You need an AI strategy consultant only if you can't yet name the one process costing your business the most time or money. If you can name it, skip the strategy engagement and hire someone who builds. More planning won't move a bottleneck you've already found.
TL;DR
Try answering four questions about the process you'd most like to fix:
If you can answer all four, you don't need strategy. You need someone to build it. If you're stuck on the first question, or you have six candidate processes and your partners disagree about which comes first, a strategy consultant earns the fee by getting you to one answer.
You may answer these faster than you expect. The process that makes you groan on Sunday night is a good place to start.
The two jobs end in different places, and that's the clearest way to tell them apart.
| AI strategy consultant | Someone who builds | |
|---|---|---|
| What you get at the end | A document: priorities, use cases, a roadmap | A working system on your accounts, plus training |
| How you judge it | Does it name one process and what fixing it is worth? | Does your team use it on real work without the builder? |
| What you still have to do | Find and pay someone to build it | Maintain it, or pay for support |
| Where it goes wrong | A long roadmap nobody acts on | A build aimed at the wrong process |
| Best when | You don't know where to start | You know exactly what's eating your week |
Both failure modes are real. The strategy risk is paying for clarity you already had. The build risk is automating something that didn't matter. The four questions above are there to tell you which risk you're facing.
Because the gap is in doing the work, not in knowing it's worth doing. In a Goldman Sachs survey of 1,256 small businesses in its 10,000 Small Businesses program, 76% said they use AI, but "only 14% say AI is fully embedded in their core operations." And "73% say they would benefit from additional access to training and implementation resources" (Goldman Sachs). Those owners aren't short on reasons to use AI. They're short on help making it part of how the work gets done.
Bigger companies show the same pattern. In a survey of 200 UK mid-market decision makers by the consultancy TXP, "just 47% of AI pilots have scaled into production" (Enterprise Times). And McKinsey's latest State of AI survey found that "nearly three-quarters of high performers report fundamentally redesigning workflows because of their AI use," compared with about a quarter of everyone else (McKinsey). The companies getting value changed how the work runs. A roadmap doesn't do that on its own.
Strategy isn't the villain here. The same TXP survey found "77% said their organisation should have allocated more time and budget to discovery before launching its AI initiatives." Skipping the thinking costs money too.
Paying for strategy makes sense when:
The key is scope. Good strategy for a small business is short, fixed-price, and ends with one named process, what it costs you now, what AI would and wouldn't do to it, and a rough price to build. Whichever you hire, be clear up front on what AI consultant services should leave you with when the engagement ends.
Before you pay anyone for this step, try the free version. AI Marketing Labs' free AI Audit is built to do exactly this job: find the one bottleneck to solve first and tell you what fixing it would take.
If you skip the strategy consultant, the builder still needs to do some strategy. It's just smaller and aimed at one process. A good builder will:
A builder who starts building before seeing how the work gets done today is guessing. Mapping the real work is the stage of AI implementation that's most tempting to skip and most worth protecting.
Yes, and for a small business it can be the simpler route. Two things to get in writing if you go that way. First, the strategy fee should be credited toward the build, so you're not paying twice for the same thinking. Second, the plan should be yours to take to a different builder if you don't like the build proposal. If the discovery only works with that firm, it was a sales process.
For a one-person business, the argument for hiring a builder gets stronger. The Small Business and Entrepreneurship Council notes that sole proprietors "may face the challenge of managing all business functions themselves, potentially lacking the bandwidth, specialized skills, or time required for AI implementation and learning" (SBE Council). A solo owner may not need to be told what's slow. What's missing is a second pair of hands.
If your real choice is between an AI specialist and a consultant who added AI, decide first whether your problem is a process or the way the business runs.
It varies with the firm and the scope. Whatever the number, ask for a fixed price, an end date, and a named deliverable, so you know exactly what you're buying.
Is an AI strategy consultant the same as a fractional AI officer?No. A strategy consultant works on a project with an end date. A fractional AI officer is part-time, ongoing leadership, closer to a staff role you share with other companies.
Can I do the strategy part myself?You can, if you're willing to track time. Log the hours your team spends on your three most annoying tasks for two weeks. The one with the most hours and the clearest rules is a strong place to start.
What should an AI strategy deliverable include?One named process, what it costs you now, what AI would do to it and what it wouldn't, a rough build cost, and the next step. If it lists twelve initiatives, it went too wide.