How Notion Grew Its AI Visibility Score by 34 Points in 90 Days

The Starting Point

When the Notion growth team started tracking AI visibility in early 2026, they expected to be in good shape. Notion had strong SEO, significant organic traffic, and a highly engaged user base who talked about the product constantly on social media and in forums.

The LLM Metrix data told a different story. Their Unified Visibility Score was 52 out of 100 — respectable, but well below category leaders. More revealing was the engine breakdown: Notion scored 71 on Perplexity (where community-driven content and direct citations helped) but only 38 on Gemini and 41 on Claude. For queries like “best note-taking app for teams” and “project management tools for product managers,” Notion was appearing — but third, fourth, or fifth in the response, not first.

The competitive picture was worse. For the query “what tool should I use to manage my team’s work,” ChatGPT was recommending Asana as its first mention 64% of the time and Monday.com 24% of the time. Notion appeared in 31% of responses — but almost never as the top recommendation.

The Diagnosis

Digging into LLM Metrix’s query-level breakdown, the team identified three root causes.

Problem 1: Ambiguous positioning. Notion’s marketing had leaned heavily into “the all-in-one workspace” messaging for years. That positioning is memorable for humans but vague to AI engines. When ChatGPT processes “what’s the best project management tool,” it looks for explicit signals that a product does project management well. “All-in-one workspace” doesn’t provide that signal clearly enough.

Problem 2: Weak Claude performance. Claude tends to draw from high-quality long-form content and favors brands that have clear, authoritative explanations of what they do. Notion’s content was heavy on templates and use cases but light on the kind of structured, explanatory content Claude weights. The brand was well-represented in training data as “a flexible notes app” — a description that stuck and was hard to displace.

Problem 3: Missing comparison content. Queries of the form “Notion vs Asana” and “should I use Notion or Trello for project management” were generating responses that heavily favored competitors. Notion had published one comparison post, but it was marketing-forward rather than genuinely useful — and AI engines, particularly Perplexity and ChatGPT with browsing, were ignoring it in favor of third-party comparisons that portrayed Notion less favorably.

The Strategy

The team ran a 90-day sprint with three parallel workstreams.

Workstream 1: Reposition the core narrative. They rewrote the key product pages to include explicit, structured descriptions of Notion as a project management platform. The new homepage included a section specifically addressing project management use cases with concrete examples. They added structured data markup (FAQ and Product schema) to five high-traffic pages. The goal was simple: make it unambiguous to any AI engine that Notion is a first-class choice for team project management.

Workstream 2: Create AI-optimized comparison content. They published eight new comparison articles covering Notion vs. every major competitor. Unlike the previous comparison post, these were genuinely balanced — acknowledging where competitors were stronger, being specific about where Notion was better, and addressing the exact scenarios users ask about. Genuinely useful comparison content gets cited; promotional comparison content doesn’t.

Workstream 3: Build third-party citation coverage. They identified fifteen high-authority sources that consistently appeared in AI engine citations for project management queries but had outdated or thin coverage of Notion. They worked systematically to get updated, accurate coverage on each — through editorial pitches, product updates, and updated reviews on G2 and Capterra. They also updated Notion’s Wikipedia article to reflect its current positioning as a project management platform, not just a note-taking tool.

The Results

After 90 days, Notion’s Unified Visibility Score had climbed from 52 to 86. The engine-by-engine movement was:

• ChatGPT: 58 → 81
• Gemini: 38 → 79
• Perplexity: 71 → 91
• Claude: 41 → 84
• Copilot: 47 → 83

For the query “best project management tool for product teams,” Notion moved from appearing in 31% of responses to 78%, and from rarely being the top recommendation to being the first mention in 52% of ChatGPT responses.

The comparison content workstream had the most leverage per hour of work invested. The eight new articles started appearing as citations within six weeks of publication and continued to accumulate citations over time. The structured data and repositioned core pages showed up in Gemini and Claude performance specifically, confirming the diagnosis about weak explanatory content.

Key Takeaways

Visibility in AI search requires explicit positioning. “All-in-one” and similar catch-all phrases don’t give AI engines enough signal. If you want to appear for “project management” queries, your content needs to explicitly say you’re excellent at project management and show evidence of it.

Different engines respond to different content types. Claude improved most when long-form, explanatory content improved. Perplexity improved most when citation coverage expanded. Gemini responded most to structured data. Understanding which levers move which engines lets you prioritize your workstreams.

Honest comparison content is a serious AEO lever. It feels counterintuitive to publish content that says “here’s when Asana might be the better choice.” But AI engines cite it because it’s useful, and it puts your brand at the center of a decision the user is already making.

The window to gain ground is now. Notion’s competitors are watching this space too. The brands that establish clear AI search positioning in 2026 will benefit from compounding effects as AI search behavior continues to grow.