How to Measure a Launch: Source Tags, One Core Number and a Verdict Date
Measure a launch by tagging every link with its own source, counting how many visitors finish the one task your product exists for, and checking who comes back after a week. Votes, views and sign-ups are context, not results. Decide before launch day what outcome counts as success and when you will judge it, because nobody can tell you in advance how many visitors a launch will bring.
What should you decide before launch day?
Write one sentence you can be held to: “I will call this launch worked if some number of people finish the core task by a certain date.” Fill in the blanks yourself. The number is yours because nobody has a figure that holds across products, and the core task is whatever your product exists for, such as sending a first invoice or running a first command.
Writing it down first matters because results are easy to flatter afterwards. A quiet week feels like a win if you were hoping for nothing and a loss if you were hoping for a crowd. The sentence stops the goalposts from drifting, and it tells you which numbers to collect.
How do you tag links so each source has its own numbers?
Add UTM parameters to every address you publish. Google Analytics Help, read on 13 November 2026, lists utm_source, utm_medium and utm_campaign as the ones to always use, and advises one unique source value per platform and one consistent name for each campaign, so a single effort does not split across several report rows. Other analytics tools read the same parameters, but check yours.
https://example.com/?utm_source=launchboard&utm_medium=listing&utm_campaign=launch-nov https://example.com/?utm_source=showhn&utm_medium=post&utm_campaign=launch-nov https://example.com/?utm_source=newsletter&utm_medium=email&utm_campaign=launch-nov https://example.com/?utm_source=forum&utm_medium=reply&utm_campaign=launch-nov
Platforms treat addresses differently, so test each one. Submit the tagged link, click it from the platform’s own page and read the address bar. The launch form on this board stores the address you give it with its query string, but a redirect on your own site, such as one that adds or drops “www”, can still strip the tags. If the parameters vanish, your analytics will file those visitors under direct or referral, and you will have lost the source. The UTM tagging checklist lists more cases.
Which numbers matter and which only look good?
| Number | What it tells you | The trap |
|---|---|---|
| Visits by source | Which places sent anyone at all | Counts run low because some visitors use blockers, and bots can raise them |
| Core-task completions | Whether the people who arrived got value | Needs an event you set up before launch day |
| Return visits after seven days | Whether it was use or curiosity | Takes a week to learn, so it is easy to skip |
| Comments and replies | What people said and asked | Easy to count, hard to judge; read them, do not tally them |
| Votes | That some people found it interesting | Says nothing about whether anyone used it |
| Sign-ups | Interest strong enough to leave an email | A list of people who never opened the product proves little |
If you can track only one thing, track completions of the core task, split by source. That figure ties the launch to the product, not to the platform. Launch metrics to track goes through setting events up on different kinds of product.
What does the arithmetic look like?
The numbers below are invented to show the sums. They are not a typical result and are not a benchmark. Suppose a tagged listing sent 40 visits, and 8 of those visitors finished the core task. Completions divided by visits is 8 over 40, one in five. Suppose 3 of those 8 came back within seven days. Returns divided by completions is 3 over 8, which is a better guide to whether the product earns a second visit.
Now compare two sources. One sent 5 visits and 3 completions. The other sent 60 visits and 4 completions. Judge them on completions per visit and on total completions together; the first has the better rate and the second brought more people. With samples this small, treat a rate as a hint, not a result, and send more traffic from the better source before you decide. The launch directory traffic calculator does the same sums once you enter your own visits and sign-ups.
When do you judge the launch, and what does each result tell you?
Take a first reading seven days after launch and a second at thirty. Those two dates are a suggestion; the rule is to pick the dates in advance. Then read the result against the sentence you wrote.
- Few visits, few completions: a reach problem. Go back to why nobody visited my launch and check the page was public, the link tracked and enough of the right people told.
- Many visits, few completions: a page or product problem. Watch two people try the core task and find the step where they stop.
- Healthy completions, few returns: the product does the job once and does not earn a second visit. Ask the people who left why they did.
- Healthy returns from one source only: put the next hours there, and drop the sources that sent visits but no completions.
What can you not measure, and how should you say so?
You cannot see how many people looked at your listing and did not click, whether someone shared your link in a private chat, or how many visitors blocked your analytics script. A board’s own impression counts, where they exist, belong to the board. Be honest about that in your notes and do not turn a gap into a number.
Keep one sheet per launch, with a row per source and columns for visits, completions, returns and what you did that day. When you launch your next product for free, the sheet is your baseline, and the ideas board is a quick way to see what others are building while you wait for the seven-day reading.
Frequently asked questions
What metrics should I track for a product launch?
Track visits by source, completions of the one task your product exists for, and return visits after seven days. Read comments for the reasons behind the numbers. Votes and sign-ups are context. If you can only follow one figure, use core-task completions split by source, since that ties the launch to the product.
How do I know which launch site sent me visitors?
Add a different utm_source to the address you submit to each site, and read the Traffic acquisition report in your analytics. Check that each platform keeps the query string, by clicking your own listing and reading the address bar. Without tags, visitors from several sites can end up under one direct or referral line.
How long should I wait before judging a launch?
Pick your dates before launch day. A reading at seven days and another at thirty works for many products, and the first shows reach while the second shows whether people returned. Nobody has a universal figure, so compare the result with the sentence you wrote about what would count as success.
Are upvotes a good measure of launch success?
Not on their own. Votes show that some people found the listing interesting, and they affect where it appears on boards that rank by them, but they say nothing about use. Pair them with visits and with how many visitors finished the core task, and give the completions more weight.
Written by Post your build ideas with AI assistance and edited for this site. Where a page names another product, the facts came from that product’s own pages on the date shown above; terms change, so check them before you rely on them. About the publisher.