Are Startup Directory Submissions Worth It? I Checked the Links Myself

I run two product directories. This site is one of them, and Makers Page is the other. I checked both this morning, and every outbound link to a listed product on both of them carries rel="nofollow".

So if you're submitting to either of mine for the backlink, you are not getting one. That's worth saying out loud before anything else in this post, because the entire "submit your startup to 100 directories" industry is built on people not checking.

What are you actually buying with a directory submission?

Strip out the pitch and a directory listing can do exactly three things for you. It can send you real visitors. It can leave behind a page that ranks or gets cited when someone searches for your category. It can pass you a backlink.

Most submission advice quietly implies you get all three from every listing. You almost never do. The three come apart constantly, and which one you're getting changes the answer to "is this worth an afternoon" completely.

The link is the one people obsess over, and it's also the one that's easiest to verify and hardest to fake. It takes about forty seconds per directory. So I went and did it.

Are directory backlinks dofollow or nofollow? I went and checked

Here's the method, so you can repeat it on any directory you're considering: fetch a real listing page with curl, find the anchor pointing at the product's own domain, and read its rel attribute. No tools, no subscriptions, no trusting a comparison table written by someone who wants your $199.

What I found on September 2, 2026:

• SaaSHub nofollows it. Every outbound link on the product page I checked, including the main "visit website" button, carries rel="nofollow". The listing had 41 plain rel="nofollow" anchors on it.
• BetaList doesn't link to your domain at all. The button points at an internal URL, /startups/your-slug/visit, which 301-redirects to your site with ?ref=betalist appended. Redirects can pass signal, so this isn't nothing, but it isn't the clean direct link people picture either, and one of the anchors pointing at it is explicitly rel="noopener nofollow".
• Uneed actually gives you the dofollow. The outbound anchor on the tool pages I checked carries rel="noopener" and nothing else, pointing straight at the product's real domain with a ?ref=uneed.best parameter. They advertise a do-follow backlink and the HTML backs it up, which after the rest of this exercise felt genuinely notable.
• Product Hunt returned a 403 to a plain curl request, so I couldn't check it the same way and I'm not going to guess.
• This site's tools directory uses rel="nofollow noopener noreferrer" on every outbound tool link. That's a line in app/tools/[slug]/page.tsx, not an accident.
• Makers Page uses rel="nofollow noopener" on the product link.

Two directories I run, both nofollow. One well-known launch platform routing through a tracked redirect. One popular comparison site nofollowing everything. One directory doing exactly what it says. That's the actual landscape, and it took twenty minutes to establish.

I'm not calling any of this dishonest. Nofollowing outbound links on a page where anyone can submit anything is the correct engineering decision, which is exactly why I made it on my own sites. It's just very different from what the submission-service landing pages describe.

What does a nofollow link actually do for your SEO?

For rankings, close to nothing directly. For Domain Rating, literally nothing: Ahrefs only counts followed links when it calculates DR, so a nofollowed listing on a DR 70 directory moves your number by zero. Not a little. Zero.

That's the first place the "100 directories" math falls apart. If most of them nofollow, and only the first link from any given domain counts anyway, then the difference between 20 submissions and 100 submissions is mostly the eighty afternoons.

And even for the dofollow ones, the ceiling is lower than the pitch suggests. Gary Illyes, an analyst on the Google Search team, said at Pubcon Pro in Austin in 2023 that links are not a top three ranking factor, and that Google needs very few links to rank pages. He cited a page ranking at position one for Porsche cars that Google had only found through a sitemap, with no links pointing at it at all. That's one Googler's framing rather than a documented rule, but it's the closest thing to an official read on how much a link is worth in 2026, and it's a lot less than the link-building industry needs you to believe.

What does Google actually say about directory links?

It's in the spam policies, and it's specific. Google's spam policies documentation lists "low-quality directory or bookmark site links" as an example of link spam, right alongside paid links and "using automated programs or services to create links to your site."

Read that second one again if you're about to buy a bulk submission package.

The important nuance is what happens when Google decides your links qualify. With the December 2022 link spam update, Google announced it was "leveraging the power of SpamBrain to neutralize the impact of unnatural links on search results." Neutralize, not punish. The usual outcome isn't a penalty and a scary message in Search Console. It's that the credit those links were passing quietly stops existing, and your rankings sit exactly where they'd sit if you'd never built them.

That's the worst-case scenario worth planning around, and it's more boring than the horror stories: you spend a weekend on 100 submissions, most of them are nofollowed so they never counted, and the handful that did count get neutralized as a cluster of low-quality directory links from an automated service. Net effect on rankings, zero. Net effect on your Saturday, gone.

Google also says plainly that having these links is fine as long as they're qualified with rel="nofollow" or rel="sponsored". Which is precisely why every well-run directory nofollows outbound links, and why the ones promising you SEO value from a $5 listing are the ones I'd look at hardest.

Are the $99 to $199 submission services worth it?

The going rate, as of August 2026, is roughly $99 for 30 directories, $149 for 60, and $199 for 100 or more, with per-directory costs across the market running from about $0.50 up to nearly $6. Some do it manually, some automate it, most promise a report at the end.

Run the math with what we just established. Say a generous 30% of the directories in a 100-pack pass a real followed link. That's 30 links. Deduct the ones from domains with no traffic and no crawl budget, deduct the ones Google files under "low-quality directory links" as a cluster, and the realistic outcome is a handful of links that survive, pointing at a site where links weren't the top-three constraint in the first place.

The referral traffic case is stronger, and it's the one nobody sells you on because it's harder to promise. A listing on a directory people actually browse can send real visitors who are in shopping mode. That's a genuine, if small, channel. But it comes from maybe five to ten directories with real audiences, not from a hundred, and you can submit to those five yourself in an afternoon for free.

There's a version of this purchase that's honest: you value your Saturday at more than $199 and you want the ten good listings without doing the data entry. Fine. That's the same trade as buying a boilerplate to skip the plumbing, and it's legitimate. Just buy it knowing you're buying back time, not buying rankings.

Which directories are actually worth submitting to?

I'd rather give you the filter than a list, because the list rots and every list you find was written by someone who wants to be on it.

Four questions, in order:

1. Does anyone actually browse it? Open the site as a visitor, not a submitter. Is there recent activity, real comments, products from this month? A directory with no audience is a page, not a channel.
2. What does the outbound link do? Run the check from earlier. Forty seconds. If they promise a dofollow backlink in the pitch and nofollow it in the HTML, that tells you what else they're willing to say.
3. Is the listing page indexable and server-rendered? If the product page only exists after JavaScript runs, crawlers and AI bots are fetching an empty shell. This is the same failure mode I ran into writing about llms.txt, and it silently invalidates the whole "durable page" argument.
4. Is it relevant to what you built? A general "SaaS tools" dump is worth less than a directory whose entire audience showed up for your specific category, on both the traffic side and the link side.

Applied to the ones I checked: Uneed passes on the link and has an audience. BetaList is worth it for the specific case it's built for, which is pre-launch and waitlists, and not for the backlink. Product Hunt is worth it for launch day attention and the durable product page, whatever its link policy turns out to be. This site's own directory will send you visitors and will not send you link equity, and now you know that before submitting instead of after.

That's four to six submissions, not a hundred. The rest of the spreadsheet is busywork with a progress bar.

Do AI search engines cite directories?

This is the one part of the argument that's actually gotten more interesting, and it's the reason I'd still submit to the good ones.

When someone asks ChatGPT or Perplexity for "the best tool for X," the model has to pull passages from somewhere. Category pages and comparison listings are exactly the shape of content that gets retrieved for that kind of query, which means a listing on a well-indexed, server-rendered directory can put your product inside an answer you'd otherwise never appear in.

I want to be careful here because I've already written about how much of the AI-visibility advice going around is unmeasured. I don't have citation data proving directory listings drive AI mentions for a specific product, and I haven't seen a study that isolates it. What I can say is that the mechanism is plausible in a way the backlink argument no longer is, and that it depends entirely on question three above. A client-rendered listing page contributes nothing to it.

The 30-minute version

If you were about to spend a Saturday on submissions, do this instead:

• Pick five directories where your actual buyer would plausibly be browsing. Not five with the highest DR. Five with humans.
• curl each listing page and read the rel on the outbound link. Write down what you find. That's your real expected return, not what the pricing page says.
• View source on one of their product pages and search for a product's description text. Not there? The page is invisible to crawlers and AI bots, and the listing is worth traffic only.
• Submit to those five, by hand, with a description written for that specific audience rather than pasted across all of them. Identical copy across a hundred sites is the pattern Google's scaled-content and link-spam policies are both built to catch.
• Put the remaining six hours into something that produces a page other people want to link to without being asked.

What the hundred-directory spreadsheet is really for

It's for feeling like distribution is a task with a finish line.

That's the same trap I keep walking into from different directions. Shipping llms.txt to feel like AI search was handled. Buying a boilerplate to skip the uncomfortable part of starting. A submission checklist is the purest version of it: a hundred rows, a satisfying number of green checkmarks, no ambiguity about whether you did the work, and almost no relationship between the checkmarks and anyone finding out your product exists.

Distribution is still the hard part, and it stays hard specifically because the useful version of it doesn't come with a progress bar. Five good listings, done properly, are worth more than ninety-five more rows in a spreadsheet, and the afternoon you get back is the actual return on this post.

I'll keep nofollowing the outbound links on both of my directories. Submit to them because people browse them. That part is real, and it's the only part I can honestly promise.