Two products, ClickUp and Wrike, were named in every answer. Nothing else was.
We put six buyer questions about project management software to ChatGPT, Gemini and Perplexity on 3 August 2026, gave each question one live web search, and recorded every product each assistant named and every page it built the answer from. All six questions returned a readable answer. Only two names appeared in all six: ClickUp and Wrike. Asana and Trello followed, each named in five of the six. That is the whole shortlist a buyer is likely to be handed today, and it is shorter than the category's own marketing would suggest.
Below ClickUp and Wrike at six of six, and Asana and Trello at five of six, sit Jira, Monday.com and Notion, each named in four of the six answers. Paymo was named in three. Everything after that, Smartsheet at two and then a long tail of single mentions (Freedcamp, GanttPro, Microsoft Project, Miro, nTask, Plaky, Teamwork, Zoho Projects), appeared in one answer and vanished from the other five. A single mention across six varied questions is noise: a buyer cannot rely on it showing up again, and a vendor cannot bank on it.
Each of the six questions produced a different shortlist
The most useful finding is not the ranking, it is how much the shortlist moved between questions. The assistants do not hold one fixed list of the best project management software and read it back. They answer the specific question asked, and the products change with it.
Ask "best free project management software" and Freedcamp appears, a name absent from every other question. Ask "most affordable project management software" and the answer swings to a budget-specific set, Zoho Projects, GanttPro, nTask and Plaky, three of which show up nowhere else in the study. Ask the bare "best project management software" with no buyer intent attached and the answer degrades: the assistants reached for dictionary.cambridge.org, merriam-webster.com and even bestbuy.com, pulling a definition of the phrase rather than a shortlist of tools. The buyer-shaped questions, "for small teams", "what should I use", "recommendations", are the ones that returned a clean list.
For a vendor this is the practical lesson, and it is the same one Google's AI surfaces show: you do not win the answer, you win a specific question. Paymo is invisible in "best" and "for small teams" but named in "most affordable", "what should I use" and "recommendations". It earned three of its slots by being the right answer to three narrower questions, not by being the biggest name. Covering more distinct buyer questions is how a smaller tool shows up more often, because each answer holds about one slot per product and the question, not the brand, decides who fills it.
The answers came from community threads, a few roundups, and the vendors' own pages
When we recorded the pages each answer was built from, three domains fed five of the six answers: reddit.com, project-management.com and paymoapp.com. icagile.com fed four, quora.com and wrike.com fed three each, and youtube.com fed three. So the recommendation set for this whole category is assembled from a short list of community threads, two or three specialist roundup sites, and, notably, the vendors' own domains.
That last part is worth stopping on. paymoapp.com, Paymo's own website, was a source in five of the six answers, and Paymo the product was named in three. wrike.com fed three answers, and Wrike was named in all six. The assistants are citing vendors' own content as evidence for recommending those vendors. That is not a flaw to report; it is a description of what works. A vendor with genuinely useful, comparison-shaped content on its own domain can feed the answer that names it, and two vendors here are doing exactly that.
The review directories a project management vendor might expect to matter barely registered. pcmag.com fed two answers and technologyadvice.com fed one. There is no G2 or Capterra anywhere in the source list for these six questions. The pages that decided who got recommended were Reddit threads, a few category roundups, and the vendors' own writing.
What a vendor in this category would do about it
If you sell project management software and were not named, the reason is in the source list, not in your product. Every product the assistants named appears on the pages they retrieved: the Reddit and Quora threads, project-management.com, icagile.com, and in two cases the vendor's own site. A tool absent from those pages has no path into the answer, whatever its market share. The work is getting into the specific threads and roundups the assistants pull for your buyers' questions, and publishing the comparison content on your own domain that the assistants proved willing to cite.
The harder part is that neither the source set nor the shortlist is stable, so knowing where you stand is not a one-time check. The list that names ClickUp and Wrike today is one Reddit thread and one roundup edit away from naming someone else next month. That is the case for measuring it on a schedule rather than guessing once.
Where this leaves us
Glotier does not sell project management software, so it is correctly absent from every answer above: we will not print our own name into a category we do not compete in. What we do is the measurement itself. This study was run with glotier.com's own tools, the same ones we point at a customer's own category: we ask the buyer questions their prospects type and record which of the three assistants named them, who got named instead, and the exact pages behind each answer. If you want that for the category you actually sell in, the check is free and takes about a minute.
The full list, counted
Every product the three assistants named across the six answers, and how many of the six named it. The share is out of the answers we read, not out of the category.
| Product | Named in | Share |
|---|---|---|
| ClickUp | 6 of 6 | 100% |
| Wrike | 6 of 6 | 100% |
| Asana | 5 of 6 | 83% |
| Trello | 5 of 6 | 83% |
| Jira | 4 of 6 | 67% |
| Monday.com | 4 of 6 | 67% |
| Notion | 4 of 6 | 67% |
| Paymo | 3 of 6 | 50% |
| Smartsheet | 2 of 6 | 33% |