The AI Consulting Trap | Lucas Sailer

Key takeaways

An Explainer teaches you about AI; an Operator uses it to run a business and builds with you.
Ask any AI consultant to show you the systems that run their own business.
Tools change every six months; start from what costs you the most time or money.
Good help leaves you able to maintain and extend what was built.

If you run a business, you've probably thought about hiring someone to help with AI. Maybe an agency, maybe a consultant, maybe a fractional CTO. And if you've done any research, you've probably noticed something: everyone claims to be an AI expert now.

Six months ago they were a marketing consultant. Now they're an "AI strategist." Their LinkedIn bio changed, but their skillset didn't.

I'm going to be blunt about this because I think business owners deserve honesty, not more sales pitches. I sell AI consulting myself, so read this with that in mind. There's a very specific trap that catches smart, well-meaning business owners when they go looking for AI help. And avoiding it comes down to understanding one thing: the difference between someone who teaches AI and someone who uses it.

The Two Types of AI Consultants

Type 1: The Explainer

This person can tell you what AI is. They've read the articles, watched the demos, maybe taken some courses. They can run a workshop for your team on "How to Use ChatGPT" and show everyone how to write better prompts.

They'll leave you with a slide deck, some tool recommendations, and a bill. A month later, your team is back to doing things the same way. Nothing actually changed in your operations.

Type 2: The Operator

This person uses AI to run their own business. They've built real systems, hit real problems, found real solutions. When they sit down with you, they don't show slides - they open your tools, look at your workflows, and start building.

The difference is night and day. An Explainer tells you about the future. An Operator shows you what's working right now, today, in a real business.

Two kinds of AI help

The Explainer

Workshops and prompt tips
A slide deck and tool list
Leaves with a bill
Your team is back to the old way in a month

The Operator

Runs their own business on AI
Opens your tools and workflows
Builds with you on real work
You can maintain what was built

Hire the person who uses it, not the person who explains it.

How I Learned This the Hard Way

I didn't start as a consultant. I started as a builder.

When I built Ad Blitz, my lead gen agency, I had a choice: hire a big team and scale the traditional way, or build AI into the core operations and run lean. I chose lean. Not because I was cheap, but because I'd seen what happens when AI is actually woven into business operations instead of bolted on.

Campaign briefs that write themselves. Creative production pipelines that generate dozens of variations in minutes. Client reports that compile automatically. Performance analysis that spots trends before humans would notice them.

The result: one person operating at the level of a full team. Not 80-hour weeks - just better systems.

And when business owners started asking "how do I do what you're doing?" - that's when I started consulting. Not because I wanted to sell advice. Because I was already living the answers.

The 5 Red Flags of Bad AI Help

Before you hire anyone for AI help, look for these warning signs:

1. They can't show you their own AI-powered systems

Ask to see how AI runs their business. If they can't show you, they're selling theory. I can show you Ad Blitz's operations and the AI systems that power it. If someone can't do the same, that's information.

2. They lead with tools instead of problems

"You need to use [Tool X]" is a red flag. The right approach is: "Tell me what's costing you the most time/money, and let's figure out what solves it." Tools change every six months. Good systems thinking doesn't. (More on that in Stop Using AI Like a Toy.)

3. They promise a big change without understanding your business

If someone tells you what AI can do for you before they understand how your business actually works, run. Every business is different. The AI strategy for a dental practice is completely different from the strategy for a SaaS company. Anyone promising one-size-fits-all solutions is selling you a template.

4. They can't explain it without jargon

"We'll implement a RAG-augmented pipeline with fine-tuned LLMs integrated via API middleware." If that's how they talk to business owners, they either don't understand it well enough to explain simply, or they're trying to impress you into paying. Neither is good.

5. They build things you can't understand or maintain

The worst outcome in AI consulting is dependency. If you can't understand the systems, you can't maintain them. You become locked into paying the consultant forever.

Good consulting should teach you as it builds. When we're done, you should be able to maintain and extend the systems yourself.

What Good AI Help Actually Looks Like

Here's what you should expect from someone who actually knows what they're doing:

They start with your business, not with AI. The first conversation is about your operations, your bottlenecks, your costs, your goals. AI is the solution - not the starting point.
They're honest about what doesn't need AI. Not every problem needs an AI solution. Sometimes a simple automation, a better tool, or a process change is enough. Good consultants will tell you when AI isn't the answer.
They work with you, not for you. The best AI engagements are collaborative. You learn as you build. You understand what was built and why. You can maintain it, improve it, and extend it after the engagement ends.
They have clear, transparent pricing. No vague "it depends" followed by a surprise invoice. You should know exactly what you're getting and what it costs before you commit.
They show results from real businesses. Not hypothetical case studies. Not "potential ROI projections." Real businesses, real systems, real results.

The uncomfortable truth

Most businesses don't need a huge AI transformation project. They need to see what's possible, get one system working, and then decide if they want to go deeper.

That's how I set up my offers:

Do it yourself. The AI OS Starter Kit: the playbook, the setup guide you hand to your AI, five ready skills and a 30-day plan.
Done with you. AI OS Launch: thirty days, a kickoff call and three check-ins.
Done for you. AI OS Build: I build the system around your top three workflows.

If you're evaluating AI help right now, ask to see their systems. Ask how AI runs their business. If they can't show you, keep looking.

Questions people ask

How do I tell a good AI consultant from a bad one?

Ask to see the AI systems that run their own business. An Operator can open them and show you. An Explainer shows you slides.

Should the first conversation be about tools?

No. It should be about your operations, your bottlenecks and your costs. The tool comes after the problem is clear.

What should I own when the engagement ends?

The systems, and enough understanding to maintain and extend them yourself. Dependency on the consultant is the worst outcome.

Sources

← Back to Writing