OST Session on Max AI

Brand & Content Offsite, October 2025

CONTEXT

• New focus on AI has prompted major resource and team shifts
• There’s alignment that AI is the future, but fragmentation on how we get there
• Marketing is lagging relative to AI product development

KEY QUESTIONS

• Is Max AI a PostHog product, or a repositioning of PostHog itself?

CURRENT POSITION & GAPS

• Pricing for AI capabilities in PostHog is still being defined discussion here
• Max AI is strongest in analytics-oriented tasks, not yet able to execute (e.g. set feature flags) breakdown here
• Max AI output is currently not meeting user expectations — some concern for early adopters losing confidence / being unwilling to engage later on with a more advanced version
• Fuzzy naming conventions (agent, mode, capability, feature, add-on) is causing confusion internally
• The scrapped “four agents” concept helped clarify purpose and output — people still need to know what Max is for. (Max Modes could run with this theme)
• We need clear, tactical use cases to drive awareness and adoption of Max AI capabilities
• Deep Research is big bet — but team questions if it’s “too late / old hat.”
• Session summaries should be marketed as an AI add-on rather than a stand-alone product
• Max AI-related content is in its infancy (so far we've done: billboards, MVP website presence, some docs and a few case studies)
• Competitors’ growth (AI-first companies) is heavily driven by marketing reach and distribution
• PostHog's AI currently lacks differentiated storytelling or a sense of “what’s new and never done before.” The tone of positioning and messaging will be critical to nail

THE ICP PERSONA PARADOX

• Primary ICP: Product engineers and technical users who want faster insight loops.
• Emerging users: Product managers, PMMs, and growth teams who want to query in natural language — “PostHog easy mode.”
• Max AI makes it easier for non-technical users to act as 'product engineers' (is this a GUI moment for PostHog? New user personas does not necessarily distract or detract from the core ICP mission)

TAKEAWAYS

• Debate is unresolved: today, Max AI behaves like a feature, but aspirations are platform-level
• The most urgent need is to clarify: Are we evolving Max AI into the interface for all of PostHog, or one product line within it?
• The main blockers aren’t technology — they’re narrative clarity, use-case definition, and cohesive go-to-market orchestration
• PostHog’s historic last mover advantage can still be leveraged — letting others test AI narratives before we define a better one

ACTION ITEMS

What can content and marketing ship in the meantime?

• Audit current AI messaging across website and docs for consistency
• Update Max AI product page (website)
• Add Max AI capabilities to all other product pages (website)
• Create sample use cases for personas that are most likely to become Max AI power users (product managers, growth marketing and product marketing teams, etc..)
• Analyze top AI-first company marketing patterns and extract learnings for PostHog’s playbook
• Identify 2–3 standout campaign themes that could break new ground in AI storytelling

NEXT STEPS

What else do we need to consider?

• MCP
• A2A
• Installation wizard