the sample includes replacement filters and several variants of the same product, so the totals are directional and not a deduplicated market size. The data supports that three of the five top listings are Veken C1 variants with almost identical sales (38,290, 38,273, and 38,249 units).
Top 5 by revenue* | Price | Est. units | Est. revenue |
|---|---|---|---|
PETLIBRO Dockstream 2 | $79.99 | 26,210 | $2.10M |
PETLIBRO RFID Multi-Cat | $99.99 | 19,194 | $1.83M |
Veken C1 (B0FPFZ3C5K) | $44.99 | 38,290 | $1.69M |
Veken C1 (B0G34FPC49) | $39.99 | 38,273 | $1.51M |
Veken C1 (B0G34DKXPL) | $39.99 | 38,249 | $1.51M |
*All figures are accurate at the time of publishing and may change as the Amazon market evolves. This post is for informational purposes only and should not be considered a recommendation, business advice, or investment advice.
An earlier run of this prompt, in a longer chat, returned $22.32M and a 38.68% top-5 share from what looks like a different set of listings. The repeat in a clean chat produced the figures above.
2. Audit a Single Listing
Prompt used in ChatGPT: Using the AMZScout Skill, pull the full stats and sales history for ASIN B0FDKTYC1V on Amazon US. Is it a safe bet right now?
PETLIBRO Dockstream 2 | Value* |
|---|---|
Price | $79.99 |
Est. monthly sales | 26,210 units |
Est. monthly revenue | $2.10M |
Reviews | 20,474 (4.2 stars) |
Amazon FBA fee | $21.59 per unit |
Listing Quality Score | 74 |
First available | Aug 11, 2025 |
Variants | 7 |
*AMZScout data for ASIN B0FDKTYC1V, as returned by ChatGPT. All figures are accurate at the time of publishing and may change as the Amazon market evolves. This post is for informational purposes only and should not be considered a recommendation, business advice, or investment advice.
ChatGPT pulled the monthly history and found the turning point: sales rose from 8,110 units in April 2026 to 21,560 in May (+166%) and stayed above 24,000 from June through September. It marked October as an incomplete month (20,154 units as of October 23).
Its verdict was a conditional NO-GO for a new private-label seller until margins, customer complaints and ad costs are validated. It also pointed out that the $58.40 left after the FBA fee ($79.99 minus $21.59) is not profit.
AMZScout marks this listing as a 7-variant family whose sales estimate may cover all variants. From this data alone we cannot tell whether the May jump is demand growth or a change in how variants are grouped, so treat it as a lead to verify on the listing.
3. Compare Two Listings
Prompt used in ChatGPT: Using the AMZScout Skill, compare B0FDKTYC1V and B0C8MHHVPC on Amazon US — which one is actually winning right now?
Metric | PETLIBRO Dockstream 2* | Neareal 108oz* |
|---|---|---|
Price | $79.99 | $38.99 |
Est. monthly units | 26,210 | 17,893 |
Est. monthly revenue | $2.10M | $670K |
Reviews | 20,474 | 12,844 |
Rating | 4.2 | 4.4 |
Pet Supplies rank | #175 | #372 |
*All figures are accurate at the time of publishing and may change as the Amazon market evolves. This post is for informational purposes only and should not be considered a recommendation, business advice, or investment advice.
ChatGPT named PETLIBRO the winner: 46% more units and 3.13 times the revenue, at more than twice the price. It gave Neareal the edge on rating (4.4 against 4.2 stars) and price ($41 lower) and noted that PETLIBRO's seven variants may inflate its estimate.
The answer rests on current estimates only. ChatGPT did not show how each listing's sales moved over time, while the comparison in our Claude test included a trend row. In AMZScout's data, Neareal's monthly sales went from 10,261 in April 2026 to 16,535 in September, and PETLIBRO's from 8,110 to 26,108. If trend matters for your decision, ask for it in the prompt, for example, by adding "and show the monthly sales trend for each".
4. Pull a Competitor's Keywords
Prompt used in ChatGPT: Using the AMZScout Skill, what keywords does B0C8MHHVPC rank for on Amazon US, and is there anything it's leaving on the table?
Keyword | Monthly searches | Organic rank | CPC |
|---|---|---|---|
cat water fountain | 49,500 | #21 | $0.54 |
cat water fountain stainless steel | 18,100 | #12 | $0.72 |
stainless steel cat water fountain | 18,100 | #9 | $0.72 |
pet water fountain | 9,900 | #20 | $0.52 |
cat water dispenser | 5,400 | #64 | $0.39 |
pet water dispenser | 2,400 | #37 | $0.33 |
ChatGPT received 100 keywords for the ASIN and showed only the commercially relevant ones. The raw AMZScout list also contains broad terms such as "stainless steel" (165,000 searches, rank #69) and "pet supplies" (90,500 searches, rank #129), and even the competitor brand "petlibro" (49,500 searches, rank #196). Its main recommendation was to move "cat water fountain" from #21 into the top 10 and to defend the #9 and #12 stainless steel positions.
It also checked the listing itself. It noticed the 28 dB pump and suggested "quiet" and "silent" searches, and it advised against targeting "cordless", "wireless", "filterless", and "ceramic", because the listing does not offer those features. The listing data in AMZScout confirms a 28 dB pump and a stainless steel body and none of those four features.
It warned that similar keywords share demand: "cat water fountain stainless steel" and "stainless steel cat water fountain" both show 18,100 searches and should not be added together. It also noted that AMZScout's keyword trend (-18% for "cat water fountain", -33% for "cat water fountain stainless steel") describes search demand, not the product's own sales.
We repeated this prompt in a new chat. The keyword rows were identical, and the written recommendations were not.
5. Ask for a Verdict
Prompt used in ChatGPT: Using the AMZScout Skill, I'm thinking about launching a private label cat water fountain on Amazon US. Is this niche worth entering right now, or is it too saturated? What would I do differently from the top sellers?
Sample | Result |
|---|---|
Broad niche, 49 listings | $25.54M revenue, 683K units, median price $39.99, median reviews 5,220 |
Veken + PETLIBRO share | 75.6% of sampled revenue |
Ceramic fountains, 20 listings | $1.46M revenue, median reviews 1,168, median price $38.99, 11 brands |
ChatGPT gave a Conditional GO. Its advice: do not launch another $30 to $40 stainless steel fountain, and test an easy-clean ceramic or a reliable cordless model instead. To compare segments, it ran two more niche analyses on its own (ceramic and cordless, 20 listings each) before answering.
It added a margin example for a $59.99 fountain: a $9.99 contribution per unit, a 16.7% margin. The inputs were its own planning assumptions ($16 landed cost, $16 Amazon fees, 25% ACOS, and $3 other costs), and it labeled them illustrative. Replace them with your supplier quote and fee estimate before you rely on the result.
Prompt 2 returned a conditional NO-GO and prompt 5 a Conditional GO. They answer different questions: whether PETLIBRO's listing is a safe bet to copy and whether the niche is worth entering with a different product.
What to Double-Check Before You Act on the Answer
Sample size. The same niche gives different totals depending on how many listings are pulled, so quote the number of listings next to every total.
Listings pulled | Est. monthly revenue | Top-5 share |
|---|---|---|
15 | $15.95M | 54.1% |
30 | $22.83M | 37.82% |
50 | $26.16M | 33% |
Variants and accessories. In the 30-listing sample, four replacement-filter listings account for $2.19M, or 9.6% of the revenue (our calculation). ChatGPT flagged variants and accessories on its own in prompts 1 and 5.
Incomplete months. ChatGPT marked October 2026 as incomplete: 20,154 units in the history against a 26,210 headline estimate.
Planning assumptions. Margin figures in verdict-style answers come from ChatGPT's assumptions, not from AMZScout data.
ChatGPT vs Claude: What Differed
Setup. ChatGPT needs Developer mode and a custom MCP server entry. In Claude we connected the AMZScout MCP from the connectors directory, as described in our Claude article.
Skill reminder. We started every ChatGPT prompt with "Using the AMZScout Skill", as the AMZScout guide advises, and did not test without it.
Format. ChatGPT answered each prompt with a structured report: charts, labeled verdicts such as "Conditional GO", and a suggested next step.
Extra searches. In prompts 4 and 5 ChatGPT ran additional AMZScout searches on its own: niche-wide keywords, then ceramic and cordless niches.
Trend. The comparison in prompt 3 used current estimates only.
The Bottom Line
With AMZScout MCP connected, ChatGPT answered each of our five research prompts in 6 to 44 seconds, and the figures it returned for the same listings matched AMZScout's data. Use it for fast first-pass research, and verify totals, trends and margins before you place an order.
For a quick check on a single product page while you browse Amazon, the PRO AI Extension shows the numbers directly on the page.
Connect AMZScout MCP to ChatGPT and run the five prompts on your own niche.
FAQs
Does ChatGPT Plus support MCP?
Yes, for AMZScout MCP. AMZScout lists ChatGPT Plus and higher as supported. Adding the connector also requires Developer mode, according to AMZScout's guide.
What is MCP vs API?
MCP (Model Context Protocol) is an open standard that lets an AI assistant call outside tools and data sources during a chat. An API is the interface a developer codes against; MCP lets an assistant use such a tool without custom code. The full explanation is in our first article on AMZScout Skill + MCP.
Do I need to type "Using the AMZScout Skill" every time?
AMZScout's guide says ChatGPT does not always apply the Skill automatically, so add the reminder to each prompt where you want it used. We used it in all five prompts and did not test without it.
Does AMZScout MCP work in a temporary ChatGPT chat?
In our test it did not. Use a regular chat.
Are the numbers exact?
No. AMZScout's sales and revenue figures are estimates, and ChatGPT labeled them that way in all five answers.
