If you've worked with an SEO agency, you've probably seen "directory submissions" as a line item on a monthly report, usually bundled under "link building" alongside the actual hard work: outreach, content, digital PR. Submit to enough AI tool directories and someone will eventually tell you it's "great for backlinks." It's worth being precise about what that claim actually means, because most directory links aren't backlinks in the sense that matters for rankings, and the ones that are don't move the needle the way a founder hoping for a quick SEO win usually expects.

What a "backlink" actually needs to do to help rankings

A link helps SEO when Google follows it and treats it as a vote of confidence from the linking page. Google's own developer documentation is specific about this: links marked nofollow, sponsored or ugc "will generally not be followed," meaning Google doesn't pass ranking credit through them, full stop, regardless of how authoritative the linking site is. Since 2019, Google has treated these attributes as hints rather than absolute rules, meaning it may occasionally crawl and consider a nofollow link anyway, but the official guidance is clear that you shouldn't plan a link strategy assuming otherwise. Most free directory listings use nofollow by default, specifically because directories are a known target for link spam and Google has said as much for years.

That single fact kills most of the "directories are great backlinks" pitch before it starts. A listing that shows your name, logo and description to a directory's visitors is real, useful discovery. A nofollow link sitting in that listing is not the same thing as the backlink an SEO means when they talk about link building, and treating the two as interchangeable is how a lot of directory-submission advice goes wrong.

How to actually check if a specific link is followed

Don't take a directory's marketing copy at its word either way. Right-click the listing page (or view the page source) and search for your own link's HTML; a followed link is a plain <a href="..."> tag, while a nofollow link includes rel="nofollow" somewhere in that tag, sometimes alongside sponsored or ugc. Google Search Console's Links report will also show your site's external backlinks, though it doesn't distinguish follow status there either, so the page-source check is the only way to know for certain without a third-party crawler tool. It takes about thirty seconds per listing and settles the question definitively, which beats trusting a directory's FAQ page or a third-party "directory review" site's claim about its own follow status, since those claims disagreed with each other for several of the directories we checked while researching our directory submission guide.

Google names this exact pattern in its own spam policy

There's a stronger statement than "nofollow doesn't pass value" worth knowing here. Google's official spam policies page lists "low-quality directory or bookmark site links" by name, in a bulleted list of link schemes, practices Google considers an attempt to manipulate rankings through artificial linking rather than earned endorsement. That doesn't mean every directory listing is spam, a relevant, curated, editorially-reviewed directory is a different thing from a mass-submission bookmark farm, and Google's own guidance distinguishes quality within the category rather than banning directories outright. But it does mean the specific tactic of submitting to dozens of low-quality directories specifically to accumulate links is a pattern Google has named and watches for, not a gray area you're getting away with because nobody's looking.

What the actual data says about backlinks and rankings

Ahrefs ran a study across one million SERPs specifically to check whether Google's own public statements about links mattering less held up against real ranking data. The honest finding: backlinks still correlate with rankings, but the relationship is weak by normal statistical standards, a Spearman correlation around 0.25 for both raw backlink count and referring domains, and that correlation has actually declined since 2019, from roughly 0.27 down to the 0.22-0.24 range. Ahrefs is careful to flag that correlation isn't causation, a page with more links also tends to be older, more established and better promoted in ways that independently help rankings. Links also matter more for some query types than others; the correlation runs stronger for local search (around 0.33) than for the kind of competitive, non-local keywords most SEO advice is written around.

None of that says backlinks are worthless. It says the relationship is real but modest, has been shrinking for years, and varies a lot by context, which is a very different claim than "get more backlinks and you'll outrank the competition," the pitch behind most directory-submission advice.

Why directory links specifically carry less weight than the pitch implies

Set aside the nofollow issue for a moment and assume a directory does pass a followed link. Directory links still carry two structural problems that reduce their value relative to an editorial link from a real article. First, they're contextually thin: a link sitting in a database row next to a hundred other listings in the same category carries far less topical signal than a link embedded in a paragraph of a relevant article that's actually about your specific product. Second, directory links decay faster than most link types. Ahrefs' link-rot research found that sites with domain authority above 45 lose about 10.9% of their referring links in the first year, while sites below that threshold lose closer to 15.9%, and most AI tool directories sit well below DA 45. A link you got from a directory submission in January has a real chance of being gone, broken, or sitting on a page Google has stopped crawling by the following January, which is a cost worth weighing against the hour it took to submit.

A worked example of what this actually costs and earns

Say a new AI tool submits to fifteen free directories in a single afternoon. If even a generous half of those pass a followed link, roughly seven real backlinks land on a site that previously had none. Plugging that into Ahrefs' own correlation numbers, a referring-domain correlation around 0.25, means seven new, mostly low-authority referring domains are a genuinely small input into a ranking factor that's already weak on its own, and that's before accounting for the 15-16% of those links likely to decay within the first year on lower-authority sites. Compare that to the effort of earning a single mention in one relevant, well-trafficked industry article or newsletter, which both counts as one editorial referring domain with real topical relevance (a stronger signal than a directory listing by nature, not just by follow status) and tends to decay far more slowly, since real publications don't disappear or get deindexed at the rate marginal directory sites do. The fifteen-directory afternoon isn't wasted, but measured purely as a link-building investment against that one editorial mention, it's a rounding error.

Where directories genuinely help, separate from the backlink question

None of this means directories are a waste of time, it means the actual value is somewhere other than classic link equity. Our directory submission guide covers which ones are worth claiming and why; the honest case for them rests on three things that have nothing to do with PageRank. A real, consistent listing is a small piece of entity SEO, one more place where a tool's name, description and URL show up the same way, which helps both Google and AI models resolve what the product actually is. A handful of the directories we reviewed, G2, Capterra, AlternativeTo among them, send genuine discovery traffic from people actively comparing or shopping for a tool like yours, traffic that has nothing to do with whether the link is followed. And a directory's own citation-worthiness inside AI answers, being one of the sources an AI system might reference when someone asks "what tools exist for X," is a separate channel from Google rankings entirely, governed by its own logic rather than classic SEO signals.

The specific link-attribute situation across the directories we checked

Worth revisiting in this context: several directories in our own roundup don't give a straight answer on whether their links are followed at all. TAAFT's own pages don't spell it out, and third-party reporting disagrees about whether standard listings there are nofollow or dofollow. AlternativeTo's situation is similarly inconsistent depending on who's reporting it and possibly on how established a listing becomes over time. Dang.ai's free tier specifically requires you to add a crawlable backlink to dang.ai on your own site as a condition of the free listing, which is a link exchange, not a one-way vote of confidence, and worth factoring into whether that "free" listing is actually free in the sense that matters. If a directory's own answer to "is this link followed" is vague or inconsistent, the safest assumption for planning purposes is nofollow, and any SEO value beyond that is a pleasant bonus rather than something to count on.

AI citations run on a different logic than backlinks entirely

It's worth separating "does this link help my Google ranking" from "does this listing help an AI system cite or recommend me," because they're genuinely different questions with different mechanics, not two names for the same thing. Google's ranking algorithm, as the Ahrefs data shows, weighs backlinks as one of many weak-to-moderate signals. An AI system deciding what to cite in an answer isn't running a ranking algorithm against a link graph at all; it's retrieving and synthesizing information from whatever sources it can find and judging relevance and credibility more directly, closer to what our information gain piece describes for how Google's own core updates increasingly evaluate content. A directory listing with a nofollow link can still be a source an AI system finds and cites, or a place it confirms your product's name and category exist, entirely independent of whether that link would ever move a classic Google ranking. Conflating these two systems is a common mistake: treating "it's nofollow, so it's useless" as true for AI citation potential the same way it's true for PageRank, when the two systems don't share that mechanism at all.

What this means for how you should actually prioritize

If link equity specifically is the goal, directory submissions are a weak, low-effort supplement at best, not a strategy. The actual lever that moves the 0.25-correlation needle Ahrefs measured is earning editorial links, a credible site covering, reviewing or linking to you in a real article because the content warranted it, which is exactly the harder, slower work our link building strategies guide covers. A single genuine mention in a relevant, topically-connected article is worth more for rankings than a dozen nofollow directory listings, even though the directory listings take an afternoon and the editorial mention might take months of outreach or genuinely remarkable content to earn.

If discovery, entity consistency, and AI-answer citation potential are the goal instead, directories are worth the afternoon regardless of follow status, because that value doesn't depend on PageRank passing through the link at all. The mistake is doing the submissions while believing they're accomplishing the first goal (meaningful SEO link equity) when what they're actually accomplishing is the second (discovery and entity signal). Both are legitimate reasons to submit. Only one of them is actually true of what a nofollow directory listing does.

Do any AI tool directories give a real, followed backlink?

Some claim to, and a few third-party directory-review sites report specific ones do under specific conditions (a paid tier, an established listing, a particular directory's own policy), but the reporting is inconsistent enough across the directories we checked that you shouldn't submit expecting a followed link as the primary payoff. Treat any followed link you do get as a bonus on top of the discovery and entity value, not the reason you submitted.

Is paying for a premium directory listing worth it just for the backlink?

No, not based on the backlink alone. If a directory's paid tier also gets you real traffic, a newsletter mention, or a featured placement in front of an audience actually looking for a tool like yours, that can be worth paying for on its own terms. Paying specifically because you've been told it's "great for SEO" through the link alone isn't supported by how nofollow links actually work or by how weak the overall backlink-to-ranking correlation already is.

Why do so many SEO guides still push directory submissions as a link-building tactic?

Partly inertia from an era before Google treated nofollow more strictly, and partly because directory submissions are an easy, scalable deliverable for an agency or a guide to recommend, cheap to execute, easy to show as "done" in a report, regardless of whether the underlying value is real. It's also the exact tactic the directory-submission content-farm ecosystem, covered in our directories guide, has a direct financial incentive to oversell, since many of those sites sell the submissions themselves.

Does a nofollow link do absolutely nothing for SEO?

Not literally nothing. Google has said since 2019 that nofollow is a hint, not an absolute rule, and it may still crawl and occasionally consider a nofollow link in some circumstances, which is part of why Google doesn't call nofollow links worthless outright. The honest summary is that you shouldn't plan a link-building strategy around that possibility, since Google's own guidance and the behavior it describes point toward "generally not followed," not "sometimes followed if you're lucky."

Should I stop submitting to directories altogether, given all this?

No, just submit for the right reason. The free, legitimate ones take an hour combined and genuinely help with discovery and entity consistency regardless of link-follow status. What you should stop doing is treating directory submission as a substitute for actual link building, or expecting a measurable ranking jump from a nofollow listing on a DR-30 directory site.