AI Is a Feature, Not a Strategy

Your board wants an AI strategy. Product teams respond to the challenge, and add a chatbot into their product. IT evaluates LLM options and provides employees with access to Claude or GPT.

That's an AI strategy, right? Right?

To troubleshoot your strategy, delete the word "AI" from your strategy decks. Did anything change? What changed for the customer? How do we win? If the deck collapses, you had an AI feature-focused slide, not a strategy.

Strategy should answer three questions:

Which customers are you serving?
Where are they struggling to get a job done? What's the Job-to-be-Done?
Why will they hire you over every alternative, including doing nothing?

While we can leverage AI to help us analyze data that goes into a strategy, or we can use it to bring customer needs into focus, AI shows up after those answers, not before. Sometimes it helps to remove real friction that customers will pay (or stay) for. Sometimes it's theater that wins the board meeting, but loses the renewal (rehire).

There's a separate question: how teams are structured and how they operate. Jeff Bezos' rule of thumb for "Two Pizza Teams" is now too big. Teams are smaller, one pizza teams of 1-3 people and 6-12 agents. These teams ship faster: daily, or every other day. They ship end-to-end bets, not button-color changes. Discovery happens in hours, not weeks or months. That advantage is real, and it shows up in your margins, not just your marketing.

So before the next AI roadmap review, ask some pivotal questions: what need or Job does this get done that customers couldn't get done before — or couldn't get done this cheaply, this quickly? How are we measuring success of this? If we place this bet, what return do we expect?

← Back to all insights