Exa raised $225M to build search for AI agents. Two days later Google announced 1 billion AI Mode users with queries doubling every quarter. Does Google's distribution moat transfer when the customer is a programmatic API call, not a human with a browser default?
Walmart failed agentic checkout. Alibaba processed 200M AI-native orders. Amazon retired a 300M-user product because the brand was the constraint. Google built a protocol for it. Four bets on agentic commerce. One of them is right. https://t.co/atXLKYdAqI
@PatrickCFOD not just "safer," but simply . . . "easier". Who has the time and mental space to be managing up 24/7? Folks rarely get fired for doing exactly as the boss says.
The dangerous version of a fast-moving CEO is not the one who makes bad decisions. It is the one who moves fast enough that no one disagrees anymore. Speed that is right 80% of the time trains teams to assume it is right 100%.
When you have one fractional client, a difficult client is a cash flow problem. When you have three, it is a decision. The pipeline is not just revenue. It is a positioning instrument.
Alibaba is moving Taobao to conversational AI. Browse, compare, buy through chat. The trust infrastructure problem nobody is talking about: at TripAdvisor we assumed trust transferred to a new context. It does not. The interface changes the contract.
Steve Kaufer ran TripAdvisor's product review himself. My first one: he asked which OTAs were the largest buyers on meta. I didn't know. "Aren't you supposed to know that?" I spent the rest of the meeting writing a HiveQL query from the back row. https://t.co/7AofDw4EFJ
At a private CRO luncheon last week with top practitioners in Boston. Consensus: organic and paid traffic is already dropping to AI search. Three things to do - qualify harder on existing traffic, test by channel before page, and keep human judgment in the testing loop.
@shreyas At TripAdvisor, the most valuable person in any product review wasn't the one with the cleanest story. It was whoever brought the thing that wasn't working into the room early enough to do something about it. Harder to manage. Never wrong.
A CPO friend at a PE-backed company told me the standard his firm set: come back when everyone is a solid A. The B player is the dangerous one - good enough that the case for action is never obvious, not good enough that the team performs at the level you need.
At EverQuote we didn't pick a second vertical when we expanded. We launched several at once. Every new vertical needed carrier integrations and volume before the supply side would compete. You can't run that in parallel. Sequencing is the strategy.
Etsy's Q1 turnaround metric is not revenue. It is GMS per buyer, up year-over-year for the first time since 2022. That means they stopped trying to fix the top of funnel and bet on the existing user instead.
When the steam engine got more efficient, demand for coal tripled - not collapsed. More efficient engines meant more machines, more fuel needed. AI is doing the same thing to product judgment. The ratio of judgment required to execution required is increasing.
@clairevo Six months ago these conversations had a center of gravity. Now they don't. That's either a sign the category is maturing into real infrastructure or fragmenting before it consolidates. The DB work and hardware hacking suggest the former.
National sign-off does not produce local adoption. A free curriculum backed by a famous founder with 300+ school partners is still fighting a last-mile distribution problem. Brand helps. Free helps. Neither solves it. You still have to close Ohio.
@sama The builders who find it useful are the leading indicator. The lagging indicator is whether the orgs those builders work inside actually change how they make decisions. That gap is where most AI ROI gets lost.
@venturetwins The model picks up signals you didn't know you were broadcasting. The unsettling part isn't that it guessed right - it's that there's probably a coherent reason why.