We checked who ranks when you search for alternatives to a tool. Reddit won nine of twelve.

When somebody has a tool and wants a different one, they search for its name plus "alternatives". It is one of the most honest queries on the internet: nobody types it casually.

We measured who Google actually shows. Twelve searches, one for each of twelve tools we list, US English, roughly the first thirty organic results each, read on 9 August 2026. That is 336 results across 248 different sites — no publisher owns this space.

Corrected the same day this was published. The first version of this article read only the first seven to nine organic results per search, which is what the search API returns by default, and reported "96 results" as though it were the top ten. On that sample we found one directory and said so. Re-read at full depth, there are fourteen directory results, and one claim in the original — that ranking pages tend to advertise first-hand testing — turned out not to survive the larger sample at all. Both corrections are below. The shallow read did not change the headline finding; it did change three of the numbers under it.

Reddit is in all twelve

Not most. All twelve, and first in nine of them.

The pages winning are threads, not articles. "Who's still using Notion and who found an alternative?" is the top result for a search with real commercial value, and it is a person asking strangers a question.

The difficulty score is wrong here, and we nearly believed it

Before pulling the results, we priced these terms through a keyword tool. It reported a difficulty of 0 or 1 for every single one of the thirty-two "alternatives" searches we measured — while every other kind of term we checked that day spread across the full range from 0 to 91.

A uniform score is not a measurement, it is a gap in the data wearing a number. The results confirm it: Gartner, IBM, PCMag, DataCamp, DigitalOcean, Zapier and New Relic all hold page-one positions on these searches. That is not a difficulty of zero by any definition.

We are writing this down because we came close to acting on it. A score of zero across dozens of commercially valuable terms looks like a discovery. It was a bug.

Directories rank here. AI-tool directories do not.

This is the finding the shallow first read got wrong, and the corrected version is sharper than what we originally published.

Of 336 results, 14 are directories — about four percent. But look at which ones:

Every one is either a long-established B2B software review platform or a narrow, opinionated catalogue. Not one general AI-tool directory appears in all 336 results — no Futurepedia, no Toolify, no There's An AI For That, and none of the rest of the category we belong to. We checked for them by name.

They also rank deep. Only 3 of the 96 top-eight slots are directories, against 8 of the 156 slots from position sixteen down.

There are two ways to read that, and only one of them is flattering. It could be an opening nobody has taken. It could equally be Google deciding that a general catalogue page is not what this search wants — a conclusion the presence of specialist directories, and the absence of general ones, mildly supports. We do not know which, and anyone who tells you they do is guessing.

What we checked and could not confirm

The first version of this article said that the pages which rank tend to claim the writer actually used the tools — "I've tried dozens", "Tried & Tested" — and concluded that a bare list is not what wins.

The larger sample kills it. Titles claiming first-hand testing are 12% of the top eight results, and 12% of results at positions sixteen to thirty. Identical. Whatever separates a page that ranks third from one that ranks twenty-eighth, saying you tested the tools is not it. We had found a real feature of this kind of page and mistaken it for an explanation.

If you make one of these tools, someone is writing your alternatives page

This is the part founders tend not to look at.

Forty-four of the 336 slots are held by tools we list in this directory — companies ranking on a rival's name. Uizard on "figma alternatives". Zapier on three separate searches. Two different tools on "quillbot alternatives", above QuillBot itself.

Three companies turned up on the search for their own alternatives, having written the page themselves: LangChain sits at number one for "langchain alternatives", Metabase at three and again at eleven with a second comparison, Sentry at four against a competitor.

That is the whole strategy in one line. The search for your alternatives is going to rank something. If you do not write it, a competitor will, and their version of the comparison will not be generous.

What we are not claiming

Twelve searches is a sample, chosen from tools in our catalogue that had measurable demand on this shape of query. It is enough to show that Reddit is dominant and that general AI directories are absent; it is not enough to put a percentage on the whole market, and we have not.

These are results for one country and one language, read on one day. Search results move.

The combined demand for those thirty-two searches is roughly 6,000 a month, which is real but modest. This is not a gold mine. It is a corner of search where the incentives are unusually visible: the people who make the tools are the people writing the comparisons, and the single most reliable winner is a forum thread where somebody unaffiliated just says what they thought.

If you would rather your tool were described by someone with no stake in the answer, add it and we will describe it.