Nobody Is Selling Them Training

Enough people asked that I've put the $9 AI Audit Workshop back up for a limited time.

It's not the old version. It's been updated with what's actually working at Dunn & Company and what's come out of over 150 students delivering audits themselves.

Get access here while it's open.

Right, this week's issue.

The line item that shows up in every audit and almost never gets sold.

August 23, 2026
Read Time: 7 minutes

Seventy three percent of small businesses say they need more training to properly use AI. Carrier Management

About half of the ones already using it have invested nothing at all in training or integration. Eatsleeplaunchrepeat

Both of those numbers were published this year. Neither of them is controversial. And almost nobody in AI consulting is selling against them.

We spend our time talking about audits, implementations, agents, integrations. Meanwhile the buyers are quietly saying the thing they actually want is for their people to know what they're doing.

It's Already in Your Audit

Here's what makes this route different from every other one.

Every single company you audit has training as a line item. Not most of them. Every one.

You go in, you map the processes, you quantify the inefficiencies, and somewhere in that report there is a section that says the team needs to be brought up to speed. It's not a bolt-on you have to invent. It's a conclusion the assessment produces on its own.

Which means you have already done the selling. You've been inside the business, you've seen how work moves, and you're recommending it off the back of a diagnosis rather than pitching it cold. It is the most logical next step in the room, and it costs you nothing to put it on the page.

Most consultants write that line item, hand over the report, and then let someone else deliver it. Or nobody delivers it at all.

Why Training Closes When Implementation Stalls

There's a reason this one gets a yes faster than a build does.

It's cheaper. An implementation is a capital decision. Training is an operating one. Different budget, different approval, different level of anxiety about being wrong.

It reduces the fear rather than feeding it. Every AI conversation inside a business carries an unspoken question, which is whether this ends with people losing their jobs. An implementation confirms that fear. Training does the opposite. You're walking in to make the existing team more capable, not to design them out of the process.

It attacks recruitment cost, which owners feel every month. A team of twenty who can actually use these tools properly is a team you don't need to add to next year. Owners understand that argument immediately, because they have already had the conversation about whether they can afford another hire.

It creates builders instead of dependents. When people genuinely understand what the tools do, they start making their own small fixes. The little workflow, the recurring prompt, the thing that saves them an hour on a Thursday. That value compounds without you being in the room.

Three Ways to Price It

There's no single correct structure. There are three that consistently work.

Per day. $3000 - 5000 for a corporate training day. You show up, you deliver, you leave. Simple to sell, simple to invoice, and it works well as a first engagement because it commits them to almost nothing.

Per month. $5000 - $15000 a month for four ninety-minute sessions, usually delivered virtually, often working through the business department by department. Roughly the same price point as the day rate but recurring, and considerably better for you.

As a cohort. A ninety-minute training session plus a ninety-minute question-and-answer session every week, running up to twelve weeks. This is the format that produces the most actual change, because people bring real problems back into the room and get them solved in front of their colleagues. I’ve seen cohorts sold for up to $20,000.

Per session or per month covers the vast majority of it. Start there.

The Honest Limit: This Doesn't Make Them an AI Business

Worth being clear about this, because it's the thing that stops training being oversold.

Training does not turn a company into an AI-native company. It turns employees into more AI-native employees. Those are different outcomes.

A business that has been through six months of good training still has undocumented workflows, disconnected systems, and no automation running without someone pressing a button. On the maturity curve I wrote about a few weeks ago, you have moved them from level one to somewhere around level two or three. You have not moved them to level four or five, and you cannot, because that requires building things.

Training is a stepping stone. It is a genuinely valuable one, and it is the stone most businesses need to stand on before anything else works, but it is not the destination.

That's also why it should never be your only route. It pairs with implementation, it pairs with advisory, and it pairs with maintenance. It sits alongside them rather than replacing them.

Why It Keeps Renewing

Here's the mechanic that makes this better than it first looks.

AI does not stop moving. Something meaningful shifts every few weeks, and every one of those shifts creates a question inside the business that nobody internally can answer.

So the contract renews for a reason that has nothing to do with your persuasiveness. They keep paying because they want to stay current, and you are the only person in their orbit whose job it is to be current on their behalf.

Monthly is the standard. Quarterly works for smaller organisations. Either way it becomes a subscription to knowing what's going on, which is a far easier thing to renew than a project.

How Big This Can Actually Get

There is a company I'm aware of that does nothing but corporate training. No audits, no implementations, no software. They come in, they upskill the staff and the owners, and they leave.

They're north of twenty million in three years.

That's not a lucky outlier. That's what happens when a service is genuinely needed, almost nobody is providing it properly, and the buyer has already told researchers in public that they want it.

The ceiling on this route is nowhere near where most people assume it is. It is one of the newest, emptiest parts of a market that is already underserved.

The Real Point

You are already writing training into every audit you deliver.

You're just letting someone else deliver it, or watching it never happen at all.

It's cheaper to buy than an implementation, it reduces the fear rather than causing it, and it renews every month because the technology never stops moving.

Put a price on the line item you're already writing.

See you next week,

– Andrew