In short: zero sign-ups after a launch has four possible causes, and they are cheap to tell apart if you check them in the right order: the sign-up is broken or not counted, nobody came, people came and the page lost them, or the page worked and the offer did not. Most founders start with the third and rewrite the headline. Check the first two before you touch the page, because they take ten minutes and they are the ones that make every later change look like it failed.
We are writing this from both sides. We run LaunchGrill, which grades landing pages against 18 published tests, and we launched it ourselves in September 2026 to very few sign-ups. The numbers below are ours: our own funnel, and the 26 launches on our public record as of 11 October 2026.
1. Is the sign-up broken, or just not counted?
This comes first because it is the cheapest to check and the most embarrassing to miss. Founders who wrote up a launch with zero sign-ups have found that the real cause was a sign-up event that never fired, or a user cap left switched on from testing. The product was fine. The page was fine. The counter was wrong.
Do this before anything else:
- Sign up yourself, from a private window, on your phone and on a laptop. Use an address you have never used. Go all the way to the first screen inside the product.
- Check that the account exists where you would look for it: your database, your auth provider's dashboard, your analytics event list. If one of them is missing it, you have found your problem.
- Look for limits you set while testing: an invite-only flag, a waitlist that never sends, a cap on accounts, an email domain blocklist, a confirmation mail that lands in spam.
- Read your sign-in mail as a stranger would. If it comes from a no-reply address with a bare link, many people will not trust it, and some providers will not deliver it.
A landing page grader cannot see any of this, ours included. It reads the page; it cannot tell whether the button behind the page works. Saying so is the honest start of this list.
2. Did anyone actually come?
"Zero sign-ups" means very different things at 30 visitors and at 3,000. Before you judge the page, write down two numbers for the launch week: unique visitors, and where they came from.
Here are ours, so you have something to compare against. In the two weeks from 17 September to 1 October 2026, our homepage had 51 visitors and 84 visits. 70% of visits left without a second page. That is not a sample anyone can learn a conversion rate from. At about 50 visitors a week, a headline change cannot be measured: the difference between two versions is smaller than the noise.
If your number is that small, the problem is traffic, not the page, and rewriting the page will not show you anything. Spend the week on getting people there instead.
Be careful about which traffic you count. Launch directories are the usual answer, and we measured ours: by 11 October we had 13 live directory listings, and together they sent 5 visits in eleven days. Three of them carry a followed link that helps search; the rest do not. The details are in which startup directories are actually dofollow, and every directory we checked has its own dated page under /directories. List on them for the record and for search, not for launch-week traffic.
3. Did the page lose them?
Only once you know the sign-up works and a few hundred people have seen the page is this the right question. Here is what fails most often on the pages graded on our record. We count each launch once, by its latest grill.
| Test | What it asks | Failed on |
|---|---|---|
| O4: reason to act | Is there an honest reason to buy now rather than bookmark? | 20 of 26 pages |
| A3: pain before product | Is the problem stated before the product is? | 18 of 26 |
| P2: specific proof | Is the proof specific (names, numbers, dates) rather than decorative? | 11 of 26 |
The median score across those 26 launches is 69.5 out of 100. The lowest is 17 and the highest is 92. A page in the high sixties is clear about what it is and still gives a visitor no reason to act today. That pattern is common enough that we wrote a guide about it alone: 20 of 26 landing pages failed the same test.
A note on A3: it is the least consistent of our 18 tests from one run to the next, so read its count as a direction, not a measurement.
What a lost visitor looks like in numbers: visits that end on the page they started on, short time on page, and almost nobody reaching the sign-up form. If people do reach the form and stop there, look at the form, not the page: how many fields, whether it asks for a card, whether it works on a phone.
4. Did the offer fail?
If the sign-up works, people came, and the page passes the basics, then the page may be doing its job and the answer is no. That is the hardest result and the most useful one. The usual signs:
- Visitors read the page, some click through to pricing or features, and still leave.
- The people who reply to your launch post say "nice" and nothing else.
- The few who sign up never come back.
No page grader can tell you this. It is a question for the people you built it for, asked directly. Five conversations with the people you meant to reach will tell you more than another headline.
What actually got us sign-ups
For honesty, here is what moved our own numbers, as of 1 October 2026. None of our six outside accounts came in cold through the homepage form. Four arrived after seeing a grill of a page: someone else's on the wall, or their own, which we had graded and sent them. The other two came from a post in a founder community and from Indie Hackers. Thirteen grills sent to named founders brought two accounts. The homepage form, asking a stranger to type their own address, brought about none from roughly 35 visitors.
That is our product and our audience, so it is not a rule. But it is the pattern behind this guide's order: make sure the counter works, make sure people come, then show them something about themselves rather than asking them to start from nothing.
The order, once more
- Sign up yourself, as a stranger, on two devices, and find the account.
- Count the visitors and their source. Under a few hundred, work on traffic.
- Grade the page against a fixed list, ours or anyone's, and fix the reason to act first.
- Ask the people you built it for. If the page passes and they still do not sign up, the page was never the problem.
You can run step 3 on your own page for free: paste your address on the homepage and you get a score, a grade and the fixes in a few minutes. Your grill is public, like the others on the record.