AI Strategy Is a Business Strategy Problem

You don't need an AI strategy. You need a business strategy that accounts for AI.

Most owner-led companies get this backward. They treat AI as a destination - a roadmap to build, a pilot to run, a vendor to pick - and start the work before they've asked what it's for. The result is motion without direction. A tool gets bought. A pilot runs. Six months later the business looks the same, and the owner is no clearer on whether any of it mattered.

AI is a lever. It sits inside decisions you already make - how you price, where you spend attention, who you hire, what you can offer that a competitor can't. The question is never "what's our AI strategy." The question is which of those decisions AI changes, and whether the change is worth the bet.

That reframe sounds small. It's the difference between spending a year chasing tools and spending a week finding the two places AI moves your business.

What an AI strategy consultant does

A good AI strategy consultant doesn't hand you a technology plan. They start with the business - your economics, your competitive position, the decisions that decide your year - and find the few places where AI shifts the odds. Tool selection comes after that, and it's the easy part.

That order matters. Most AI consulting runs the other way: pick the technology, then hunt for a problem it can solve. You end up with a solution looking for a home, and a bill. Start from the business and the technology becomes a means - you know what you're buying and why before you spend.

For an owner-led company the job is narrower and harder than it sounds. You can't chase every use case. You have limited capital and limited attention, and every bet you make is a bet you're not making somewhere else. The work is selection - naming the handful of places AI earns its cost, and walking past the rest.

Why AI strategy is different for an owner-led company

Big-company AI strategy is an infrastructure problem. Scale, data pipelines, model governance, a team to run it. Enterprise consultants solve for that because their clients have the size to justify it.

You have a different problem. You make important decisions without the data a larger company would have, and you make them with your own capital on the line. AI reaches you as a question of judgment under scarcity: where does a tool sharpen a call you're already making by feel, and where does it just add cost and false confidence.

This is the Second Stage version of the AI question. A company between roughly $1 and $20 million in revenue has outgrown the owner-does-everything phase but hasn't bought its way into enterprise systems. AI meets that company at a specific moment - enough complexity to need leverage, not enough slack to waste a year finding out. The right AI strategy for a business this size looks nothing like the enterprise playbook. It's a short list of high-conviction bets, sized to what you can afford to lose.

Five questions that decide whether an AI move is worth the bet

Before you spend a dollar on AI, run the move through five questions. They come from how an investor sizes a position, not how a technologist scopes a project. The point is to separate the two or three moves that could matter from the dozen that will only cost you.

1. Does it touch a decision that moves the business? AI applied to a peripheral task saves a little time. AI applied to one of the few decisions that determine your year - pricing, where demand comes from, what you can deliver that rivals can't - changes the trajectory. Start only where the lever counts.

2. If it fails, is the loss bounded? The best bets have a floor. You can name the most this costs you - the license, the hours, the switching cost - and it's survivable. If a failure takes down a client relationship or your reputation, the downside isn't bounded, and the bet is a different kind of decision.

3. If it works, does the upside compound? Some AI moves save a fixed amount and stop. Others get better the more you use them - the model learns, the advantage widens, the lead compounds. Weight the second kind. Bounded downside with compounding upside is the shape worth paying for.

4. Can you feed it, or are you automating a guess? AI runs on data and judgment. In a data-poor business, ask whether you have the inputs to trust the output, or the experience to catch it when it's wrong. Automate a process you don't understand and you scale the error, not the insight.

5. What breaks if it works? Speed exposes whatever was already bent. Automate a broken intake process and you get broken intake faster. Scale a pricing model built on a flawed assumption and the flaw compounds into a reinforcing feedback loop before anyone notices. Ask what a win puts under pressure downstream.

A move that clears all five is rare, and it's where an owner-led company should spend. Most AI pitches fail at question one.

When AI makes your problems worse

AI is an accelerant. Point it at something that works and it goes faster. Point it at something broken and it breaks faster, with more conviction, at a scale that's harder to see.

This is where AI meets cognitive runaway - the pattern where a company's own responses feed the problem they were meant to fix. A team overwhelmed by demand automates its outreach and generates more demand than it can serve. The fix accelerates the failure. Drop AI into a process like that and you don't solve anything. You pour speed on a reinforcing feedback loop and buy yourself a bigger version of the same problem.

None of this is an argument against AI. It's an argument for getting the sequence right - fix the process first, then automate it. An AI strategy consultant who understands systems will look at what you want to speed up and ask whether speeding it up is safe. That question is worth more than any tool recommendation.

When you need an AI strategy consultant, and when you don't

You don't need one to buy a few software licenses or to let your team try a chatbot. That's tool adoption, and you can run it yourself. Bringing in a strategist for that is overkill, and a good one will tell you so.

You need one when AI touches something that decides your future - a shift in what your industry expects, a competitor changing the economics, a core process you're tempted to rebuild around a model. At that point the question stops being technical and becomes strategic: what bet are we making, how big, and what does it cost us to be wrong. That deserves the same rigor as any other major move you make.

The test is simple. If the AI question is about a task, handle it in-house. If it's about the direction of the business, it belongs with the rest of your strategy.

Frequently asked questions

What does an AI strategy consultant do?

An AI strategy consultant helps you decide where AI fits your business and where it doesn't. The good ones start from your economics and your competitive position, find the few decisions AI changes, and size those bets before recommending any tool.

Do small or owner-led companies need an AI strategy?

Not as a separate document. An owner-led company needs a business strategy that accounts for AI - a clear view of the two or three places AI could move the business, and a decision about which bets are worth making. The standalone "AI strategy" most vendors sell is usually a solution hunting for a problem.

How is AI strategy different for a company under $20 million in revenue?

For a company this size, AI strategy is a problem of selection. Enterprises build AI capability at scale, with the data and budget to match. A Second Stage company has neither, so the work is naming a short list of high-conviction bets and skipping everything else.

How do I choose an AI strategy consultant?

Pick one who starts from your business rather than from the technology. If the first conversation is about your decisions, your economics, and your risks, you're in the right room. Open with a tool or a platform, and you're being sold.

Isn't AI moving too fast to build a strategy around?

The tools move fast. The business questions don't. What you're pricing, who you serve, what you can offer that others can't - those change slowly, and they're what a strategy is built on. Anchor to the business and the pace of the tools stops being a threat.

Working with Phimation on AI

Phimation treats AI the way it treats every other lever - inside your strategy, weighed against everything else on the table. Dave Haviland works with owners and leaders of Second Stage companies to find the few moves that matter and size them as bets. One of them is AI.

That's why this work lives inside strategy advisory rather than a separate AI practice. The value is a clear read on where AI earns its place in your business, and the discipline to walk past the places it doesn't.

If AI has moved from a curiosity to a decision you have to make, that's the conversation worth having.