How to Choose an AI Consultant: 7 Questions to Ask
The AI consulting market is full of strategists who disappear after the presentation. Here is how to find someone who builds and stays accountable for results.
Choosing an AI consultant is one of the most important decisions a growing business makes, and the market makes it harder than it should be. There are genuine builders who stay accountable for outcomes. There are slide deck merchants who bill by the hour and vanish. Knowing which is which before you sign saves you months and significant money.
Why most AI consulting engagements fail
The number one reason engagements fail is a mismatch between what was sold and what was delivered. McKinsey's State of AI research points to the same gap: plenty of AI activity, far less measurable value. You hired someone to get AI working in your business. They gave you a strategy document and a handshake.
A strategy without a build plan is just advice with a high price tag. Before you evaluate anyone, get clear on what you actually want at the end: a report or a working system.
Our guide on what AI consulting actually covers gives you a clear baseline for what good looks like.
7 questions to ask before you hire
Use these questions in your first conversation. The answers will tell you almost everything you need to know.
- What does the deliverable look like? Ask them to describe exactly what you will have at the end. A working system is a different answer from a roadmap or a set of recommendations.
- Can you show me something you built? Not a case study with vague percentages. Ask to see a live system, a demo or a client they will let you speak to.
- How do you price? Time-and-materials billing with no fixed outcome is a red flag. Look for scoped engagements tied to a defined result.
- Who does the build? Some consultants outsource the implementation. That is not always bad, but you need to know who is accountable end to end.
- What does your team need to do? The best consultants tell you exactly what they need from your side and what they will handle. Vagueness here creates scope disputes later.
- What happens after the project? Do they train your team? Is there a handover? Or does the knowledge walk out the door with the consultant?
- What metric will we use to judge success? If they cannot name a number, the project has no clear finish line. That is a problem for both of you.
Build vs advise: the most important distinction
Some consultants are strategists. They are excellent at mapping opportunity and designing the roadmap. But if they do not build, you need a separate implementation partner, which doubles your coordination cost and dilutes accountability.
Others build but have no strategic discipline. They will automate the wrong thing well.
The best outcome is a consultant who does both. Strategy that comes with a build plan. Implementation guided by clear commercial priorities.
See how the models compare in our breakdown of AI consulting vs AI agency.
What to look for in their track record
Be sceptical of:
- Case studies with no named clients and no specific outcomes
- Testimonials that could apply to any services business
- Claims of ROI with no methodology behind them
Look for:
- Specific problems they solved and how they measured the result
- Evidence they stayed involved through implementation, not just strategy
- Clients in a similar industry or with a similar challenge
If they work with businesses like yours, that matters. Our AI consulting services are focused on B2B businesses with sales and revenue bottlenecks, not a catch-all list of use cases.
Red flags to walk away from
- They lead with the technology, not your business problem
- They cannot explain what AI will actually do in your workflow
- They cannot give you a clear scope within the first two conversations
- They have never had to stand behind a result
How to evaluate fit quickly
Before spending hours in discovery calls, get a preliminary read on your own situation. If you know what you need, you can evaluate consultants against it instead of letting them define the brief for you.
You can also read our full guide on hiring an AI consultant, which covers contract structure and onboarding.
Ready to evaluate your fit in a couple of minutes? The AI Consulting Fit Scorecard gives you a clear read before you talk to anyone. Take the Scorecard
Common questions
How much should I expect to pay for an AI consultant?
Rates vary significantly. A scoped engagement for a specific outcome is a more reliable way to budget than an hourly rate. See our AI consulting cost guide for a breakdown of what drives price.
Should I hire a freelancer or a consultancy?
A freelancer can be excellent for a narrow, well-defined task. A consultancy brings team depth, broader capability and more consistent accountability. For a full AI build across your business, a team usually delivers better results.
How long should I give a consultant before expecting results?
A focused engagement should show measurable progress within four to six weeks. If you are three months in with no clear result and no clear end in sight, that is a problem to raise directly.
Can I check references before signing?
Always. Any consultant worth hiring will offer references without being asked. If they hesitate, treat that as data.
