AI Strategy vs Execution: The $99 Call AI Couldn't Make (Simply Explained)

The Build Took 20 Minutes. The Decision Took Two Weeks.

A few months back I was building a membership product for a telehealth service. Simple idea: a monthly subscription that gave members ongoing access to care.

We had two ways to price it, both with tiers. A basic level, a premium level. The usual ladder.

Here's the thing. Building either version was easy. An afternoon, maybe less. The signup page, the billing, the customer records. I had AI do most of the typing. The technical work was never the problem.

What stalled the whole project for two weeks was one decision the AI couldn't make for me. Not a technical call. A business call. And no matter how I asked it, the AI couldn't carry the weight of that decision, because it lived in a place AI can't reach.

This is the difference between strategy and execution in one project. When you build with AI, the building stops being the hard part. The thinking becomes the hard part.

What the AI Did Well

Let me be fair, because the AI did real work here.

It drafted both pricing plans in minutes. Not vague ideas, actual structures with tiers and prices and features. Then it ran the math. Given some assumptions about how many people would sign up and how many would cancel, it showed me what each plan looked like at 100 members, 500 members, 2,000 members.

It wrote the signup page. It connected the billing. It built the system to track members and their plans. It even pulled up pricing patterns from similar subscription businesses so I wasn't guessing.

This is the part people get excited about, and they're right to. AI is fast and competent at this kind of work.

My DTC fashion brand runs a pricing system that sets prices on more than 564 products automatically. It makes those decisions while I sleep, and it works because the rules are clear. When the boundaries are well-defined, AI pricing is fantastic. Faster and more consistent than I'd ever be by hand.

So on the telehealth project, none of the execution scared me. If the question had been "which of these two plans makes more money," the AI would have answered it cleanly and been right.

That's not where it failed. It failed on a question that looked like a pricing question but wasn't one.

The Wall the AI Couldn't Cross

Here's the problem, plainly.

This membership had real healthcare on the back end. Actual medical care, delivered under rules that don't bend for marketing.

A higher-priced tier could charge more. Fine. But by law, the actual care had to be identical between tiers. A "premium" member had to get the exact same care as a "basic" member. Same standard, same treatment. That's not a preference. That's the law.

So a tiered price would imply something we couldn't deliver and legally must not deliver: better care for more money. The premium member thinks they're buying something extra. They're not. They can't be. That gap is a legal problem and an ethics problem stacked on top of each other.

Here's the part that matters.

When I explained the medical rules to the AI, it understood the problem and repeated it back to me. But it would have happily built the tiered version if I hadn't told it. Nothing in the pricing math flagged a problem. The tiered version looked slightly better on paper, so the AI pushed toward it. That's what I'd pointed it at.

It had no way to weigh the legal risk against the extra revenue. It couldn't measure the danger of a misled customer or a regulator's attention. Those aren't numbers in a spreadsheet. They're judgments about what could go wrong.

That judgment was the actual decision. The AI handed me two clean options and no way to know that one was a trap dressed up as a revenue plan.

Why This Happens Every Time

This wasn't a fluke. It's the shape of almost every AI project I run.

The bottleneck is never the building. I build fast, and AI makes me faster. The bottleneck is deciding what should be built and what shouldn't.

Here's how to think about it. AI works inside the boundaries you give it. It does not question those boundaries. Tell it to maximize revenue and it will, brilliantly, without ever asking whether one option implies a promise you can't keep. Setting the boundaries is your job. AI just lives inside them.

And it will cheerfully build the wrong thing faster and cleaner than any human team. A junior employee might pause and say "wait, doesn't this seem off?" AI doesn't pause. It just builds. With confidence, with clean work, the wrong thing.

Speed without judgment just means making mistakes faster.

The Decision I Made

I scrapped both tiered plans. I shipped one flat price.

The reasoning was simple once I stepped away from the spreadsheet. One price means one clear promise. No implied difference in care. No legal gap to defend. No customer ever wondering if someone paying more was getting better treatment. The signup got simpler. The whole product got easier to explain, which in healthcare is worth more than it sounds.

On paper, the tiered plan looked slightly better. A few percent more projected revenue. The AI showed me that accurately. But the tiered version carried a risk the AI couldn't put a number on, because the risk wasn't a number.

Here's the order of operations I want you to remember. I made the decision first. Then I used AI to pressure-test it. I had it argue against the flat price, find the holes, model the downside. It did that well, and the decision held.

Human decides. AI stress-tests. Not the other way around. When you let AI make the call and then rubber-stamp it, you've handed the decision to the one thing that can't carry it.

What This Means for You

If you're a business owner worried AI is going to start making big decisions and make them badly, let me point the fear in the right direction.

AI isn't a decision-maker. It's a multiplier on judgment. The risk was never that it's too smart and will outthink you. The risk is that it's confidently fast inside whatever boundaries it's given. Point it at the wrong target and it'll hit the bullseye every time.

So the safeguard isn't smarter AI. It's a human who knows where the boundaries are.

The real question isn't "will AI make a bad call." It's "who is setting the boundaries it works inside." That telehealth project didn't need smarter AI. It needed someone who knew the legal rules and could see that one of the obvious options was a trap.

That's what I do as a Chief AI Officer. I'm in the build and in the strategy. I'll ship the billing in an afternoon, and I'll also be the one who says "this tiered plan is a legal trap, we're going flat." Those two things in one person is the whole point.

If your team can build but keeps building the wrong things, that's the gap I fill.

Want to explore what AI could do for your business?

Book a free 30-minute strategy call. No pitch deck, no sales team, just a real conversation about your operations and where AI fits.

Book a Discovery Call