TORONTO, Canada - July 1, 2026 - When buyers ask an AI assistant to recommend project-management software, the winners are not who traditional search rankings would predict, according to new original research from AI visibility platform LLMRanks.
The study recorded 342 answers from three leading AI systems — ChatGPT, Google's Gemini, and Google AI Overviews — across 30 unbranded buyer questions, each asked three times per engine to average out the run-to-run variation these systems are known for. It measured how often six leading tools — Asana, ClickUp, Monday.com, Trello, Notion and Jira — were named, and how often each was offered as a top recommendation.
The leaderboard
Across 270 open buyer answers, Asana was named most often, appearing in 81% of answers, followed by ClickUp (74%), Monday.com (70%), Trello (60%), Notion (44%) and Jira (43%). The order held loosely across engines, but the runner-up changed depending on which AI was asked: ClickUp pushed hardest on ChatGPT, Monday.com on Gemini, and Trello on Google AI Overviews.
"You can't optimize once and win everywhere," said Abubakr Latif, founder of LLMRanks. "Your visibility depends on the specific engine, so it has to be measured on each one separately."
Strong SEO didn't translate into AI visibility
The most striking result concerns the link between search authority and AI recommendations. Using authority data from Ahrefs (US, June 2026), the study found all six tools sit at the very top of the domain-authority scale — Domain Ratings of 90 to 92, effectively a tie. Yet their AI visibility ranged from dominant to nearly invisible.
Notion holds the group's highest domain authority and largest backlink footprint, yet placed last in AI naming (44%) and was cited by ChatGPT just twice. ClickUp, by contrast, ranks for the smallest keyword footprint of the six and has the lowest domain rating — yet it was the most-cited single product and the second-most-named tool overall. Jira's parent domain ranks for the most keywords and traffic of any tool studied, yet Jira was named the least.
"Great SEO is the foundation — it gets a brand eligible," added Abubakr Latif. "But a separate layer on top decides who actually gets chosen inside an AI answer. Domain authority, traffic and keyword footprint didn't reliably predict who the AI named."
The review sites buyers trust are nearly absent
The research also classified all 781 source domains the engines cited. Big-name software-review sites — G2, Capterra and TrustRadius — together accounted for roughly 1% of citations. Instead, AI leaned on a long tail of niche blogs and listicles (59% of citations), vendor pages (16%), Reddit and forums (8%), industry and education sites (6%), and YouTube (6%). Reddit, YouTube and a niche Agile-training site were the most-cited individual domains.
A further finding: AI systems name brands far more often than they link a source. Every engine named a brand in essentially 100% of answers, but Gemini linked to a source only about half the time, versus 83% for ChatGPT and 99% for Google AI Overviews — suggesting that "being recommended" and "earning a citation" are two distinct goals.
Methodology and data
The findings are a point-in-time snapshot — AI answers drift over time — and describe correlation rather than proven cause. The full study, methodology and complete dataset (every question, engine and answer) are published for open inspection at https://llmranks.io/blog/seo-vs-aeo.
About LLMRanks
LLMRanks is an AI visibility platform that helps brands track and improve how ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews cite them. Companies can run a free AI visibility check at https://llmranks.io/ai-visibility-check.
Media Contact
Company Name: LLMRanks
Contact Person: Abubakr Latif
Email: Send Email
Phone: 4165095797
Address:338 Baverstock Crescent
City: Milton
State: ON
Country: Canada
Website: https://llmranks.io/
