Do You Need an AI Strategy? You Need One Win First.

Every few weeks a business owner asks me to build them an “AI strategy.” My honest answer usually disappoints them: you probably don’t need one yet. A strategy is a plan to spend money. What you actually need first is proof the money was worth spending — one working use case that moves a number you care about.

I spent three decades building companies before I did this. In shrimp farming, nobody wrote a five-year genetics strategy before they had a single tank of animals surviving. You prove one batch — one protocol, one pond — and you scale what worked. In my case the right approach took survival from a stuck 15–20% to over 90%. That number came from doing the work at small scale, not from a deck that predicted it.

What an “AI strategy” usually buys you

In practice, most AI-strategy engagements produce a 40-page document: a maturity assessment, a landscape of tools, a roadmap with quarters on it, and a lot of the word “transformation.” It feels like progress. It costs real money. And six months later the business runs exactly as it did before — because a roadmap isn’t a result, it’s a plan to maybe get one.

Vendors love it. A strategy engagement is recurring, low-risk revenue and it commits you to the next phase. I’m vendor-neutral by principle, so I’ll say the quiet part: the strategy deck is usually where momentum goes to die.

Start with one win, not one strategy

The operator move is the opposite. Instead of mapping everything, find the single workflow where a week of the right automation saves ten hours every week after — and build that first. Finding it is a process, not a guess: you look at where your team’s time actually goes, not where the tool demos are flashiest. That’s the whole point of finding your first use case before you buy anything.

One win does three things a 40-page strategy can’t. It proves the ROI is real, with an actual number instead of a projection. It earns your team’s trust in the tools, because they watched one work. And it funds the next step out of the savings it created — so momentum compounds instead of stalling in committee.

The number is the strategy

Here’s the reframe: you don’t need an AI strategy. You need a scoreboard. Once one project pays for itself, the “strategy” writes itself — you do more of what moved the number and less of what didn’t. That’s not a document; it’s a habit. Ship the boring, proven thing, bank the return, and let thatdecide what’s next.

If you’re not sure where your first win is hiding, that’s exactly what a short AI Opportunity Audit is for: one problem worth solving, found before you spend. A strategy without a win is a wish list. Get the win first — the strategy is just the story you tell afterward about why it worked.