Startup Directories in 2026: We Checked 1,276 of Them

Why we did this, and our own interest

Every founder who searches for a list of startup directories hits the same problem: almost every list ranking for that search is published by a directory or a submission service. They are grading themselves, and they have a reason to make the list long and the links sound valuable.

So we checked the directories from the outside, with a method anyone can repeat. We read the actual HTML of real listing pages instead of trusting what a list or a directory says about itself.

We are a party here too, so we will say it up front. LaunchList runs launch.li, a launch directory. We excluded launch.li and getlaunchlist.com from every number in this post. Of 1,278 candidates we found, those 2 are ours, which leaves 1,276.

This is part 1. In part 2 we will launch one real product on the top 30 or so directories from this audit and report the visitors and signups each one sends. We will update this post when that is done.

The short answer

• Under half of all directories are active. 596 of 1,276 (47%) published a listing in the last 90 days. 74 (6%) are dead, and 127 (10%) are dormant.
• The popular lists are stale. 35% of the directories on public lists were active within 90 days, against 66% of the ones we found only by following links. 479 of the directories we measured appear on no public list we read.
• The "dofollow" claims mostly do not hold. Of 95 claimed-dofollow directories we could read, 26 (27%) gave a followed link on every sampled listing. 63 were mixed.
• Followed links are rare. 142 (11%) of all directories gave a followed product link on every sampled listing. 700 (55%) had no measurable product link at all.
• Pricing is hidden. 888 (70%) state no price before you sign up or start submitting.

How we checked

How the list was built

1. Public lists. We started from 35 public lists: directory roundups, GitHub "awesome" lists and lists published by submission services. That gave 797 directories after removing duplicates and our own two sites.
2. One list in full. ScrollLaunch's list only renders its first page to a simple fetch, so we loaded the full list in a real browser.
3. Following links between directories. Launch boards link to each other through "featured on" badges, partner strips and footers. Any site linked from 2 or more directories whose homepage invites submissions became a candidate. We ran this for three rounds. New finds fell from 267 in round one to 173 in round two to 39 in round three, so we stopped there.
4. Filtering. We filtered out product sites, AI vendors and domain-rating checkers by their homepage text, and set aside 7 more submission services (finding 8).

That gave 1,276 directories measured, 479 of which appear on no public list we read.

We sorted them by type:

What we checked on each directory

• The homepage: alive, parked, redirecting, or failing TLS.
• robots.txt.
• The submit page and any pricing text on it.
• The newest listing date we could find.
• Up to 6 real listing pages. Where a directory had ordinary listings, we skipped featured or sponsored cards so we were looking at what a normal submission gets.

On each listing page we found the link to the listed product and read it from the HTML. We recorded its rel attribute (nofollow, ugc or sponsored), whether it went through the directory's own redirect, and whether the listing page itself carried noindex or pointed its canonical somewhere else. We sampled 4,684 listing pages in total. Pages that only build their content with JavaScript were rendered in a headless browser: 663 of them.

We also looked up domain age from public RDAP records, which we found for 815 domains. The median age is 28 months: 230 are under 12 months old and 378 under 24. We checked each directory against the free Tranco top-1M list of 2026-10-13, and 239 directories are in it. Tranco is a popularity rank, not a traffic number, so we use it only as a rough authority signal.

A site that blocked our requests (a 403 or 429 response, or a bot wall) is counted as blocked, never as dead.

To check our own crawler, on the first pass (the 797 directories from public lists) a separate script with no shared code re-fetched 40 random listing pages. Where it could find the same product link (31 of 40), it agreed with the crawler 31 out of 31 times. In the other 9, the page had changed or the link was built by JavaScript.

What this method cannot tell you

• We did not log in, submit a product or pay anyone.
• "Not marked featured" does not prove a listing was free. Some directories may give followed links to paid listings that look ordinary.
• We cannot see whether Google has indexed a page. We can only see whether the page tells Google not to.
• Pricing could not be read for most sites (finding 5).
• The link-following stopped after three rounds, when new finds dropped to 39. New boards appear every week, so we will re-run this audit quarterly.

Finding 1: under half are alive

We dated each directory by its newest listing. Here is where the 1,276 landed:

Add the first two rows and you get 596 directories (47%) with signs of life in the last 90 days. Everything else is dormant, dead, blocked, undated or unverifiable.

Among the dead, 20 sites were parked domains and 10 redirected to a different domain entirely.

The type matters. Launch boards were the healthiest large group: 162 of 265 were active within 90 days, with 12 dead. Startup directories did much worse: only 39 of 141 were active within 90 days. AI tool directories sat in between, with 171 of 353 active and 23 dead.

Finding 2: two populations, and the popular lists are the stale one

This is the finding that surprised us most. The directories split into two groups depending on how we found them, and the groups look very different.

The directories we found by following links are also young. Their median domain age is 11 months, and 176 of the 347 with a known age are under a year old.

What this means in practice: the popular lists are stale. They keep naming directories that stopped publishing long ago, while the live, new directories are mostly missing from them. If you build your submission plan from the top search results for "startup directories", you are starting from the weaker half.

One caution about the comparison. Following links favors live sites by design: an active launch board is the kind of site that shows up in another board's "featured on" strip, and a dead one is not. So part of the gap comes from how each group was found. The direction is still clear, but the size of the gap is partly a product of the method.

The lists also copy each other. While building the candidate list we saw several large lists repeat one another, and four big roundups descend from one older list that still includes 2016-era press outlets. So when many lists name the same directory, that is often copying, not confirmation that the directory is good.

Finding 3: the "dofollow" claims mostly do not hold

"Dofollow" is the selling point on most directory lists. Across the 35 source lists, 200 directories were claimed to give a dofollow link. Only the public lists made these claims, so this finding covers them alone. We tested the claim.

For 95 of those 200 we could actually read the product link on listing pages. Here is what we found:

So 26 of 95 (27%) matched the claim fully. The 63 mixed directories typically give a followed link only for paid or featured tiers, or only on some listings. If a list says "dofollow" and the directory turns out to be mixed, the free listing you get may well be the nofollow kind.

For the other 105 claims, we could not find a product link on any sampled listing at all, so the claim cannot be confirmed either way.

Finding 4: link class across all 1,276

Looking beyond the dofollow claims, here is the link picture for every directory we measured:

"No measurable product link" covers dead and blocked sites, directories where we found no listings, and listing pages with no outbound link to the product. That last case exists more than you might expect: some directories show your product on their page but never link to it.

Activity and links overlap less than you would hope. Of the 142 fully followed directories, 102 were also active in the last 90 days. Of the 356 mixed directories, 276 were active.

We also checked whether listing pages tell search engines to stay away. 14 directories mark every listing page we sampled noindex, and 98 do so on some listings. 729 were indexable on every listing we sampled. A followed link on a noindex page is worth much less than it sounds, so this is worth a quick look before you submit.

Finding 5: pricing is opaque

We tried to read the price from each directory's submit page. We could do that for 388 of 1,276:

That is 70% of directories that state no price before you sign up or start submitting. Some of them are free. Some reveal a fee or an upsell halfway through the form. You only find out by starting.

Where prices were stated, the lowest paid price we saw was $1 and the highest was $299 for one listing.

None of this means a paid directory is a bad deal. It means you should know the price, and what it buys, before you put in the time.

Finding 6: badge and backlink requirements

82 directories ask you to place their badge or a link back to them on your site. Often it is the condition for a free listing, or for getting a followed link instead of a nofollow one. Examples include fazier.com, twelve.tools, startupfa.st, huzzler.so and launchdirectories.com.

A reciprocal link is a trade. You give a directory a link from your site, and in return you get a listing or a better link from theirs. That can be a fair deal, especially on an active board with real visitors. Founders should just know they are making it, and decide whether they want that badge on their homepage.

Finding 7: shared infrastructure

We compared analytics IDs, ad publisher IDs and footer links across directories. 14 groups, covering 46 sites, share an AdSense ID, an analytics ID or a link to the same site in their footers. The larger groups:

• One AdSense publisher ID across 9 sites: aitoolsmagazine.com, allesora.com, cssnectar.com, feedough.com, flattrendz.com, linkvertex.com, nextbigwhat.com, startupsoflondon.com and toolspedia.io.
• The same site (direasy.com) linked in the footer of 6 launch boards: makerpocket.com, smolfind.com, smolhunt.com, smolmaker.com, smolshow.com and smoltop.com.
• One AdSense ID across 4 sites: dayslaunch.com, starterbest.com, startuptostartup.com and theonestartup.com.
• One AdSense ID across 3 sites: ainewtool.site, newtool.site and saastool.site.
• One Google Analytics ID across 3 sites: dododirectory.com, postioo.com and saasbison.com.

We also found 36 sites with the identical homepage title pattern <Name> - Discover the Best Tech Products, which points to one shared off-the-shelf directory template.

All this tells you is that these sites share infrastructure or a template. It says nothing about the quality of any one of them, and several appear in our recommended lists below. It matters for one reason: if you submit to every site in a group, you may be reaching fewer separate audiences than the count suggests.

Finding 8: submission services

The public lists led us to 11 services that sell "we will submit you to 100 to 500 directories", and following links turned up 7 more: 18 in all. Examples include launcheverywhere.com, submitsaas.com, 300aidirectories.com and startupsubmit.app.

We did not test these services, and we are not saying any of them misrepresents what they do. But our numbers have a direct consequence for anyone buying one. Across the 1,276 directories we measured, 596 showed activity in the last 90 days, only 142 gave a followed product link on every sampled listing, and 700 had no measurable product link at all. And if a service builds its list from the popular public lists, it is drawing from the group where only 35% were active. A list of 500 directories will, by those numbers, include many dead, dormant or not-followed sites.

If you are considering a service, ask for its list, and check a handful of the names against the findings above before you pay.

Where to submit in 2026

This is the useful part. Below are the directories that passed our checks on 2026-10-14.

Print this with any table you share: "followed" means the links we sampled on 2026-10-14. Policies change, and "mixed" usually means the followed link is a paid or featured perk.

Launch boards: active, with followed or mixed links

119 launch boards published a listing in the last 90 days and gave followed or mixed product links on our samples. Here are the 20 best known, ranked by Tranco popularity and then by how many lists and directories name them:

The other 99 passed the same checks. We left them out here to keep this page readable; they will be in the full sortable list we publish with the quarterly update.

If a board is mixed, assume the free listing sends visitors first and link value second.

Directories: active, followed on every sample, free or freemium

54 directories were active, gave a followed link on every listing we sampled, and are free or freemium according to their own submit page. The 20 best known, ranked the same way:

"No" in the badge column means we found no badge request on the submit page, not that none exists. A directory that is followed and free but has nothing to do with your product will not send you the right visitors.

What about Product Hunt and BetaList?

Neither made the followed lists, and that is fine. Product Hunt marks product links ugc. BetaList sends visitors through its own redirect. Both are still worth it for visitors, not for link value. We cover how to use them, and the boards that work like them, in our guide to Product Hunt alternatives.

A simple decision rule for founders

A directory can give you three things: visitors, a link, or nothing. It always costs you effort. Here is how we would decide.

1. Decide what you want first. If you want visitors, an active launch board with a real audience beats a followed link on a site nobody reads. If you want link value, it has to be a followed link on an indexable page of an active site. Most directories give you one or neither, rarely both.

2. Check three things before you submit. Is the newest listing recent? Is the link on an existing listing followed, and does it go straight to the product? Is the price stated, and is there a badge requirement? You can check all three in a few minutes by opening one listing page and viewing its source.

3. Submit to 10 to 20 active directories, not 500. Pick the ones that match your product and your buyers, from the tables above or your own checks. A short list you actually prepare for, with a good description and real screenshots, beats a long list of dead sites.

4. Send visitors somewhere that keeps them. A directory visitor who lands on a homepage with no signup form is usually gone for good. Send them to a page that captures the email: a waitlist page before launch, a signup page after. Directory traffic comes in small waves, so each visitor you keep matters.

For the wider picture of where to launch beyond directories, see where to launch your startup and our list of places to promote your startup for free.

Disclosure, once: LaunchList is pre-launch waitlist software for founders, and a waitlist page is exactly where we think directory visitors should land before you launch.

What part 2 will measure

Links are only half the question. The other half is whether a directory sends anyone who cares.

In part 2 we will launch one real product on the top 30 or so directories from this audit and measure, per directory:

• visitors sent
• signups from those visitors

We will update this post with the results.

FAQ

Are startup directories worth it in 2026?

Some are. 596 of the 1,276 we checked (47%) published a listing in the last 90 days. The active ones can send a small wave of interested visitors and a durable listing. Pick 10 to 20 active directories that fit your product instead of submitting everywhere.

Do directory backlinks help SEO?

Less often than the lists claim. Of 95 directories advertised as dofollow where we could read the listing links, only 26 (27%) gave a followed link on every listing we sampled. Across all 1,276 directories, 142 (11%) did, and 356 were mixed, which usually means the followed link is a paid or featured perk. Another 98 directories mark some listing pages noindex. Treat link value as a bonus, not the reason to submit.

How many directories should I submit my startup to?

10 to 20 active ones that match your product and audience. Most directories on the popular public lists are dormant, dead, blocked or give no measurable product link, so a list of hundreds mostly adds work, not visitors.

Are directory submission services worth it?

Check what is on the list first. We found 18 services selling bulk submission, often to 100 to 500 directories. Across the 1,276 we measured, 596 showed activity in the last 90 days and only 142 gave a followed link on every sampled listing, so a list of 500 will include many sites that send nothing.

What does dofollow mean on a directory?

It means the link from your listing to your site carries no rel attribute that tells search engines to ignore it, such as nofollow, ugc or sponsored, and it points straight at your site rather than through the directory's own redirect. Many directories give a followed link only to paid or featured listings, so check an ordinary listing, not the featured cards at the top.