AI chat aides are not being utilized the means many marketing professionals expect. As opposed to acting as buying tools or online search engine, they operate mostly as support systems for cognitive jobs– writing, intending, evaluating and learning, according to a report by AI search engine optimization firm Dejan. In 2026, maximizing for AI implies comprehending how customers in fact engage with assistants, not just thinking industrial actions.

Dan Petrovic, supervisor of Dejan, carried out an extensive evaluation of 3 9 million conversational turns– covering 613 million words and 4 4 billion characters– to uncover patterns in real-world AI usage. The results test numerous presumptions in AI material approach.

Usage patterns: most chats are brief and task-oriented

• Typical conversation size: 2 turns (a solitary concern and answer)
• Typical session size: 430 words
• Over 80 % of chats are under 1, 000 words
• Only 4 2 % of sessions surpass 2, 500 words, commonly including modifying, summarization, tutoring, coding, or data analysis
• Mean session size: 732 words, altered by lengthy paper entries
• Assistant output volume: about 1 5 times that of individual input

Petrovic kept in mind that individual contributions usually make up simply 16– 17 % of the session, meaning the assistant creates the bulk of the web content.

Dig deeper: What takes place when nobody clicks anymore

Intent breakdown: commercial usage is restricted

Petrovic analyzed 24, 259 classified sessions across 42 intent categories. He located that 64 6 % of sessions had no industrial intent. Individuals largely engaged AI for jobs like:

• Brainstorming (7 7 %)
• Planning (6 5 %)
• Psychological assistance or discussion (6 2 %)
• Evaluation (5 7 %)
• Knowing (4 7 %)
• Text improvement, including summaries and translations (4 6 %)
• Web content development (3 9 %)

Even among the 35 4 % of sessions showing commercial intent, the bulk were early in the funnel:

• Recognition stage: 10 %
• Factor to consider: 8 5 %
• Exploration and decision support: 6 9 % consolidated
• Transactional support: 4 8 %
• Post-purchase assistance: 5 1 %, including troubleshooting and usage assistance

Key understanding: AI conversations are cognitive operations, not questions

Lots of marketing professionals and Search engine optimizations are enhancing web content with a search-first attitude– thinking AI chats imitate keyword-based questions. Petrovic’s findings recommend otherwise. AI aides are more frequently made use of to sustain multi-step jobs, instead of making instant purchases.

“AI chat usage is extremely non-commercial,” Petrovic stated. “Individuals treat assistants as co-pilots– not sales reps.”

Effects for AI material strategy

Marketing experts need to change their technique to AI optimization by:

• Focusing on content that sustains awareness and early-funnel exploration
• Producing structured, high-context information that agents in longer workflows can recycle
• Preventing overemphasis on transactional search phrases in AI-facing content

Bottom line: The future of AI material exposure isn’t around gaming intent– it has to do with meeting it. Assistants are becoming devices for cognition, not simply conversion. If your web content aids individuals find out, plan, create, or review, it’s even more likely to turn up in real-world AI usage.

The report. How do individuals utilize AI aides?