OneSignal was named in 5 of the 6 answers we collected for in-app messaging tools, more than any other product. On 2026-08-03 we put six buyer questions to ChatGPT and Gemini, one live web search per question, and wrote down every product each assistant named and every source domain it drew from. OneSignal, at 5 of 6 (83%), was the only tool to clear five answers.
Refiner came second, named in 4 of the 6 answers (67%). Braze and Pushwoosh tied for third, each named in 3 of 6 (50%). After that the drop is steep: Appcues, Pendo and Userpilot were each named in 2 of 6 (33%), and everyone else appeared once.
In total, 19 different products were named at least once across the six answers, and 12 of those 19 appeared in only a single answer. So the agreement between ChatGPT and Gemini is thinner than a top-line list suggests: only 4 products out of 19 were named in half or more of the questions.
When the question changed, the whole list changed
The bare "best in-app messaging tools" question returned 8 products: Userpilot, Refiner, OneSignal, Userlist, Customer.io, Appcues, Pendo and Braze. Two of those, Userlist and Customer.io, showed up in this one question and nowhere else in the six.
Adding three words flipped the category. "Best in-app messaging tools for small teams" returned a completely different 7 products: Slack, Google Chat, Microsoft Teams, Pumble, Discord, Zoho Cliq and Element. Not one of those 7 appeared in any of the other five questions. The phrase "for small teams" moved the answer from product-adoption messaging to workplace chat, and it was the only question where OneSignal did not appear at all.
"What in-app messaging tools should I use" returned 8 products and was the only place Chameleon was named, alongside the familiar Userpilot, Refiner, Pushwoosh, OneSignal, Pendo, Braze and Appcues.
The two price-shaped questions split as well. "Best free in-app messaging tools" narrowed to just 3 products: OneSignal, Pushwoosh and Refiner, all three of which also appeared in other questions, so "free" filtered the list down rather than surfacing anything new. "Most affordable in-app messaging tools" returned 4 products: OneSignal, Pushwoosh, Quickhunt and UserGuiding, and both Quickhunt and UserGuiding appeared only in this affordability question.
So "free" and "affordable" overlapped on just 2 products, OneSignal and Pushwoosh. Refiner made the free list but not the affordable one; Quickhunt and UserGuiding made the affordable list but not the free one. Same buyer intent, two different shortlists, decided by which word you use.
Across five of the six questions one small group kept reappearing: OneSignal, Refiner, Braze and Pushwoosh. The exception was "for small teams," which shared none of them. If a buyer types the plain category name, those four are roughly what the assistants converge on; add "for small teams" and all four vanish.
Where the answers came from
Twenty different domains fed the six answers. The three most-cited each fed 5 of the 6: pushwoosh.com, refiner.io and userpilot.com. All three are vendor sites in this category, not neutral reviewers.
This is the pattern worth staring at, and it is the key insight of the whole measurement: the vendors whose own domains fed the search are the vendors who got named. OneSignal's own site, onesignal.com, was cited in 3 of the 6 answers, and OneSignal was the most-named product at 5 of 6. Refiner's domain fed 5 answers and Refiner was named in 4. Pushwoosh's domain fed 5 and Pushwoosh was named in 3. Your own ranking pages are doing the citing.
Community threads carried real weight too. reddit.com was cited in 4 of the 6 answers, the most-cited domain that is not a vendor, and quora.com in 1. The "small teams" answer in particular leaned almost entirely on community and roundup pages: reddit, facebook, youtube, zapier, jimdo, pcmag and slack.com.
Independent roundups fed the rest. pcmag.com appeared in 3 answers, businessofapps.com and zapier.com in 2 each, and a cluster of product-marketing blogs (productfruits.com, produktly.com, knock.app) in 2 answers each.
One absence stands out: no review directory appeared at all. G2, Capterra, TrustRadius and GetApp were cited in 0 of the 6 answers. In this category the AI answer is built from vendor blogs, Reddit and tech-press roundups, not from the directories vendors usually spend the most time on.
What it takes to get named here
The source list, not market share, decided who got named. Every product that cleared 3 of 6 answers (OneSignal, Refiner, Braze and Pushwoosh) had either its own domain or a Reddit thread among the pages feeding those answers.
If you sell an in-app messaging tool and want to be in these answers, be present on the exact pages that fed them: your own domain ranking for these queries (refiner.io, userpilot.com and pushwoosh.com each did that in 5 of 6 answers), a Reddit presence (cited in 4 of 6), and inclusion in the roundups that ranked (pcmag.com in 3, businessofapps.com and zapier.com in 2 each). Being the biggest name is not what got tools named. Being on those pages is.
And pick the query you are fighting for. Ranking for "best free" put a tool in a 3-product answer; ranking for "for small teams" put it in a different 7-product answer next to Slack and Microsoft Teams. Same category label, different competition, different sources.
The full list, counted
| Product | Named in | Share |
|---|---|---|
| OneSignal | 5 of 6 | 83% |
| Refiner | 4 of 6 | 67% |
| Braze | 3 of 6 | 50% |
| Pushwoosh | 3 of 6 | 50% |
| Appcues | 2 of 6 | 33% |
| Pendo | 2 of 6 | 33% |
| Userpilot | 2 of 6 | 33% |
| Chameleon | 1 of 6 | 17% |
| Customer.io | 1 of 6 | 17% |
Glotier is not in this list, and that is correct
Glotier does not sell an in-app messaging tool, so we were named in 0 of these 6 answers. That is the honest and expected result: we do not compete in this category, so ChatGPT and Gemini should not name us here, and they did not.
But this measurement is exactly what we run for a customer's own category. Six real buyer questions, the actual assistants, every product and every source domain recorded, so you can see whether you are named in your category and which pages are deciding it. If you sell something like an in-app messaging tool, you can run the same check on your own category: it is free, no account, about a minute.