Loading…
Guide
A launch is a distribution event. It is not a growth strategy, and the most common way to be disappointed by one is to have quietly expected it to be. The traffic a launch sends converts far worse than search traffic — something around a quarter of a percent is typical — and most of the spike does not persist past the week it happened in.
That is not an argument against launching. It is an argument for launching with the right expectation, because what a launch reliably leaves behind is a page and a link, and those are worth planning a week around. The graph is not.
Start with the arithmetic, because it is the part nobody puts in a launch checklist.
Launch-day visitors are browsing. They arrived at a board, scrolled a list of products they had never heard of, and clicked one out of curiosity. Someone who found you through a search arrived with a problem already in mind. Those are different people, and they behave differently: a conversion rate around 0.25% is a normal outcome for launch traffic, and it is a poor outcome for search traffic.
So a day that sends a few hundred visitors is a day that sends you a small handful of signups, if it goes well. That is worth knowing before you build a week of work on top of it, and it is worth knowing before you compare yourself unfavourably to a screenshot someone posted of a better day.
The second half of the picture is that spikes flatten. Traffic from a launch is traffic from a page that will be pushed down its own board within a day or two. Unless the product itself is doing the retaining, the following week looks much like the week before the launch. Plenty of good products launch, get a quiet day, and are fine.
The honest frame
Do a launch for the page and the link. If the traffic is good too, treat it as a bonus rather than the point. A plan built on the bonus is a plan that fails most weeks.
Two things outlive the day, and they are the reason the exercise is worth an afternoon.
A listing is an indexable page, on a domain that is not yours, that describes your product in your own words. It gets found by people searching for a category or for an alternative to a competitor, long after the board it launched on has moved on. That is a slower and much duller thing than a spike, and it is the thing that is still there in six months.
The other output is a backlink, and whether it is worth anything is a mechanical question you can answer rather than a claim you have to trust. A link that carries nofollow, sponsored or ugc passes no ranking signal, and a followed link on a page nobody indexes passes nothing either. Both are checkable in about a minute.
It is worth being blunt about how rarely anyone checks. We tried to verify the outbound link type on 51 directories by fetching a real listing page and reading the rel attribute. We could confirm 22 as dofollow. Two were confirmed nofollow. Twenty-seven we could not verify at all — mostly because the sites returned an error to an automated request, which is ordinary CDN behaviour rather than anything sinister. That is what a deliberate attempt looks like, and it still leaves more than half the list unconfirmed. Any list asserting a followed link beside two hundred entries did not do this.
The per-directory results, with the date each was last looked at, are on the directories page. Use it to build a shortlist you can defend rather than one you copied.
Here is the planning error that costs more than any other, and it is entirely about timing rather than effort.
Directories differ enormously in queue length. Some publish immediately. Some run a weekly batch and will surface you on their schedule, not yours. Some take four to five weeks to review a free submission. If you sit down on launch morning and submit to all of them in one pass, the fast ones publish into your launch day and the slow ones publish weeks later, into an ordinary Tuesday when you are not watching, not sharing anything, and have nothing to point people at.
Everything you did to make launch day land — the post you wrote, the people you told, the announcement in a community you are actually part of — applies only to the listings that went live that day. The rest arrive alone.
The fix is not more work. It is sorting the list by queue length and submitting backwards: slowest first, weeks ahead, so they surface near the day rather than long after it. Nothing about that costs extra time; it just requires knowing the queue before you plan the date.
Which queues are which is documented in the blog, where we went through the submission and pricing pages of the platforms makers actually use and wrote down the free-tier rules as they stand — who queues, who batches weekly, who wants a reciprocal badge, and who has closed the free path entirely.
A caveat on the whole exercise
Launch platforms change pricing and queue rules more often than they change their homepage copy. Anything written down about a queue, here or anywhere else, is a starting point for your own check rather than a fact you can plan a date around unverified.
We run a directory, so treat this section as the interested party explaining its own rules. The rules are public and they are the reason the link here is followed, so they are worth stating exactly.
When you submit a product, it is held rather than published. It does not appear in any public listing, it is not in the sitemap, and its page is marked noindex — but it is reachable by direct link, so you can open it and copy the badge snippet.
The listing publishes when the reciprocal badge is found on your own site. At that moment the launch time is moved to the moment of publication, so you land on the current week’s board rather than one that already closed while you were installing the badge. There is no queue and no reviewer’s afternoon to wait for; the gate is mechanical and you control when you pass it.
That trade is a real cost, not a gift. A badge occupies a slot in your footer and carries an ongoing obligation to leave it there. Decide once whether you are willing to run one, and if you are, put every badge in a single component so a redesign six months from now cannot quietly retire the links you launched for.
Once published, a listing needs 5 qualifying upvotes within 24 hours. Clear it and the listing is permanent, with a permanent followed link. Fall short and the listing is hidden from browse, its page goes noindex, and its outbound link becomes nofollow. You get one free relaunch.
5 is a low bar on purpose — it is a check that a real person cared, not a popularity contest — but it is a real one, and it is the reason the links here are followed rather than sold. Plan for it the way you would plan for anything with a deadline: know when your 24 hours start, and be somewhere you can tell people about it during them.
This one genuinely is a sequence, because each step depends on a queue that has already started running. The dates are relative to the day you want to be visible.
Two of those steps are worth expanding. The one-week-out check is partly about crawlers and partly about the directories themselves: a submission form usually fetches your page and reads your tags to prefill the listing, so a thin or missing set of tags becomes a thin listing on a dozen sites at once. That mechanism, and the fix, is in what a launch directory actually reads off your page. And the launch-day step assumes the badge is already installed if you are launching somewhere that gates on one — it is the difference between publishing that morning and publishing whenever you get round to editing your footer.
Nothing in that list requires a budget, an audience or a coordinated campaign. It requires knowing the queue lengths before you pick the date, which is the only genuinely scarce input.
The day itself is less dramatic than launch-day advice implies, and the useful version of it is short.
The one thing that is genuinely time-sensitive is answering people. Someone who comments on a launch post, files a bug, or asks whether it does the specific thing they need is worth more than the next ten visitors, and they are gone in an hour if nobody replies. Refreshing a leaderboard produces nothing.
A launch that needs a handful of votes needs a handful of people who already know you to see it. Telling them is normal. What is not normal — and what most platforms detect — is manufacturing accounts to vote for yourself. On this site, votes from accounts created minutes ago do not count toward the threshold, which raises the cost of vote rings without punishing anyone who launched honestly.
Most launch days are quiet. A quiet day is not a verdict on the product, and it is not a reason to start submitting to another forty directories in a panic. It means the boards were busy, or the category was crowded, or the day was Tuesday. The page and the link are the same either way.
Half of launch-day disappointment is measuring the wrong number and the other half is measuring nothing at all.
Referral traffic per platform, so you know which ones sent anybody at all and can skip the rest next time. Signups attributable to the day, in absolute numbers rather than as a rate, because the rate on a few hundred visitors is noise. Whether each link exists, what its rel says, and whether the page carrying it is indexable. Which listings are still findable a month later.
The ranking contribution of any individual link, because nobody outside a search engine can see it — not us, not the tool vendors, not the people selling packages of links. Domain authority scores, which are third-party estimates rather than anything a search engine publishes. Your position on a board an hour after it opened. And the comparison between your day and someone else’s screenshot, which you cannot audit and which was selected for being impressive.
The week after a launch is where the actual return shows up, and it is almost entirely unglamorous.
Go through every submission you made and record what happened to it: published or still queued, link followed or suppressed, page indexable or not. Half an hour with the link checker turns a vague sense that you launched into a list of what you have.
Then re-check it periodically. Sites redesign, and a followed link becomes a nofollow one without anyone telling you. A badge removed during your own redesign can retire a link on the other end. Neither failure announces itself, and both are invisible in every dashboard you normally look at.
The listings that are still queued will surface in their own time, and when they do, that is fine. They were never going to convert anyone; they were going to be a page and a link, and they still are.
Submitting to fifty platforms. The evening it costs is the same evening you could have spent on the five whose readers plausibly overlap your users, and the output is a set of near-identical pages that helps nobody. Pick the ones where you can picture the reader.
Launching into a site that cannot be read. A link pointing at a client-rendered shell, or at a page a crawler gets a 403 from, is a wasted link. This is a one-hour check and it belongs before the launch, not after it.
Paying for a followed link. Buying links that pass ranking signal is against Google’s link spam policy; a paid placement is supposed to carry sponsored. Paying for a faster review or an earlier date is a different thing and is perfectly legitimate — know which one you are buying.
Treating the launch as the marketing plan. A launch is one day of borrowed distribution. If there is nothing after it — no reason for the people who came to stay, no second thing to say — then the day was always going to be the whole of it.
LetsLaunch is a launch directory and makes money from listings, so every claim above about what a launch is worth runs against our own interest where it should. None of it asks you to trust us: the queue lengths are on the platforms’ own submission pages, and the link types you can read yourself with curl.