It is Perplexity's free programme for retailers to share product data with its shopping answers. A small store can apply, although the launch post pitched it at large retailers and Perplexity decides who gets in. Guides say you must sell and ship to the United States; the form asks for brand, website, name, email and category, then a product feed.
One honest warning before the detail. Beyond two launch posts, Perplexity publishes no eligibility rules, approval timeline or feed specification, so parts of what follows rest on third-party guides and a feed-tool vendor's help page. Treat anything below marked "guides say" as a lead to check, not a rule. The preparation work is worth doing either way, because clean product data helps every shopping channel.
What the programme is and what it is not
Perplexity launched shopping in its answer engine in November 2024, with one-click Buy with Pro checkout for its Pro subscribers and a photo search called Snap to Shop, and announced the Merchant Program in the same post as the way for retailers to share their product specs. The post says the programme is free for merchants and lists four benefits: a better chance of being a recommended product, payment integrations, free API access and a dashboard of search and shopping trends. A year later, in November 2025, Perplexity made shopping free for all its users in its launch market, with instant checkout through PayPal, and said retailers stay the merchant of record.
Three things it is not. It is not a shop: customers still buy on your site or through a checkout route Perplexity offers, and you stay the seller. It is not a ranking guarantee, since Perplexity has not published how it orders products. And it is not the only way to appear: Perplexity's launch post said product results draw on Shopify stores that ship to its launch market, and that shoppers are sent to the merchant's site when its own checkout is not available, so joining is about completeness and control, not about being allowed in at all.
Three levels of participation
One AI-shopping vendor's guide separates a mention in an answer, inclusion in the product catalogue and integrated checkout as different things. That is a useful way to ask the right question of Perplexity or your platform: which level am I applying for, and what does it need from me? An interest form is not an approved integration, so do not plan around the first two weeks as if it were.
The Complete Excel Course with AI Integration
From formulas and pivot tables to Power Query, DAX, VBA and AI with Copilot, ChatGPT and Claude — hands-on from the first lesson.
By Sagnik Bhattacharya and Virginia Thorn, Primetime Emmy Award winner
Buy nowWho can join: what guides say and what to check
| Question | What the guides say | What to check yourself |
|---|---|---|
| Cost | Free, no commission | Written terms; payment processing or fulfilment costs elsewhere |
| Eligibility | Sell and ship to the United States; or use Shopify or BigCommerce; or integrate PayPal | Current wording on Perplexity's page and your platform's settings |
| How to apply | Short form: brand, website, name, email, category | That the form still accepts applications |
| Approval | No published timeline; acceptance comes first and Perplexity sets the timing | Ask in writing; do not promise a launch date |
| Feed format | No public specification; Perplexity assigns the delivery method at onboarding (feed tools offer a pull URL, FTP or SFTP) | Field names and delivery method from Perplexity or your integration partner |
| Small stores | Perplexity's 2024 launch post invited large retailers | Apply anyway if you are eligible; the form is short |
The silence on approval is itself useful: nobody outside Perplexity knows, and some of the guides are written by agencies who sell help with the process. Take the cheap route first. Read the official page, then your platform's app or settings page, then ask.
Shopify stores
Shopify merchants may already be partly covered. Perplexity's launch post said its product results draw on Shopify stores that sell and ship to the United States. Separately, Shopify's agentic storefronts are active by default for eligible stores, but its help page names ChatGPT, Google AI Mode and Gemini, Microsoft Copilot and Meta as the channels, not Perplexity, so do not assume that switch covers Perplexity. Shopify also has an Agentic plan with no monthly fee for stores that keep another platform. Open your Shopify admin and read what is already switched on, then decide whether to apply separately. Catalogue access does not prove that each item is included or recommended, so check a few products afterwards. For the full picture, Shopify Agentic Storefronts explained sets out what you still control.
What to submit: a feed a small store can build in an afternoon
A product feed is a spreadsheet with one row per product (or per variant) and a column for each fact a shopping channel wants. If you use Shopify, WooCommerce or another platform, you may export most of it. If not, a spreadsheet will do. Perplexity's field list is not public as far as I could verify, and one guide calls its own checklist "a preparation guide, not an official Perplexity feed specification". The safe approach is to build the standard set that Google Merchant Center uses, since most platforms and feed tools already produce it, and then adjust once Perplexity or your integration partner confirms field names.
The core fields, filled in for one product
id: KW-0042
title: Stainless Steel Kitchen Scissors, 21 cm, Dishwasher Safe
description: Heavy-duty 21 cm scissors with serrated blade for herbs and poultry. Comes apart for cleaning. Stainless steel, black handles.
link: https://example-store.test/products/kitchen-scissors
image_link: https://example-store.test/images/kw-0042-front.jpg
price: 14.99 USD
availability: in stock
brand: [your brand]
condition: new
gtin: [only if one has been assigned; leave blank if not]
Two rules save you from trouble. Include a GTIN (the barcode number) only when one has actually been assigned to the product; never invent one, because a wrong code can get a listing rejected or suspended on other channels. The tutorial on missing GTINs in Google Merchant Center explains what AI can and cannot fill. And keep price and availability identical to what your product page shows, since a mismatch is the quickest way to lose trust with a shopping system.
Using a chat assistant to prepare the feed
This is the job where AI helps most. Paste ten rows of your product export, without any customer data, and ask for a check.
Here are ten rows of a product feed. For each row, list: titles over 80 characters or without a brand or product type; descriptions under 100 characters or that repeat the title; any missing field from this list: id, title, description, link, image_link, price, availability, brand, condition.
Do not invent identifiers, sizes or materials. If a fact is missing, write [MISSING].
Then suggest an improved title only using facts already in the row.
An illustrative result: row 4's title is "Scissors" and gets flagged, with the suggestion "[brand] Kitchen Scissors, Stainless Steel" using only what was in the row, and "[MISSING] size" because the size is in the image, not the data. The assistant did not guess a length. That is the behaviour you want, and it is why the [MISSING] rule exists. For title work on a larger scale, see rewriting Google Shopping titles with AI safely.
Should a small store join? A sum with illustrative numbers
The programme is free, so the cost is your time and a share of your attention. Run the sum before you decide how much effort to spend. The store below is invented: 40 products, an average order of $45 with $16 of contribution after product cost and fees, a part-timer valued at $30 an hour.
| Item | Time | Cost at $30 an hour |
|---|---|---|
| Read the programme page, apply, confirm Shopify settings | 1 hour | $30 |
| Clean titles and descriptions for 40 products with an assistant | 4 hours | $120 |
| Fix identifiers, images and stock accuracy | 2 hours | $60 |
| Monthly upkeep (price and stock checks, reading referrals) | 1 hour a month, 12 hours a year | $360 a year |
| First-year total | 19 hours | $570 |
At $16 of contribution an order, the first year breaks even at $570 divided by $16, about 36 orders, or three a month, attributed to Perplexity. Whether you can win that depends on your category. A store selling items people ask about in conversation, such as gifts, gadgets and kitchenware, has a plausible case. A store selling made-to-order bespoke items, where shoppers rarely compare products by chat, has less. Most of the $570 also improves your listings everywhere else, so the real cost is less than it looks. I would not count it as a pure cost.
Join now, prepare and wait, or skip
- Join now if you ship to the eligible market, your product data is already clean, and your platform makes it a switch rather than a project.
- Prepare and wait if your data is messy. Fix it, because every channel rewards it, and apply once the page confirms what it needs.
- Skip for now if you do not ship where the programme operates, or you sell bespoke work with no comparable products.
Making it measurable before you join
Do this first, so you can tell afterwards whether it did anything. In your analytics, create a source filter for Perplexity referrals: the referrer domain, and any tracking parameter on the links. Guides say links from Perplexity carry tracking parameters, so look for the pattern on your first visits. Record the last four weeks of orders from every source as your baseline. Then check weekly. The method is the same as for other AI tools, and tracking ChatGPT traffic to your store walks through the setup.
What a result looks like, illustratively
Say you join in January and by March your analytics show 90 visits from Perplexity and 6 orders. Six orders at $16 of contribution is $96 a month on a $570 first-year investment, so a 7-month payback if the pace holds. If instead you see 12 visits and no orders, do not keep tinkering with the feed. Look at whether the listings that appeared were your priciest or your most generic products, and weigh the effort against other channels.
Where programme details can mislead a small store
Reading a marketing guide as a rulebook
Several guides quote percentage lifts in appearances or sales that I could not trace to any published source. Ignore them. A tool vendor or agency has a reason to make a free programme sound urgent, and Perplexity has not published figures for merchants. The only evidence that counts is your own referral data.
Letting a feed drift from the page
A feed exported once and never refreshed shows old prices and discontinued items. If a shopper clicks through to a different price or "sold out", you have spent their trust. Schedule the export or let the platform sync it, and review monthly.
Assuming it replaces your other channels
Search and shopping behaviour changes. For a store with stable Google Shopping traffic, fixing Merchant Center problems for AI Mode shopping is likely to matter more in the near term, and the same clean data serves both. Treat AI shopping channels as additional, and measure each separately.
A product-data audit, five rows before and after
Most of the work in joining any shopping channel is fixing the data you already have. Here are five rows from the invented kitchenware store, as exported and as improved. All of it is illustrative, and every improved value uses only facts that were in the original row or on the product page.
| Row | Title as exported | Problem | Improved title |
|---|---|---|---|
| 1 | Scissors | No brand, material or size | [brand] Kitchen Scissors, Stainless Steel, 21 cm |
| 2 | SALE!!! Chopping board NEW | Promotional words, no material | [brand] Bamboo Chopping Board, 40 x 28 cm |
| 3 | Pan 24 | Unclear size and type | [brand] Non-Stick Frying Pan, 24 cm |
| 4 | Set of 3 bowls red | Colour and material missing from the start | [brand] Mixing Bowl Set of 3, Red Enamel |
| 5 | Peeler | Too short to match a search | [brand] Swivel Vegetable Peeler, Stainless Steel Blade |
Notice what the improved titles do: brand first, then product type, then the details a person would say out loud when asking an assistant for the item. A shopper who asks for "a 24 cm non-stick frying pan" is matched by row 3 and not by "Pan 24". Check each result against the product page before you accept it. If the assistant adds "oven-safe" and you have not said so, delete it. Claims about safety, materials or certifications are the ones that cause returns and complaints.
Variants and the rows people forget
A product with three sizes and four colours is twelve rows in most feeds, each with its own id, price, stock and image. Mistakes cluster here: the red one shows the blue photo, or a sold-out size remains "in stock". Sort the sheet by parent product and read each group once. In the invented store, a 90-minute pass over the 11 products that have variants turned up four mismatched photos and two sizes that were out of stock but still listed. Those errors would have sent shoppers to a page that did not match what they were shown.
Writing descriptions that answer what shoppers ask
Product cards show a title, a price and a line or two from the description. The rest feeds the answer the assistant gives. A description that says "great quality, must have" gives it nothing to quote. One that names the size, the material, who it suits and how to care for it gives several facts. Here is a prompt that keeps the assistant honest.
Rewrite this product description in 60 to 90 words. Use only the facts below. Include: what it is, size, material, who it suits, how to clean it.
Facts: stainless steel kitchen scissors, 21 cm, serrated blade, comes apart for cleaning, dishwasher safe, black handles.
Do not say anything about safety, certifications, warranty or "best". If a fact is missing, write [MISSING].
An illustrative result: "Heavy-duty kitchen scissors in stainless steel, 21 cm long, with a serrated blade that grips herbs and poultry. The blades come apart for cleaning and the scissors are dishwasher safe. Black handles. A practical choice for cooks who want one tool for trimming and snipping. [MISSING] warranty." The assistant followed the rules, but "grips herbs and poultry" is an inference from "serrated". Is it true for your scissors? Only you know, so decide whether to keep it. For a longer method, see writing Shopify product descriptions with AI.
Reviews and what to do about them
Guides list reviews among the data shared. Reviews are the one part of a feed you cannot write yourself and should not try to. Do not pay for them or write them, and do not copy reviews between products. If your store has few reviews, a polite post-delivery email that asks for one honest review is the legitimate route; the number grows slowly. A new store with ten honest reviews is better placed than one with fifty that look bought.
The four weeks after you apply
An application is a start, not a result. This plan keeps you moving without waiting for a reply.
- Week 1: apply, or switch on the platform route. Record the baseline of orders by source. Export the feed and run the audit above on the ten best-selling products.
- Week 2: fix the remaining products in batches of ten. Check stock and price against the live pages.
- Week 3: search for five of your products on Perplexity by describing them as a shopper would. Note whether they appear, with what title and price. If they do not, check eligibility and ask your platform or Perplexity.
- Week 4: review referrals and orders. Apply the decision rule from earlier. If nothing has arrived, set a 12-week review and move on to a channel with clearer demand.
If you hear nothing, or are told no
One feed-tool help page says programme eligibility decisions are made entirely by Perplexity and appeals go to Perplexity directly. That is a sensible expectation. If you are declined or hear nothing, do not pay a third party to chase it. Your cleaned data is still useful for Google, your own site search and any other shopping channel, so the time is not wasted. Reapply after a few months if your eligibility changes, for instance if you start shipping to the eligible market.
How a small store should think about the risk
The risks are small because the programme is free, but they exist. Your price and stock are shown in a channel you do not control, so errors are public. A product you stop selling can linger in an answer after you delist it. And you take on a small admin habit that someone must own. Name one person, set a monthly reminder, and write one line in your risk list. Whether a paid Perplexity plan is worth it for your own research is a separate question: the merchant side is free and does not require you to subscribe.
A one-page checklist to copy
PERPLEXITY MERCHANT PROGRAM CHECK (store: ________) Date: ______
[ ] Read Perplexity's merchant page: eligibility wording, fee terms, form status
[ ] We ship to the eligible market: yes / no
[ ] Platform integration available and switched on: Shopify / other / none
[ ] Application sent on ______ with brand, website, contact, category
[ ] Feed has: id, title, description, link, image_link, price, availability, brand, condition
[ ] GTIN present only where assigned
[ ] Prices and stock match the product pages
[ ] Analytics source filter created; baseline orders recorded: ______ a month
[ ] Review date set (4 weeks, then 12 weeks): ______
[ ] Decision rule: keep if orders attributed >= ______ a month, otherwise stop
If you would like another person to look at your feed and your numbers, that is part of what my AI implementation consultation covers, and you can ask about it here.
Questions small stores ask about the programme
Does the Perplexity Merchant Program cost anything?
Perplexity's launch post says the programme is free for merchants, and guides add that there is no listing fee or commission. The application form and Perplexity's reply are where to confirm that today, along with any payment-processing or fulfilment costs that sit outside the programme. Get the terms in writing before you assume your total cost of selling is zero.
Do I need a Shopify store to join?
No. Guides describe a short application form for any eligible merchant, and say Shopify and BigCommerce stores can come in through those platforms' integrations instead. Which route applies to your store depends on your platform and where you ship, so check Perplexity's page and your platform's settings.
Will joining guarantee my products appear in answers?
No. Joining gives the system product data to work with, but Perplexity does not publish a ranking method. Complete listings, clear titles, accurate price and stock and good reviews improve your chances, and nothing guarantees placement.
What if my store ships outside the United States only?
Guides give selling and shipping to the United States as the main condition, so a store that does not ship there may be ineligible for now. Check Perplexity's current eligibility wording, as programmes widen and change.
How do I know whether it sent me any customers?
Guides say links from Perplexity carry tracking parameters, so look for them in your analytics referrals and orders. Set up a simple filter for the source before you join, so you have a baseline of zero to compare against.
Further reads
- AI Product Recommendations on Shopify: Built-In vs Paid Apps — Compare built-in and paid recommendation tools.
- How to Sync Stock Across Shopify, Amazon and eBay Automatically — Keep stock accurate across every channel you use.
- Is Perplexity Pro Worth Paying For in a Small Business? — Decide whether the paid Perplexity plan earns its fee.
- AI Marketing Tools for Shopify: Which App for Which Job — Pick marketing apps for the jobs around your listings.
- Perplexity vs ChatGPT for Business Research — Compare the two assistants for business research.
- How to Use AI in a Small Business: Perplexity for Research Jobs — Use Perplexity for the research jobs in your own business.
- Google Merchant Center Image Requirements for AI-Generated Product Photos — Check the image rules before using AI product photos.
- Help ChatGPT Discover Your Products: A Checklist for Small Shops — A grouped checklist for small online shops, with a scored audit of twenty products, test prompts and the feed fields OpenAI documents.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: Perplexity's announcements launching shopping and the Merchant Program (18 November 2024) and its PayPal checkout (25 November 2025); GoDataFeed's Perplexity feed template help article; WebFX and Alhena guides to the programme; Shopify's pricing page and help pages on agentic storefronts and its product catalogue; all read November 2026. Perplexity publishes no eligibility rules or feed specification beyond these, so confirm details when you apply.