How to get feedback from your first users

You have a few users or buyers, and you do not know why they came, what nearly stopped them, or why the others left. To get feedback from your first users, ask every one of them one short question about something they just did, write down their exact words, and fix the problem that comes up most. If you send one "Any feedback?" email and wait, you will hear nothing, or "looks great", and keep guessing.

Do these three things

About 2 hours today, then 1 to 2 hours a day, plus about 3 hours on days you send 100 messages. Judge it after 30 answers, not after a week.

Show the three actions in fullHide the full list

1. 1Today: email every user and customer you have one personal question, start a feedback log in their exact words, and put a one-question survey on the page where most visitors leave.
2. 2Every day: ask every new user or buyer one question within 24 hours, ask everyone who cancels or goes quiet what stopped them, and hold every 15-minute call you can book. With under 5 new users a day, also send 100 personal messages asking people who fit to try it while you watch. AI drafts each message; you add one true line.
3. 3Every Sunday: count which problem came up most, fix it if 3 or more of the right people said it, and write to each of them to say it is fixed.

Get help doing these

ChatGPT or Claude reads this guide and works through it with you.

Do this with ChatGPT(opens in a new tab)Do this with Claude(opens in a new tab)

1. Email every user and customer you have one question today

The fastest way to get feedback from your first users is a personal email to each one, from you, with one question about something they already did. One question gets a one-line reply. A link to a 10-question survey gets put off and forgotten.

Start the feedback log first. Make one Google Sheet with these columns. Every answer from every step in this guide goes here, in the person's own words, never your summary. A sheet is enough until it gets slow; then move it into a CRM.

Date | Name | Who they are (paying / trial / signed up, quiet / visitor / left / client) | Fits your buyer? (yes / no) | Where (email / call / website / leaving) | The question you asked | Their exact words | The problem in 3 words | Told them it is fixed (date)

Fill in one row for every user and customer you have: everyone who paid, signed up, ordered, or asked for a quote. Mark "Fits your buyer?" against your ideal customer profile. A friend who bought to help you is a "no": friends are too polite, and their answers do not tell you what a stranger needs.

Pick the question by who they are. Each one asks about the past, never "would you", because people are polite about the future and honest about what they did. This is the same rule the customer interview questions use.

Send them today. Ask ChatGPT or Claude to draft each email from your list. You add one true line about the person: what they bought or did, and when. Send each one on its own from your own Gmail, never as a mail-out. Then follow up on day 3, day 7 and day 14 with a short "Just checking you saw this", up to 3 times. Reply to every answer within a day with "Thank you" and one follow-up question: "Why was that?" or "What happened next?"

Subject: one question about [product]Hi [first name],I'm [your name], I make [product]. [One true line: what they bought or did, and when.]Can I ask you one question? [The question from the table.]A one-line reply is perfect. I read every answer myself.Thanks,
[Your first name]

• Tool: Gmail and Google Sheets. Free.
• Time: about 2 hours for 30 people: 3 minutes an email, plus the sheet.
• You will have: a feedback log with a row for every user and customer, and a sent date on each.
• It worked if: every person on your list got an email today. Expect about 1 or 2 replies from every 10; one founder with 20 users reported 15% in 2026. Under 1 in 20 after the 3 follow-ups means the question is too big or the email looks automatic.
• Common mistake: asking "Any feedback?" or "What do you think?". People answer "looks great" and you learn nothing. The other mistake is counting your friends' answers the same as strangers'.

2. Ask every new user or buyer one question within 24 hours

From today, ask every new user or buyer one question within 24 hours of signing up or buying, while they still remember why they came and what nearly stopped them. Asked a month later, they have forgotten.

Every morning, open your list of new signups or orders since yesterday and send each one a personal email with one question. AI drafts it from the sign-up or order details; you add the one true line. Ask:

• New signups: "What made you sign up today, and what were you using before?"
• New buyers: "What nearly stopped you from buying?"

If you already send onboarding emails or an order email flow, keep them. This question goes in a separate, plain email from you, so a reply reaches a person.

Then add one question at the moment that matters most for your business:

If you run a SaaS: ask in the app, right after a new user gets their first win. In PostHog, open Surveys and click New survey, choose Popover, and write one free-text question: "What almost stopped you from getting here?" Under Display conditions, use User sends events with the event for your first win. PostHog's free plan gives 1,500 survey answers a month and one question per survey, which is all you need.

If you run a D2C store: ask on the Shopify Thank you page. Install Zigpoll (the free Lite plan takes 50 answers a month) and add one question: "What almost stopped you from buying today?" Since 26 August 2026, Shopify no longer runs survey code pasted into "Additional scripts" on stores not on Plus, so add it as an app block: Settings, then Checkout, then Customize, switch the page to Thank you, click Add app block, choose Zigpoll and save. Then email each buyer 7 days after delivery: "How was the first time you used it? Anything you didn't expect?" Happy answers become your first product reviews.

If you run a B2B company or an agency: ask at the end of the first week of every order or project, on a call or by email: "What has been harder than it should have been so far?" For agencies, put the same question in your client onboarding and in every monthly report.

Answers to "What made you sign up today?" also tell you which channel brought each person, so add it to the log.

• Tool: Gmail; PostHog Surveys (free up to 1,500 answers a month) for SaaS; Zigpoll (free up to 50 answers a month) for Shopify.
• Time: 30 minutes to set up, then 3 minutes for each new user or buyer.
• You will have: a new row in the log for every new user or buyer, within a day.
• It worked if: at the end of each day, no one who signed up or bought more than 24 hours ago is missing from the log, and a test signup or order you make yourself triggers the question.
• Common mistake: asking only "How did you hear about us?". It tells you where they came from, not what nearly stopped them, and that second answer is the one that brings the next sale.

3. Get user feedback on your website with a one-question survey

To get feedback from the people who visit and leave, show one question on the page where most of them leave, and watch recordings of a few who left. These are the people who never become users, so this is the only way you hear from them.

1. Find the page where most people leave. It is usually the pricing page, the product page or the checkout. If you track conversions, look at your numbers; if not, pick the page with the price on it.
2. Add PostHog to your site if it is not there yet. On Shopify, paste the PostHog snippet into theme.liquid, just before </head>. Session recording is free for 5,000 recordings a month.
3. Make one popover survey. In PostHog, Surveys, New survey, Popover. One free-text question: "What's stopping you from [signing up / buying / booking a call] today?" Under Display conditions, use URL targeting to show it only on that page. Under Customization, set the delay to 30 seconds, so only people who read the page see it.
4. Watch 5 recordings a day of people who left that page without buying. Write one line in the log for each: where they scrolled, what they clicked, where they stopped.

If you run a SaaS: answers here usually point to your pricing page or a home page that does not say what the product does. The full fix list is in traffic but no signups.

If you run a D2C store: put the question on your best product page, not the checkout. Common answers are about price, shipping and "will it work for me", which you fix in the product description and in traffic but no sales. For people who leave at checkout, use abandoned cart emails instead.

• Tool: PostHog Surveys and session replay, both free at your size.
• Time: 45 minutes to set up, then 15 minutes a day.
• You will have: answers from visitors who did not buy, and notes from 5 recordings a day.
• It worked if: the question shows on your phone after 30 seconds, and a test answer appears in PostHog. Then expect about 1 answer for every 13 people who see it: the median for website surveys in Survicate's 2026 report was 7.64%.
• Common mistake: showing a survey to everyone the second they land on the home page, or asking for a 1-to-10 rating. Both get ignored, and a number does not tell you what to fix.

4. Book a 15-minute call with every user who will talk, and watch them use it

Offer a 15-minute call to every person who replies, and on the call, watch them use your product instead of asking what they think. Calls give the best answers: people show you where they get stuck, which they cannot explain in an email.

Offer a call to everyone who replies, in your reply to their answer: "That really helps. Could I watch you use it for 15 minutes? You share your screen, I stay quiet: [booking link]." Make the link with Cal.com's free plan, which has no limit on bookings. Hold every call you book. Follow up up to 3 times with anyone who said yes but did not book.

On the call, they do; you watch. Do not show them the product: a demo gets you polite nods. Use these questions, and write their exact words in the log within 10 minutes of hanging up.

Opening: "Thanks for this. I'm not selling anything. I want to see where [product] is confusing. Can you share your screen and [do the main task] the way you would on your own? Please talk out loud. I'll stay quiet."While they use it, when they pause: "What did you expect to see there?"
At which step did you hesitate, and what did you expect to see next?
Tell me about the last time you had [the problem]. What did you do?
What nearly stopped you from [signing up / buying]?
What's the one thing that annoyed you most so far?
Who else do you know who has this problem?
Closing: "This really helped. Can I write to you when I fix what you showed me?"

Question 1 is the one a founder on the Shopify Community was given in August 2026 when users left without saying why. Question 5 often finds your next user, and the people who say good things are the ones to ask for a testimonial.

If you get fewer than 5 new users a day, you will not have enough people to learn from. Send 100 personal messages a day to people who fit your buyer, asking for 15 minutes "to watch you try it and see where it is confusing". AI drafts each message from the person's profile or post in about 30 seconds; you add one true line about them, so each one takes about 2 minutes. Find the names in the communities where your buyers talk and your lead list. Many of them become beta users. On a Google Workspace account, Gmail's limit is 2,000 messages a day (500 on a trial), so 100 is well inside it. LinkedIn caps connection requests each week, so send most of the 100 by email and in communities, and the rest on LinkedIn.

If you run a D2C store: most shoppers will not share a screen. Ask buyers for a 10-minute phone call or a voice note about their first use instead. To see your store through a stranger's eyes, ask people who fit your buyer to shop it on a video call while you watch.

If you run a B2B company or an agency: run it like a short discovery call with an existing client: ask them to walk you through the last order, report or project, step by step.

• Tool: Cal.com (free) and your usual video call link; Gmail.
• Time: 15 minutes a call plus 10 minutes for notes. On days you send 100 messages, about 3 hours more.
• You will have: call notes in the person's exact words, with the step where they hesitated.
• It worked if: every person who replied got a call offer, and you hold 1 to 3 calls a day. If fewer than 1 in 20 of your 100 messages books a call after 500 messages, change who you message, not how many.
• Common mistake: talking for 12 of the 15 minutes. If you explain the product, you hear your own words back. Ask the task, stay quiet, and write down where they stop.

5. Ask everyone who cancels, leaves or goes quiet what stopped them

Ask every person who cancels, stops using your product, does not reorder or says no one question, within 24 hours: what was the main reason? People who churn tell you the problems your happy users have learned to live with.

Pick the leaving moments for your business:

If you run a SaaS on Stripe: turn on cancellation reasons. In the Stripe Dashboard, go to Settings, then Customer portal, open Cancellations, and choose the reasons people see: "It's too expensive", "I need more features", "I found an alternative", "I no longer need it", "Ease of use was less than expected", "Quality was less than expected" and "Other reason", which lets them type. You see each answer on the subscription's page under Billing, then Subscriptions. Also email every signup who has not been back for 7 days, and every trial that ended without paying: the free trial guide has the timing.

If you run a D2C store: email every buyer who has not reordered by the time they would run out. Do not put a discount in this email; the win-back email comes later. What they tell you is how you increase repeat purchases.

If you run a B2B company or an agency: ask after every lost proposal and every client who ends: "What made you choose the other option?" For agencies this is the start of keeping clients longer.

Send it the same day you see it. AI drafts each one from what the person did; you add the true line. If they answer, reply with one question: "What would have had to be different?" Do not argue with the answer.

Subject: one question before you goHi [first name],I saw you [cancelled / haven't been back since (date) / haven't reordered / went with another option]. No sales pitch: what was the main reason?Reply with a letter if that's easiest:
A) Too expensive
B) Missing something I need
C) Too hard to use
D) Found something else
E) Don't need it right now
F) Something else: ___Thanks for trying [product].
[Your first name]

• Tool: the Stripe customer portal (free with Stripe Billing), Gmail.
• Time: 10 minutes to set up, then 3 minutes a person.
• You will have: a reason, or a sent question, in the log for every person who left.
• It worked if: every person who left in the last 7 days has a row. Aim for a reason from at least 1 in 4.
• Common mistake: putting "Here's 20% off to come back" in the same email. Now the answer is about the discount, not the real reason. Ask first, offer later, only if the reason was price.

6. Every Sunday, count what people said, fix the top problem and tell them

Every Sunday, group the week's answers by the problem they describe, count how many of the right people said each one, fix the top problem if 3 or more said it, and write to each of them to say it is fixed. Without this step, feedback is a pile of emails.

1. Group the answers. Paste the week's rows from the log into ChatGPT or Claude and ask: "Group these answers by the problem they describe. Give each group a 3-word name, a count, and the row numbers. Do not merge different problems." Check each group against the rows yourself, and write the 3-word name into the log.
2. Count only the right people. Count answers where "Fits your buyer?" is yes. A request from someone who will never buy is not a reason to change your product.
3. Decide with these rules:
   • 3 or more of the right people said it: fix it this week. One fix at a time.
   • 1 or 2 said it: do not build anything yet. Ask the next 10 people about it directly: "Some people told me [problem]. Was that true for you?"
   • One person asked for a new feature: log it, and ask what they were trying to do. The problem behind it is often already on your list.
   • Many say they do not see why they need it: this is not a fix-the-product problem. Check why no one is buying and your value proposition.
   • Many say price: check you are asking the right people first, then read how to price a SaaS product or how to price a product.
   • 0 answers after 30 asks: the problem is who you ask or when. Ask closer to the moment, and ask people who fit your buyer.
4. Tell them it is fixed. Write to every person who raised the problem you fixed: "You told me [their words]. I changed it: [what changed]. Thank you." Fill in the last column of the log. People who see their answer turn into a change answer again, and many become your first case study or referral.
5. Fill in the Sunday row, and once a month put it in your monthly marketing review.

Week | People asked | Answers | Calls held | Top problem (count of right people) | Fix shipped | People told it is fixed

Once you have 40 active users, ask them the product-market fit question once a month. Superhuman's founder made it famous: "How would you feel if you could no longer use [product]?" with three answers: very disappointed, somewhat disappointed, not disappointed. Ask only people who used it at least twice in the last two weeks. About 40 answers give you a direction; the common bar, from Sean Ellis's survey of about 100 startups, is 40% "very disappointed". Superhuman started at 22%. Below 40 users, the percentage swings too much to mean anything; keep asking one question at a time.

• Tool: Google Sheets, and ChatGPT or Claude for the grouping.
• Time: 45 minutes every Sunday, plus the fix. Block it in your week as a solo founder so it happens.
• You will have: one Sunday row, one fix chosen, and a note sent to everyone who raised it.
• It worked if: you can name this week's top problem, how many of the right people said it, and the rule that picked it, and every person who raised a fixed problem got a note.
• Common mistake: building whatever the loudest person asked for, or changing three things in one week. Fix the problem most of the right people share, one at a time, so you can see what each fix changed.

Your first-user feedback checklist

0 of 15 done

Work through this list with an AI assistant

It drafts the work and keeps track of what is done.

Do this with ChatGPT(opens in a new tab)Do this with Claude(opens in a new tab)

Step 1 · Email everyone one question

• Start the feedback log with a row for every user and customer you have.
• Email every user and customer one personal question.
• Every dayFollow up up to 3 times and reply to every answer within a day.

Step 2 · Ask every new user within 24 hours

• Set up the one-question moment for your business: the in-app question, the Thank you page or the first-week question.
• Every dayAsk every new user or buyer one question within 24 hours.

Step 3 · Ask on your website

• Add a one-question survey to the page where most visitors leave.
• Every dayWatch 5 recordings of people who left and log what you saw.

Step 4 · Hold every call you can book

• Every dayOffer a 15-minute call to everyone who replies and hold every call.
• Every dayIf you get fewer than 5 new users a day, send 100 personal messages asking people who fit to try it while you watch.

Step 5 · Ask everyone who leaves

• Set up the leaving question for your business.
• Every dayAsk everyone who left what the main reason was.

Step 6 · Count, fix and tell them

• Every SundayGroup and count the week's answers, counting only the right people.
• Every SundayPick the one problem 3 or more of the right people raised, and fix only that this week.
• Every SundayTell every person who raised a fixed problem that it is fixed.
• Once a monthOnce you have 40 active users, ask the product-market fit question.

Copy listDownloadReset