the score is the
headline. the reason is the story.

Two questions, one after the other: the standard 0–10 rating, then an open box asking what drove it. The rating is required here on purpose — a sentence with no number attached cannot be filed under promoter or detractor, and filing it is the whole point.

Install free on Shopifyfree plan · AI summaries on Pro · $6/mo flat

the template, exactly as it ships

One question on screen at a time. The rating is submitted the moment it is tapped, so the second question can never cost you the first answer.

On your Thank You page

How likely are you to recommend us to a friend or colleague?

What's the main reason for your score?

The rating is required; the written reason is optional and appears after it. Because each answer is saved as it is given, a buyer who taps a 9 and closes the tab has still told you they are a promoter — the follow-up is upside, never a gate.
In your response feed

Reasons, filed by bucket

Promoters who wrote47
“Second order this month. The refill arrived before I ran out, which has never happened with anyone else.”
Detractors who wrote11
“Checkout made me create an account before it would take my card. Nearly gave up.”
Illustrative. Both quotes are attached to the order, the customer and the rating that came with them — so “nearly gave up” is a 4, from a first-time buyer, on a specific day.

a number tells you there is a problem

It will not tell you which one. That gap is the only reason this template exists as a separate thing.

A score that drops four points this month is an alarm with no address. You can guess — new courier, a price change, that batch of packaging — and you will spend a fortnight guessing. Eleven people writing the same sentence about checkout is not an alarm, it is a ticket, and you can close it on Thursday.

This is also the difference between a survey you look at and a survey you act on. Ratings accumulate into a chart that goes on a slide. Reasons accumulate into a list of specific, fixable things with a count next to each one, which is a fundamentally more useful object.

Read the promoters too

The instinct is to read the complaints and skip the praise, and it is the wrong instinct. Promoter reasons tell you which of your many virtues customers actually noticed — and it is usually not the one your homepage leads with. Stores discovering that nobody mentions the thing they are proudest of and everybody mentions the delivery speed have learned something worth more than a point of NPS.

They also hand you copy. A sentence a real customer typed about why they would recommend you outperforms a sentence you wrote about yourself, and it comes with the order it belongs to if you ever want to ask permission to quote it.

What the AI does with the text

On Pro, the open answers are summarized for you: recurring themes with a count, a sentiment split, and verbatim quotes checked against the saved response so a summary cannot invent one. A digest arrives once enough new written answers have gathered to be worth summarizing — there is no fixed schedule, and each one picks up where the last one ended.

On the free plan you collect every comment, without limit, and read them in the feed for the last 24 hours. The written summaries are the Pro part. There is a sample of one on the open-feedback template, which is the template that leans on it hardest.

Or run the short version first

If you have never measured this at all, the one-question NPS template gets you a baseline with the least possible friction, and you can switch to this one once you know which direction the number is moving. Coming from a tool that shut down? The Delighted migration guide is the honest version of what transfers. Everything else is in the template library.

questions about this one

Why does this template make the rating compulsory?

Because a written reason with no rating attached cannot be sorted into anything. “The box was crushed” is a complaint if a 3 sent it and a forgivable aside if a 10 did, and you cannot tell which without the number. Making the rating the gate means every sentence you collect arrives already filed under happy or unhappy.

Doesn't requiring an answer cost me responses?

It costs you some, and that is a deliberate trade. The rating is one tap, so the people it turns away are mostly the ones who would have tapped past both questions anyway — but if your priority is the highest possible count of scores rather than usable explanations, run the single-question version instead.

Is the written part compulsory too?

No, and it should not be. The box is optional and sits under the rating once the rating is given, so a customer who wants to leave a number and go can. The ones who do type something are self-selecting for having an opinion, which is exactly the population worth reading.

Do I have to read every comment myself?

On the free plan, effectively yes — you read the feed. Pro adds a written summary that groups the comments into recurring themes, splits sentiment, and quotes real answers word for word so you can check the summary against what was actually said. The comments themselves are collected on both plans and never capped.

What do I do with the reasons once I have them?

Split them by bucket and treat them as two different documents. The unhappy ones are a defect list ranked by how often each thing comes up. The happy ones are your product page and ad copy, already written in the words your customers use — which is usually better than the words you would have chosen.

ask why.
then fix that.

toldby is live on the Shopify App Store. Install free, publish the score and the follow-up together, and read the first reasons today.

Install free on Shopify