Cake Equity was the only cap table tool ChatGPT and Gemini named in all six of our buyer questions, 6 of 6 answers. On 2026-08-04 we asked the two assistants six ways a founder might ask for cap table software, gave each one live web search, and wrote down every product they named and every source domain behind the answers. Cake Equity came back every single time.
Carta finished second, named in 5 of 6 answers (83%). The one question it missed is the most revealing result in the whole set, and we get to it below. Behind Carta, three tools tied at 4 of 6 answers (67%): Astrella, Ledgy and Pulley. So from two assistants running one search each, the top tier is narrow: one product at 100%, one at 83%, and three at 67%.
Below the top five the list thins fast. Clara landed in 3 of 6 answers (50%). Two products tied at 2 of 6 (33%): Eqvista, and Aegis (written by the assistants as "Aegis by Story.law"). Two more surfaced once each, 1 of 6 (17%): Vestd, and Aegis again under a second spelling, "Story.law's Aegis." Ten distinct product labels appeared in total, but only five of them cleared half the answers.
The full list, counted
| Product | Named in | Share |
|---|---|---|
| Cake Equity | 6 of 6 | 100% |
| Carta | 5 of 6 | 83% |
| Astrella | 4 of 6 | 67% |
| Ledgy | 4 of 6 | 67% |
| Pulley | 4 of 6 | 67% |
| Clara | 3 of 6 | 50% |
| Aegis by Story.law | 2 of 6 | 33% |
| Eqvista | 2 of 6 | 33% |
| Story.law's Aegis | 1 of 6 | 17% |
| Vestd | 1 of 6 | 17% |
Two rows above are the same product. The Story.law tool called Aegis got written two different ways: "Aegis by Story.law" in 2 answers and "Story.law's Aegis" in 1. We are leaving both rows exactly as the measurement recorded them, at 2 and 1, rather than merging them into a "3 of 6" that the data never actually printed. If you run cap table software, this is worth noticing on its own: an assistant that spells your name two ways is splitting your own visibility in half.
How the shortlist changed from question to question
The six questions did not return the same list, and the differences are the point. The bare "best cap table software" question pulled six names: Carta, Cake Equity, Pulley, Ledgy, Astrella and Clara. That is the widest single answer we got, 6 products from one general question.
Reframe the identical category as "best free cap table software" and the set moves. That question returned Carta, Pulley, Clara, Cake Equity, Ledgy and Eqvista, again 6 products, but Eqvista walks in here, a name that appears nowhere in the bare "best" answer.
Push harder on price with "most affordable cap table software" and Carta disappears completely. That answer named Cake Equity, Pulley, Ledgy, Eqvista and Aegis (as "Story.law's Aegis"), 5 products, and the biggest incumbent in the entire measurement, named in 5 of the other answers, is simply not one of them. Astrella and Clara drop out of the affordable answer as well.
So the price framing rewrites the shortlist. Eqvista shows up only in the two price questions, "best free" and "most affordable," 2 of 6 answers, and in no other phrasing. The cheaper Aegis label, "Story.law's Aegis," shows up only in the "most affordable" answer, 1 of 6. A vendor with clear pricing content can win a question the market leader loses outright.
Two products rode in on a single question each. Vestd appears only in "cap table software recommendations," 1 of 6, and in no other wording. "Story.law's Aegis" appears only in "most affordable," 1 of 6. Miss the exact question a buyer types and you can miss the entire answer, even when you are in the category.
The spread across phrasings was large. "Best cap table software for small teams" produced the shortest answer of all, just Carta, Cake Equity and Astrella, 3 products. The open-ended "cap table software recommendations" produced the longest, 7 names, adding Vestd and Aegis to the core five. Same category, and the count of who gets named ran from 3 to 7 purely on wording.
The source map: who fed the answers
Three domains fed all six answers: cakeequity.com, carta.com and reddit.com, 6 of 6 each. A community forum sits level with the two biggest vendors' own websites as the most-cited source in the whole category.
Reddit was the only community source doing real work, cited in all 6 answers. Facebook turned up once, 1 of 6. No Quora thread appeared at all. The "what do real users say" signal here is almost entirely a Reddit signal, present in 6 of 6 answers.
On review directories, G2 was cited in 5 of 6 answers (83%). Capterra did not appear a single time, 0 of 6. If you are a cap table vendor pinning your AI visibility on directory presence, note that exactly one directory, G2, actually fed these answers, and even it missed one question.
Roundup and comparison pages filled out the rest. LTSE (ltse.com) was cited in 4 of 6 answers, a non-vendor roundup pulling real weight. Qapita, Fidelity Private Shares and Cheqly each contributed pages too, in 1, 1 and 2 answers, even though none of those three was itself named as a recommended product. Their content got read; their names did not get picked.
Here is the insight that matters most for anyone selling in this category. The products named most often are the ones whose own websites the assistants were reading. Cake Equity's own domain, cakeequity.com, was cited in all 6 answers, and Cake Equity was named in all 6. Carta's carta.com fed 6 answers and Carta was named in 5. Astrella's astrella.com fed 4 answers and Astrella was named in 4, an exact match. Pulley, Story.law and Eqvista repeat the pattern down the list. In cap table software, getting your own pages into the assistant's reading tracks almost one-to-one with getting recommended.
What a vendor here would do to get named
If you sell cap table software and want to be in these answers, the source list is the map, not your market share. Cake Equity is not the largest name in this market, yet it was named in 6 of 6 answers, because its own pages, the Reddit threads, and the G2 listing the assistants read all carried it. Presence in the sources, not size, decided the outcome.
Concretely, this data points at four places worth being. Your own domain (cited for the named vendors in up to 6 of 6 answers), Reddit (6 of 6), G2 (5 of 6) and the roundups that rank for the category, like LTSE (4 of 6). Being on Capterra did nothing measurable here, because Capterra was cited 0 times.
And answer the price questions on purpose. Carta is named in 5 of 6 answers and still loses "most affordable" outright, 0 mentions in that one. A smaller tool with clear, cited pricing content, the way Eqvista appears only in the two price questions, can take a question the incumbent cannot. This shortlist is decided question by question, not once, so a vendor covering "free" and "affordable" pages explicitly is playing a different game from one that only ranks for the generic "best."
Where Glotier fits
We did not find Glotier in any of these six answers, and that is correct. We do not sell cap table software, so ChatGPT and Gemini should not be naming us in this category, and they did not. We are not going to pretend otherwise.
This page is exactly the thing we do run for a customer's own category. We took six real buyer questions, put them to ChatGPT and Gemini with live search, and counted who got named and which pages earned it. If you sell in a category, the same check tells you whether the assistants name you, who they name instead, and the precise source pages you would need to be on to change that. It is free, no account, and takes about a minute.