7 Questions to Ask Before Hiring an AI Consultant

The AI consulting space has a problem. There are a lot of people talking about AI and not a lot of people building AI. When everyone and their dog is suddenly an “AI expert,” knowing who to hire gets tricky.

I’m going to give you seven questions to ask any AI consultant — including me — before you hand over a dime. The answers will tell you everything you need to know.

1. “Can You Show Me Something You’ve Actually Built?”

This is the single most important question. A real AI consultant should be able to show you working systems they’ve built and deployed. Not slide decks. Not concept demos. Working systems in production.

Red flag: “We have case studies” (but only abstract descriptions, no demo). “Due to NDAs, we can’t show specifics” (convenient). “Let me walk you through our methodology” (deflection).

Green flag: “Here, let me show you this customer service agent I built for a client. Watch what happens when I ask it a question about their products.” Tangible. Demonstrable. Real.

2. “What Happens After You Deliver?”

A lot of consultants build something, hand it over, and vanish. Then when something breaks or needs updating, you’re stuck.

Ask specifically:

• Is there post-launch support included?
• How do I update the system when my data changes?
• What’s the ongoing maintenance look like?
• If I need changes in 6 months, what does that cost?

What you want to hear: A clear support plan with specifics. Training for your team so you’re not dependent. Reasonable maintenance pricing. A relationship, not a transaction.

3. “What Will This Actually Cost — All In?”

If a consultant can’t give you a clear number (or at least a tight range) after understanding your project, that’s a problem.

Watch out for:

• Hourly billing with no cap (“it takes as long as it takes”)
• Vague “it depends” without follow-up specifics
• Low initial quotes that don’t include “phase 2” or “ongoing optimization”
• Avoiding the money conversation entirely

What you want to hear: “Based on what you’ve described, the build will cost between $X and $Y. Ongoing costs will be approximately $Z per month. Here’s what drives the price in either direction.”

4. “When Will I See ROI?”

Any consultant promising immediate, massive returns is selling you hype. But they should be able to articulate a realistic path to ROI.

Good answer: “Based on the 8 hours per week your team spends on this process, at your loaded labour cost, you should see payback in about 3-4 months. Here’s the math.”

Bad answer: “AI will transform your entire business.” (How? When? What does that mean in dollars?)

You want someone who talks in specific numbers tied to your specific situation. Not generic promises about the AI revolution.

5. “What Happens If It Doesn’t Work?”

Things don’t always go perfectly. How a consultant handles this question tells you about their integrity.

Red flags: “That won’t happen.” “Our solution always works.” (Nothing always works.)

Green flags: “Here’s how we de-risk the project. We start with a focused pilot, set clear success metrics upfront, and iterate based on results. If the pilot doesn’t meet the criteria, we adjust the approach or refund the pilot cost.”

You want someone who’s thought about failure scenarios and has a plan for them. Confidence is good. Arrogance is dangerous.

6. “Will My Team Be Able to Use This?”

A technically brilliant AI system that nobody on your team can operate is worthless. This is a bigger problem than most people realize.

Ask about:

• What training is included?
• How technical does someone need to be to manage this?
• Is there documentation?
• Can I add or update data myself?

What you want: A system designed for non-technical users, with training included, documentation provided, and a path to independence. You shouldn’t need to call the consultant every time you want to update a FAQ.

7. “Why Should I Do This Now?”

This question tests whether the consultant is selling you something or genuinely advising you. A good consultant will sometimes tell you not to do something.

Red flag: “AI is moving fast and if you don’t act now you’ll be left behind!” (Fear-based selling.)

Green flag: “Based on what you told me about your current volume and pain points, the ROI makes sense now. But if you’re dealing with [other priority] first, it might make sense to tackle that and come back to AI in Q3.” (Honest, context-specific advice.)

The best consultants earn trust by telling you the truth, even when the truth is “not yet.”

The Meta-Question

Beyond these seven, there’s one overarching thing to evaluate: does this person understand my business, or just AI?

The best AI implementations come from consultants who take the time to understand your industry, your customers, your team, and your specific challenges. AI is the tool — your business problem is the target. If the consultant leads with technology instead of your problem, that’s backwards.

My Standing Offer

I’m happy to answer all seven of these questions — and any others you have — during a free consultation. No commitment, no pitch. If I’m not the right fit for your project, I’ll tell you. And if now isn’t the right time for AI, I’ll tell you that too.

That’s how I think this should work. Book a time and let’s have a real conversation.

---

Andy Doucet is an AI consultant based in Grande Prairie, Alberta, serving businesses across Alberta and Western Canada. He builds things, not slide decks.