Blog
How to write a product launch postmortem after a quiet launch (template)
A product launch postmortem template for indie builders: collect the numbers, find where visitors dropped off, talk to five people, and choose one fix before relaunching.
Nov 22, 2026 · All posts

Why write a product launch postmortem at all
Most launches are quiet. You post the product, a directory lists it for a day, a few dozen people visit, two sign up, and by Thursday the traffic line is flat again. The tempting move is to shrug and start building the next feature.
A short postmortem turns that quiet week into a decision. Done well, it takes about two hours and answers one question: what is the single change most likely to make the next launch better? This guide gives you a template you can copy, with the steps in the order that actually works.
Step 1: Collect the numbers within 72 hours
Memory fades fast and analytics dashboards hide detail behind date filters. Within three days of launch, write the raw numbers into one document. Do not interpret yet.
The minimum data set
- Visits to your product page, per source (each directory, each social post, each community thread).
- Signups or trial starts, per source if you can tag them with UTM parameters.
- Activations: how many new users did the one thing that shows value (created a project, sent a first message, exported a file).
- Payments, if you charge.
- Replies and comments: copy every comment, question, and complaint word for word into the document.
If you launched on several channels, put them in a simple table with one row per source. Keep the links you posted so you can compare later. A public log of where you shared the product, like the ibuildinpublic links page, makes this easier because the posts are still findable weeks later.
Step 2: Find the step where people dropped off
Numbers alone do not tell you what to fix. Line them up as a funnel and look for the biggest drop between two steps.

Sort every number into a stage before you decide what went wrong.
Read the funnel like this
- Low traffic (under about 100 visits): the problem is distribution, not the product. The fix is where and how you launched.
- Traffic but few signups (under 2 to 3 percent): the problem is the product page. Visitors did not understand what it is or who it is for.
- Signups but few activations: the problem is onboarding. People wanted it but got stuck before the first useful result.
- Activations but no payments: the problem is pricing or the value is not strong enough yet.
Only one of these is your biggest leak. Write it down in one sentence: "62 visits from the directory, 11 signups, 1 activation. The leak is onboarding."
If the leak is the product page, work through our launch directory submission checklist before anything else. If the leak is payments, revisit how to price your SaaS before launch.
Step 3: Talk to five people before you guess
Data shows where people left. Only people can tell you why. Message five of them within a week, while they still remember visiting.
Who to ask
- Two people who signed up but never activated.
- Two people who visited or commented but did not sign up.
- One person who used it and liked it.

Fifteen minutes with someone who signed up and left is worth more than a week of dashboard watching.
Questions that get honest answers
- "When you first saw the page, what did you think it did?"
- "What were you hoping to do when you signed up?"
- "Where did you stop, and what were you doing right before?"
- "What would you use instead, or what do you use today?"
Do not pitch, defend, or explain the product during the call. Write down exact phrases. If three of five people describe the product differently from your tagline, your description is the problem, whatever the funnel says. The same outreach habits from getting your first 10 users apply here: short, personal messages get replies; generic surveys do not.
Step 4: Fill in the postmortem template
Put everything on one page. Keep it short enough that you will actually reread it before the next launch.
Copy this template
- What I launched: one sentence, plus the date and the channels.
- What I expected: the numbers you hoped for, written honestly.
- What happened: the funnel table from Steps 1 and 2.
- Biggest leak: one sentence naming the stage.
- Why, in users' words: three to five direct quotes from Step 3.
- What went right: at least two things worth repeating (a channel that converted, a comment that showed real need).
- One change for next time: a single, specific fix you can ship in under two weeks.
- How I will measure it: the number that should move, and by how much.
The discipline is in "one change". A list of ten improvements becomes a month of busywork. One change can be shipped, relaunched, and measured.
Step 5: Share it publicly and relaunch
Building in public is partly about showing the quiet weeks, not only the wins. A postmortem post is often more useful to other builders than the original launch announcement, and it brings back the people who missed the first round.
How to share it
- Post a short version: the funnel numbers, the biggest leak, and the one change. Skip the self-criticism.
- Update your builder profile and product page so the description matches what users actually said. Your page on ibuildinpublic builders and your listing in products should reflect the new version.
- Set a relaunch date two to four weeks out, after the fix ships. Tell the five people you interviewed; they are your first audience.
When to stop relaunching
If two relaunches with different fixes still show the same leak, the issue may be the problem itself, not the execution. That is the moment for a bigger decision: change the audience, narrow the use case, or move on. Writing each postmortem in the same format makes that call much clearer.
A quiet launch is data, not a verdict
Nobody remembers your first launch except you. What compounds is the habit of measuring, asking, and fixing one thing at a time. Write the postmortem, ship the one change, and launch again. For more on launching and getting found, browse the ibuildinpublic blog or check the FAQ.