The Startup Launch Checklist That Sets Numbers
A startup launch checklist with the exact site to open and the number to set for each step, not seven vague things to feel ready before you launch.
Nobody's launch dies because the favicon was the wrong size.
It dies two weeks after the "big day," when the founder has done everything seven different checklists told them to do and still can't say which of it worked.
Money went out. Clicks came in from somewhere. Nobody wrote down which spend produced which click, so nobody can say what to do again next time.
A startup launch checklist that's worth running skips the feelings. It's not a list of things that make you feel ready.
It's a list of exact pages to open and exact numbers to type into them, run in an order that gets you an answer before the money to ask twice is gone.
Here's one built that way. Sixteen steps, each naming what to click and what number to land on, run in the order a launch actually happens rather than the order a listicle happens to publish it.
Why Most Launch Checklists Don't Move Anything
Read five checklists against this exact term and you'll find the same eleven items rearranged. Lock the brand. Set up analytics. Build a press list. Tell your network.
None of it is wrong.
All of it is unmeasured.
One popular version tells founders to "confirm all teams approve the launch state." Another, aimed at founding teams generally, says to make sure "all technological systems are fully operational." Both are structurally sound, organized into clean phases a reader can follow start to finish.
Neither one tells a reader what page to open or what number means done. A step without a number attached is a feeling dressed up as a task, and a founder running on feelings during launch week is exactly how a $500 ad budget turns into a $4,000 one with nothing to show for the difference.
The deeper problem sits upstream of any checklist at all. CB Insights analyzed 385 startups that had shut down and found 43% failed on poor product-market fit, with another 29% citing bad timing (CB Insights, Why Startups Fail: Top 9 Reasons). No submission list and no press round fixes either of those. A perfectly executed checklist for a product nobody wanted just spends the marketing budget faster than a sloppy one would have.
So step one isn't a marketing step.
It's a receipt.
The Startup Launch Checklist, In Order
Before You Touch a Marketing Channel
Test demand for under $100 before you write a line of launch copy. Build one page with a real "reserve" or "pre-order" button, not a "notify me" form, and put $100 behind driving strangers to it. Say that buys 200 clicks at $0.50 each. Twenty of those 200 clicking the actual button is a working signal.
Two clicking it is a different problem than a launch problem. If the test comes back at two, the fix is the offer or the audience, and nothing further down this list will change that number. Rewrite the pitch, run another $100, and don't queue a single directory until this number moves.
Register the domain and turn autorenew on the same day. Open Namecheap's .com registration page, checked at $11.28 for the first year with renewal at $18.48, and buy it now rather than launch week. A domain under 30 days old reads as new to some directories and gets quietly held out of the daily queue.
Verify the property in Google Search Console and submit the sitemap that day. Add the domain as a property. Verify it with the DNS TXT record method through your registrar's panel. Verification usually confirms in under ten minutes once the record propagates. Then submit
/sitemap.xml.Google's own documentation is direct about what this does and doesn't guarantee: a submitted sitemap is a hint to the crawler, never a promise of a crawl date (Google Search Central, Build and Submit a Sitemap). It's still the fastest hint available, and it costs nothing but ten minutes.
Three steps in, and nothing here is marketing yet.
That's deliberate. A domain bought during launch week and a sitemap submitted an hour before the post goes live are both self-inflicted delays that show up as missing traffic three days later, with no obvious cause to blame.
The Week Before
Write the one-line pitch and cut it until it's under twelve words. Read it to three people who've never seen the product. If two of them can't repeat back what it does, the pitch isn't ready, no matter how ready the product is.
Build the outreach list to 20 names, not 200. Twenty people who'd plausibly write about or link to the specific thing you built beats two hundred generic contacts pulled from a media database who've never heard of the category.
Search for who covered the two or three closest comparable launches in the last 90 days. Open each of those articles, note the byline, and add that name before adding a single contact from a list you didn't build yourself.
Set the launch date now and don't move it for anything short of a broken product. Pick a Tuesday or Wednesday. A weekend launch skips two of the five weekdays that directories and reporters actually run their queues on.
Monday is the wrong instinct for a different reason. Everyone else who read the same advice picked it too, and a launch queue with fifty other Monday launches ahead of yours behaves nothing like a launch queue with five. Pick the date in your own morning, not a reporter's, and confirm the time zone matches the directory's, not the one on your laptop.
Queue the submission list in the order that compounds. Product Hunt still produces the single biggest one-day spike available to a consumer product, so it goes first.
Follow it with the startup directories worth the ten minutes each one takes, and add the earlier-stage lists if the product isn't public yet, using the BetaList alternatives built for that stage.
Then work down everywhere else worth a submission on its own day, not all at once.
Reserve one paid placement for launch day, with a dollar figure decided in advance. A submission list gets found by people already looking. A paid placement gets seen by people who weren't looking yet, and that's a different kind of reach worth pricing on purpose rather than backing into at midnight.
The Board is one place to buy that second kind of reach. A dollar puts a brand on it, rank is sorted strictly by money paid, and that's disclosed here as exactly what it is: a paid placement, not an editorial pick. It's one of nine specific reasons founders are choosing to spend that dollar during launch week rather than saving it for later, more on The Board's launches shelf.
Set the total launch-week budget ceiling before launch week starts. Pick the number while you're calm, not on the day a channel is clearly working and every extra dollar feels obviously worth it.
Split it on paper first. A $500 ceiling might set aside $200 for the paid placement and hold $150 back for whichever channel proves itself on day one. The remaining $150 already went to the demand test in step one, so it isn't spent twice by accident.
Writing the split down before launch day removes the decision you'd otherwise be making at 11pm with adrenaline still in your system.
Launch Day
Publish, then request indexing manually instead of waiting for the crawl. Open the URL Inspection tool inside Search Console and click "Request Indexing" on the launch page itself, the same hour it goes live.
Post to the queued list in order, spaced by the hour, not fired all at once. A single 9am blast across ten platforms reads like automation to more than one review queue, and getting flagged as a bot costs you the exact discovery you queued the list to earn.
Track clicks hourly for the first four hours. Not impressions. Impressions tell you the algorithm noticed the post. Clicks tell you the pitch worked on a person.
Say hour one shows 400 impressions and 3 clicks. Hour two shows 900 impressions and 4 clicks. Impressions are climbing; clicks aren't. That's the headline failing, not the channel, and rewriting it on the spot beats waiting for the day to end and hoping the next platform does better with the same words.
Reinforce whichever channel is already working instead of adding a new one. If one submission or the paid placement is producing clicks, put more attention, and where it applies more budget, into that one before trying a second thing. Split focus on day one and by evening you won't know which half of the effort actually moved the number.
The First Seven Days
Check position in Search Console on day 7, not day 1. New pages rarely rank inside the first week no matter what got done right; checking on day one just manufactures anxiety with nothing underneath it yet.
Set the kill number before you look at the metric, never after. Decide it in advance, per channel, and hold to it:
| Channel type | Kill signal, decided in advance |
|---|---|
| Paid clicks to a landing page | Cost per click over $2 after the first 100 clicks |
| Directory or list submission | Zero referral clicks after 7 days live |
| Cold outreach | Zero replies after 20 sent |
| Paid placement | Impressions delivered but clicks under half your site's normal rate |
Checking the number after you've already decided you like a channel just produces reasons to keep it.
Deciding first removes that option entirely. A number set in advance doesn't care how much you liked the channel by Thursday.
- Write down what happened while it's still fresh. Note what it cost and what it sent. Then decide whether it's worth doing again next time, in one line per channel, no more.
The next launch only gets easier than this one if this one left notes behind it.
What the Numbers Actually Tell You
None of the sixteen steps above works if the only thing being tracked is total traffic. Traffic without a source attached is a mystery, not a metric, and a mystery doesn't tell you what to do differently next week.
This site holds its own numbers to the same standard it's asking of yours: The Board's own traffic, live and dated, checkable by anyone, same as every figure above ought to be.
More of this kind of step-by-step, numbered rather than felt, lives on the launches shelf of this blog, added to as the real prices and the real queues change under it.
A launch checklist that ends on "post everywhere and see what happens" isn't a checklist.
It's a hope with steps written underneath it. The founders who come back to run a second launch aren't the ones who felt the most ready. They're the ones who wrote a number next to every step and still have the page open where they checked it.
Run the sixteen above in order. Write the number down before each one, then check it against the number after. That's the entire method. Everything past this point is just deciding how fast to move through it.