← All use cases · Shops & e-commerce
Count the online orders that come from ChatGPT using Shopify's visit records, then fix the pages ChatGPT sends shoppers to
José Pérez Sánchez, owner of Persa Flores, a family florist in Albox, Almería, Spain (open since 1975, delivers across mainland Spain), on a data page on the shop's own website
The shop reviewed every online order from 1 July to 27 October 2026 and checked the visit Shopify records before each purchase. If that visit came from ChatGPT (links tagged utm_source=chatgpt.com), the order counted as a ChatGPT order. It found most of those shoppers landed on its existing city delivery pages, and published a six-point checklist: keep facts consistent across the site, update older pages, put the address, phone and delivery area in every page's structured data in every language, let ChatGPT's crawler read the delivery and returns pages, show the full price including delivery, and keep options few with a price for each.
Medium · or we set it up for $197
Shopify (order visit and landing-page records), ChatGPT (where shoppers found the shop)
The problem
Shoppers were starting to ask ChatGPT how to send flowers to a town, and the shop's standard referrer reports showed almost none of that traffic.
What happened
ChatGPT's share of online orders rose from 4% in July to 23% in October (to the 27th); from September to 27 October it was 1 in 5 orders and 22% of online revenue. 100% of ChatGPT orders came from first-time customers, and 84% of those customers landed on a city page. The owner told Practical Ecommerce that only about one in eight of these orders also showed ChatGPT as the referrer, so checking referrers alone misses most of it. Self-reported; the shop says the real share may be a little higher because Shopify records no source for some orders.
Quoted in the source: José Pérez Sánchez, owner, Persa Flores. Source: Persa Flores: How people find us through ChatGPT (shop data; also covered by Practical Ecommerce, 2026-11-01) (2026-10) · Added 2026-11-05.
How to set it up yourself
- What Persa Flores did: it checked the visit Shopify records before each online order and counted the ones that arrived from ChatGPT, using the utm_source=chatgpt.com tag ChatGPT adds to its links, not just the referrer.
- In your store admin, open each order from the last three months and look at its conversion details (first and last visit, referring site and landing page with the full URL). Note every order whose landing URL contains utm_source=chatgpt.com, and the page it landed on.
- Paste the list into a spreadsheet and group by month and landing page, or paste it into an AI assistant with the prompt below. Look for the pages that win ChatGPT orders (for this florist, city delivery pages).
- Work through the shop's checklist on those pages: same facts everywhere, older pages updated, address, phone and delivery area in structured data, delivery and returns pages not blocked to ChatGPT's crawler, full price including delivery shown, few clear options with prices. Re-count each month.
Copy-paste prompt
Below is a list of my online orders for {MONTHS}, one per line: order date, order total, landing page URL, referring site. 1) Mark each order as ChatGPT if the landing URL contains utm_source=chatgpt.com or the referrer is chatgpt.com. 2) For each month, give the share of orders and of revenue that came from ChatGPT. 3) List the landing pages ChatGPT shoppers arrived on, most common first. 4) Say how many ChatGPT orders had no ChatGPT referrer. Show your counts so I can check them.
Orders: {PASTE ORDER LIST}We set up one AI automation in your tools, live within 48 hours of getting access, for $197. No call needed. Founding price for our first 10 customers (regular $297).
One short email a day: 3 new AI use cases from real small businesses, each with its source and setup steps, plus a Tool of the Day. Free, no spam, unsubscribe anytime.
Rather not do it yourself?
The $197 Quick-Win: we set up one AI automation that saves you hours each week, in accounts you own, live within 48 hours of getting access. No call needed. Refund if it isn't live in 48 hours or isn't doing its job in the first 14 days. Founding price for our first 10 customers (regular $297). Terms.
See the $197 Quick-WinNeed more than one automation or ongoing care? Bigger setups: $149/mo with setup included (6-month minimum), or $497 pay-once. Satisfaction guarantee: we'll fix it until it works. Pay-once setups: or we refund your setup fee. $149/mo plan: cancel in the first 30 days and we refund your first month. We reply within 1 business day.
Related for Shops & e-commerce
Use ChatGPT to bulk-add book listings and spot titles to feature
The owner uses ChatGPT to fill in book details in bulk, uses AI on Google Analytics data to spot rising interest (romance before Valentin…
ChatGPT
Canva AI (Magic Studio)GAGoogle AnalyticsAdd a catalogue-only AI 'librarian' to your bookshop website
'Ask the Librarian' answers in the reader's language, only recommends books in the store's catalogue with live price and stock, and puts …
Use AI to research, price and catalogue thousands of used books before opening
AI helps find prices for comparable copies, check whether sets are complete, decipher old inscriptions, and turn his dictated book detail…
Running something like this in your business? Submit your use case and we'll credit you with a link.
Get 3 new AI use cases every morning
One email every morning: 3 new AI use cases from real small businesses, each linked to its source, plus a Tool of the Day. Free. No spam, unsubscribe anytime. Privacy
Rather have one set up for you? The $197 Quick-Win.