There are a lot of people selling AI strategy consulting right now. Most of them were selling digital transformation 18 months ago. Before that it was blockchain. Before that it was big data. The pitch changes, the slide deck doesn’t.
I’m not saying everyone in this space is a fraud. But the market is flooded, the stakes are real, and most companies buying AI consulting services for the first time have no way to tell who’s good. So here’s what I’ve learned after years of doing this work and watching others do it badly.
The two kinds of AI strategy consultants
There are people who assess and people who build. That distinction matters more than anything else in this post.
Assessors show up, interview your team, benchmark you against a maturity model, and deliver a report. The report usually says you need better data infrastructure, more executive buy-in, and a phased implementation roadmap. It costs $50,000 to $200,000. Then the assessors leave.
Builders show up, interview your team, and then stay to construct the first systems. They work alongside your people. They ship things. When something doesn’t work, they fix it. When people resist the change, they deal with it. They’re accountable for outcomes, not deliverables.
I’m a builder. I have opinions about this. But even if you hire an assessor, you should know which one you’re getting before you sign.
What good AI strategy consulting looks like
The best AI consulting engagements I’ve seen, mine and others, share a few things.
A good consultant spends the first few weeks understanding how the business works before recommending anything. If someone shows up with a technology recommendation in week one, they decided what to sell you before they understood your problem.
Good engagements are specific about outcomes. Not “improved efficiency” or “data-driven decision making.” Specific. This process takes 40 hours a month and we’re going to cut it to four. This team makes pricing decisions based on gut feel and we’re going to give them actual market data. Measurable results with timelines attached.
The people side gets addressed early. Any AI strategy consultant who talks only about technology is going to leave you with a system nobody uses. The reason most AI initiatives fail isn’t technical. It’s that nobody dealt with the change management. The fear. The “we’ve always done it this way.” The middle manager who sees AI as a threat to their team.
And there’s a handoff plan from the start. The point of outside help is to build capability inside your organization. If the engagement doesn’t include training your people and transferring ownership, you’re creating a dependency.
Red flags
I’ve watched enough bad engagements to spot the patterns.
The biggest one is leading with technology. “We’ll implement GPT-4 across your customer service workflow.” If the technology is decided before the problem is defined, you’re buying a solution looking for a problem.
Ask to see a working system they deployed. Not a case study. Not a testimonial. A real system, running in production, that replaced an actual process. If they can’t show you one, they advise on AI. They don’t build it.
Watch how they scope the work. If the proposal is organized around deliverables (“60-page strategy document, executive presentation, technology roadmap”) instead of outcomes, ask yourself what happens after you read the document. The deliverable is not the value. The change is.
Pay attention to what they don’t ask. If nobody asks who’s going to use the system, who’s skeptical, who’s threatened, and who needs to champion it internally, the engagement will produce a beautiful system that collects dust.
Be skeptical of long assessment phases. Some discovery is necessary. But if the assessment takes three months and $100,000 before anything gets built, you should wonder whether building was ever part of the plan.
And ask how they use AI in their own work. Not how they advise clients on it. How they use it themselves, every day. If the answer is vague, they’re selling expertise they don’t have.
Questions to ask before you hire
These are the questions I wish more companies asked before signing an AI consulting engagement.
What will be different in my operations six months from now? If the answer is a strategy document, keep looking.
Who on your team has built and deployed AI systems? Not researched or advised on. Built, deployed, and maintained in production for real users.
What happens when our people resist the change? This question alone will separate the consultants who’ve done this from the ones who haven’t. The ones who’ve done it will have specific stories. The ones who haven’t will give you a line about “change management frameworks.”
How do you measure success? The answer should be concrete. Hours saved. Decisions improved. Revenue impact. If the answer is “adoption rates” or “stakeholder satisfaction,” the bar is too low.
When do we stop needing you? A good AI strategy consultant wants to make themselves unnecessary. If there’s no clear path to independence, you’re signing up for a long-term invoice.
What I do differently
I work as a fractional chief AI officer. One day a week, embedded with your leadership team, for as long as the transition takes. I assess, build, get adoption, and hand off. Same person through the entire arc.
I built and sold three companies before doing this work. I spent 16 years running a food marketing agency. I teach AI implementation at the Ivey accelerator. I spend my days in the terminal writing code, not in PowerPoint writing slides. That background is why I see AI strategy consulting the way I do.
Most organizations I work with are in food, agriculture, and related industries. Established companies with 10 to 500 people who know they need to move on AI but aren’t sure how. If that sounds like you, let’s talk.
If you’re evaluating AI strategy consulting options and want a second opinion, I’m happy to have a conversation. No pitch, no proposal. Just an honest take on what you’re dealing with. Book a call.