AI adoption in companies tends to follow a predictable vertical pattern.
The top floor adopts quickly. Executives have assistants, strategy documents, high-level pattern questions — these map naturally onto what general-purpose AI does well.
The bottom floor adopts well too. Individual contributors doing research, writing, coding, data analysis find AI genuinely useful almost immediately.
The second floor stalls. Team leads, department heads, managers who sit between strategy and execution — these are the people AI adoption most often fails to reach. And they're the ones whose adoption actually determines whether the organization changes.
Why the second floor is different
The second floor's work is inherently contextual. A department head doesn't just need to summarize text or generate a draft — they need to answer questions like: Is this rock on track given what I know about the team's capacity? Does this decision contradict what we aligned on in Q1? What should I flag in Thursday's L10 that might otherwise get buried?
Generic AI answers these questions generically. And generic answers, for people who know the specifics, feel like more work than just thinking it through themselves.
So they stop using the tool. And the adoption curve flatlines at the second floor.
What makes AI useful for the second floor
The answer is context. Specifically, company context that's deep enough to make the AI's answers feel accurate rather than plausible.
When an AI knows your company's rocks, your accountability structure, your decision history — the answers it gives to second-floor questions become useful instead of approximate. The department head doesn't get a framework for evaluating quarterly priorities; they get an assessment of whether this particular rock is on track given the last three weeks of updates.
That specificity is the difference between an AI that gets used and one that gets ignored.
The compounding effect
Second-floor adoption matters most because second-floor users are the ones who shape team behavior. When a department head consistently uses AI to prepare for L10s, the meeting runs better. When they use it to brief new hires, onboarding improves. When they use it to track rock trajectory, issues surface earlier.
Fixing the second floor doesn't just add users. It changes how the team operates.
That's the problem Freddy is designed to solve — not by making AI easier to use, but by making AI contextual enough to be worth using for the people who most need it.
BrainGem builds Freddy, an AI that lives in Slack and learns your company's operating context — built for the second floor, not just the top.
Top comments (0)