Recall says chat tripled knowledge retrieval in 2026 report

Recall published a 2026 report on how more than 700,000 users build and use AI “second brains,” finding that retrieval rose threefold as chat became the main way people returned to saved material. The company says natural-language search is turning passive archives into active knowledge tools.

Why it matters: - Recall’s latest product data suggests AI chat is changing saved-knowledge habits from hoarding to actual reuse. - The shift matters because second brains only work if people return to the material they saved. - The report frames natural-language retrieval as a practical improvement over folders, tags and keyword search.

What happened: - Recall published The Second Brain Report 2026: How People Build and Use an AI Second Brain. - The report draws on product data from more than 700,000 registered users. - Retrieval increased from 10 actions per 100 saved items in August 2025 to 30 in August 2026. - The share of monthly savers who also retrieved saved content rose from about 27% in April 2025 to 52% in August 2026. - Recall said chat became the most common route back to saved knowledge after it introduced chat in August 2025 and expanded it with Recall 2.0 in April 2026.

The details: - Among paying users who returned to their libraries, 81% used chat. - In the same group, 54% reopened a source and 15% used conventional search. - During the report’s 90-day comparison period, YouTube videos accounted for 49% of saved items. - Web pages made up 36% of saved items. - Posts from six social platforms represented about 6% of saved items. - PDFs followed in the mix of saved content. - User-written notes were only 2.5% of saved items, or about one in 40. - Recall said most saved material was something people learned from, not material they created themselves. - The report says saving information is not the same as learning it. - Recall uses quizzes built from saved material with active recall and spaced repetition. - Among paying users who completed a first quiz, 64% returned for a second. - Among the smaller self-selected group that reached a tenth quiz, 93% returned again. - The report analyzed aggregate product data from April 1, 2025, through September 18, 2026. - The dataset included 966 voluntary survey responses collected between 2023 and 2026. - Saving data covered paying and non-paying users. - Retrieval and feature-use findings covered paying users because free-account limits could affect those measures. - Recall said it handled the data in line with its Privacy Policy.

Between the lines: - The report argues that an AI second brain works best when it feels conversational, not archival. - That framing positions Recall against the older idea that saving information is the end goal. - The data also suggests users are building personal libraries of things they consumed, especially video and web content, rather than only storing original notes. - Tiago Forte’s team tested Recall for three weeks on real projects and preferred its response in one comparison with Anthropic’s Claude because Recall could use the exact content the team member had saved. - Forte wrote that a user’s own knowledge and context are the fuel for what AI can do on their behalf. - Recall cites Forte’s test as evidence that curated personal context can improve AI usefulness. - Forte’s full write-up is available as the full report.

What's next: - Recall says people who regularly encounter information they expect to revisit are the best fit for a second brain. - The company is pushing three habits: use a second brain you can talk to, save material with minimal friction and test what you want to retain. - Recall’s YouTube video summarizer is one example of that strategy because it turns a video into a timestamped summary saved directly to the user’s library. - Recall is continuing to position its platform as a searchable AI second brain for videos, podcasts, web pages, PDFs, ebooks, images, documents, social posts and user notes.

The bottom line: - Recall’s 2026 report says the future of personal knowledge management is less about storing more and more about asking better questions of what you already saved.