Every list of software directories tells you where to submit. None of them tells you what the link is worth once you are listed, because nobody opens the pages and reads the attribute on the anchor.
This file does that. Fifty five directories, launch platforms and listing sites that carry feedback tool vendors were opened on a real vendor's listing, the anchor pointing at that vendor's own domain was found, and its rel attribute was read. Twenty six of them could be read. Seven pass ranking credit.
directory,rel_attribute,verdict
spotsaas.com,noreferrer,passes ranking credit
featuredcustomers.com,noopener noreferrer,passes ranking credit
stackshare.io,noopener noreferrer,passes ranking credit
saashub.com,nofollow,marked nofollow sponsored or user generated
alternativeto.net,nofollow noopener,marked nofollow sponsored or user generated
g2.com,,no outbound link to the vendor
Take the file. It may be copied, quoted and used commercially with no attribution and no link back.
None of it needs attribution. If you would rather point at it than copy it, the address is https://votefirst.app/en/data/feedback-tool-directory-links and the figures on it were read on 16 September 2026.
What was measured, exactly
For each directory, a listing page for a vendor in this category was opened as an ordinary reader. Canny was the vendor wherever a Canny listing existed, and where it did not, Productboard, Fider, Frill, Nolt or UserJot stood in. A second pass repeated the test on a listing for a large product outside this category, so that a directory's policy was read twice rather than once.
Every anchor on the page was collected, and the ones kept were those whose address points at the vendor's own domain. Anchors that merely mention the vendor in their title text were discarded, which matters more than it sounds: a first pass that matched on title text reported four directories as passing credit that do not.
The verdict on each row is then mechanical.
| Verdict | Rule |
|---|---|
| Passes ranking credit | At least one anchor to the vendor's domain carries no nofollow, sponsored or ugc value |
| Marked nofollow, sponsored or user generated | Every such anchor carries one of those three |
| No outbound link to the vendor | The listing exists and contains no anchor pointing at the vendor's own domain |
| Not measured | The page refused a reader, or no listing was found at the address tried |
How many software directories are dofollow?
Seven of the twenty six that could be read.
| Verdict | Directories |
|---|---|
| Passes ranking credit | 7 |
| Marked nofollow, sponsored or user generated | 9 |
| No outbound link to the vendor at all | 10 |
| Not measured, the page refused a reader | 16 |
| Not measured, no listing found at the address tried | 13 |
The seven are appsumo.com, featuredcustomers.com, libhunt.com, spotsaas.com, stackshare.io, subscribed.fyi and uneed.best. Two of those carry a condition. LibHunt lists repositories, so only an open source project qualifies. Uneed's own page states that a free launch keeps its followed link only above a launch day score, so the link is conditional on an audience turning up.
That leaves five directories in this whole sample that hand a closed source product a followed link for filling in a form.
Which of the big review sites pass credit?
None of the ones that could be read, and the reason is more interesting than a nofollow.
G2's review page was read from a web archive copy, because the live page refuses an ordinary reader. The archived copy is seventy eight thousand bytes of markup and carries no anchor whose address points at the vendor's own domain at all. The route to the vendor's site is a tracked redirect rather than a link. SourceForge does the same thing.
Capterra, GetApp, Software Advice, TrustRadius, Crozdesk, SoftwareSuggest, Crunchbase and Wellfound all refused a reader and are recorded as not measured rather than guessed at. That is a real limitation of this file and it is stated rather than papered over.
Does a nofollow link still do anything?
Yes, and this file is not the place that decides how much.
Google's own documentation describes nofollow as a hint for ranking and indexing purposes rather than a directive. So a nofollowed listing is not worth nothing. It is also not the thing that moves a page from the second results page to the first, and any plan that treats a directory submission campaign as an authority strategy is working from an assumption nobody checked.
The more useful way to read the table is commercial. SaaSHub carries every vendor in this category. AlternativeTo's page for one of them lists eighty five competitors. Being the one product absent from those pages is a problem a buyer notices long before a crawler does, and that is a good enough reason to submit.
Why does a feedback tool publish this?
Because the same question came up here and the answer had to be measured before any time was spent on it.
This site publishes its own measurements of this category: the link authority of seventy domains, the page inventory of sixty nine sitemaps, and thirty two vendors' prices read by hand. This is the fourth file of the same kind.
If you are choosing a feedback tool rather than researching directories, every vendor in this category is priced on one page.
How to repeat this
The script is forty lines and it runs in under a minute. It takes a list of directories, a vendor page address for each, and the vendor domain to look for. It fetches each page, finds the anchors whose address matches that domain, reads the rel attribute on each and reports the verdict.
The one thing worth copying from it is the part that got the first run wrong. Match on the anchor's address, never on its text, and treat an anchor with no rel attribute and an anchor with rel="noopener noreferrer" as the same thing, because neither of them withholds credit.
Read on 16 September 2026. A directory can change its policy without announcing it, so the date on every row is part of the row.