AI prompts for your Shopify survey data
Connect Grapevine to Claude or ChatGPT and you can ask your survey data questions in plain English. The hard bit is knowing what to ask. So here are our favourite prompts, sorted by the survey question you’re running. Copy one, paste it into your AI assistant, job done.
Not connected yet? Here’s how to connect Grapevine to Claude or ChatGPT. It’s quick.
Good to know: your AI assistant only sees survey data, never customer names, emails or order details. Individual responses go back 90 days. Summaries and question reports can cover longer stretches, like this quarter vs last.
Attribution survey prompts
Attribution surveys ask customers the simplest question in marketing: how did you hear about us? It usually sits on the thank you page right after checkout, when the purchase is fresh and people are happy to answer. The answers fill the gaps your analytics can’t see, like the friend who recommended you, the podcast ad heard on a dog walk, or the Instagram post someone saw weeks before searching your name. Over time it builds a clear picture of which channels actually create customers, not just which ones get the last click.
-
Break down the responses to my “How did you hear about us?” question for the last 30 days and compare them to the 30 days before. Which channels moved the most?
-
Read the “Other” responses on my attribution question from the last 90 days. Which channels keep coming up that aren’t answer options yet?
-
Find every response from the last 90 days that names a podcast, YouTuber or influencer. Group them by who was mentioned and count them.
What to do with the answers
- If people keep responding with the same channel, add it as a proper answer option. Next month’s data will be cleaner.
- If a channel suddenly jumps and you didn’t change anything, find out why before you move budget. See how to spot attribution anomalies with Claude.
- Send the answers to Klaviyo and build segments by channel, so podcast listeners and TikTok finds get different emails. How Klaviyo profile enrichment works.
Post-delivery NPS prompts
NPS asks customers how likely they are to recommend you, on a scale of 0 to 10. When you ask matters more than the question itself. Ask at checkout and you’re scoring your website. Ask once the order has arrived and you’re scoring the whole experience: how fast it came, how it was packed, and whether the product lived up to the hype. That’s why post-delivery NPS is the one worth tracking, usually sent by email a few days after the parcel lands. See how NPS surveys work in Grapevine.
-
What’s the NPS score on my post-delivery survey for the last 90 days, and how does it compare to the 90 days before?
-
Read the comments on my post-delivery NPS survey from the last 90 days. Sort the complaints into delivery, packaging and the product itself. Which group is dragging my score down most?
-
Show me my post-delivery NPS week by week for the last 12 weeks. If any week dipped, read the comments from that week and tell me if it was a courier problem, a stock problem or a product problem.
What to do with the answers
- Delivery complaints: take the comments to your courier or 3PL as evidence. If the problem is late parcels, set clearer delivery times at checkout. Disappointment is usually about expectations.
- Product complaints: if people keep saying it wasn’t what they expected, fix the product page first. Photos, sizing and descriptions are quicker to change than the product.
- Promoters: a 9 or 10 straight after delivery means they’ve got the product in hand and love it, which is the best moment to ask for a review. Send them to Klaviyo and trigger a review request. How Klaviyo profile enrichment works.
Shopping experience CSAT prompts
CSAT asks customers how satisfied they were with one specific moment, from very dissatisfied to very satisfied, using words or emojis. The moment that matters most for an online store is the shopping itself: finding the product, getting answers, and getting through checkout without swearing at the screen. Ask it on the thank you page and you get a read on your website while it’s still fresh in the customer’s mind. Low scores here are usually fixable in an afternoon. See how CSAT surveys work in Grapevine.
-
Show me my CSAT results for the last 90 days. What share of customers said they were satisfied or very satisfied, and how does that compare to the 90 days before?
-
Read the comments from customers who were neutral, dissatisfied or very dissatisfied in the last 90 days. What made shopping with us harder than it should have been? Group them into themes.
-
We changed [what you changed] on [date]. Compare my CSAT results for the 30 days before and after that date, then read the comments from after the change. Did it help?
What to do with the answers
- Fix the top theme first. If people keep struggling to find sizing, delivery costs or returns info, put it on the product page or in the cart where they actually look.
- Check your changes. Run the before-and-after prompt every time you change the site, the theme or checkout, so you know whether the change actually helped.
- Look after low scorers. They bought anyway, but they’re shaky. Tag them in Klaviyo and send a friendly check-in after delivery, so one clunky checkout doesn’t cost you the second order. How Klaviyo profile enrichment works.
Purchase barrier survey prompts
A purchase barrier question asks customers what nearly put them off buying. It sits on the thank you page, so you’re asking people who made it through checkout what almost stopped them. Their answers are gold, because the worries that nearly lost you this sale are the same ones that lost you the people who left. Delivery costs, sizing doubts, “is this brand legit?”: you’ll hear it all in their own words. It’s the cheapest conversion research you’ll ever do. See how website performance surveys work in Grapevine.
-
Read the answers to my “What almost stopped you from buying?” question from the last 90 days. What are the top five reasons people nearly didn’t buy, and roughly what share mentions each one?
-
From the last 90 days of answers to my “What almost stopped you from buying?” question, pull out every mention of price, delivery cost or delivery time. Show me a few typical answers for each so I can see the exact wording customers use.
-
Compare the reasons people nearly didn’t buy in the last 30 days with the 60 days before that. Is anything new showing up, or is an old problem getting worse?
What to do with the answers
- Answer the worry where it happens. If delivery cost is the issue, show it on the product page, not as a surprise at checkout. If it’s sizing, put the size guide right next to the size picker.
- Turn the top three into FAQs. Use the customer’s own wording for the question, add them to your product pages, and answer them plainly.
- Hear from the ones who left. These answers come from people who bought anyway. To catch the ones who didn’t, run an exit-intent survey alongside it. See how exit-intent surveys work.
Purchase motivation survey prompts
A purchase motivation question asks customers what made them buy today. It sits on the thank you page, usually right after your attribution question, so you learn where people came from and what actually sold them. Maybe it was the reviews, a gift deadline, the free returns, or that one product photo. Whatever it is, it’s the message you should be shouting about, and most stores are guessing at it. Browse our question library for wording ideas.
-
Read the answers to my “What made you buy today?” question from the last 90 days. Group them into the main reasons people bought and rank them by how often they come up.
-
Pull out the exact phrases customers use in their answers to my “What made you buy today?” question from the last 90 days. Give me the ten that come up most, word for word, so I can use them in ads and emails.
-
Using responses from the last 90 days where customers answered both “How did you hear about us?” and “What made you buy today?”, compare the reasons people bought by channel. Do customers from different channels buy for different reasons?
What to do with the answers
- Lead with the top reason. Whatever sold most people should be the first thing on your homepage and product pages, not buried halfway down the description.
- Use their words, not yours. Drop the exact phrases into ad headlines, email subject lines and product copy. Customers describe your product better than your marketing does.
- Match the message to the channel. If podcast listeners buy for quality and TikTok finds buy for price, tell each group what they care about. Send both answers to Klaviyo and segment on them. How Klaviyo profile enrichment works.
Ask your survey data anything with Grapevine
Grapevine runs your surveys on Shopify and connects the answers straight to Claude or ChatGPT. No exports, no spreadsheets, just ask your question and get an answer.
- Attribution, NPS, CSAT and custom surveys in one app
- Connects to Claude and ChatGPT with a secure, read-only AI connection
- Your AI never sees customer names, emails or order details
- Sends answers to Klaviyo and Shopify Flow, ready to act on
- $25 a month flat, unlimited responses, 14-day free trial