The agency model isn't dying. It's molting.

There's a kind of panic that sets in when the thing you've built starts to look obsolete because the ground underneath it shifted and you haven't decided yet whether to move with it.

I've watched a version of this happen to agency owners over the past two years.

It’s agency owners that were running perfectly reasonable businesses, with good clients and phenomenal results, who started hearing a question they didn't have a clear answer to: why can't I just do this myself with AI?

That question has ended more client relationships than bad work has. And most agency owners still don't know what to do with it.

Here's what I've learned from running Mogul Media through this wave, from watching where we got it right and where we were slower than we should have been.

The first thing worth understanding is that efficiency and effectiveness are not the same thing, and AI made it very easy to confuse them.

A founder with a Claude subscription can produce content faster than they could two years ago.

But speed of production has almost nothing to do with whether the content is positioned correctly, targeted at the right conversations, or building anything that compounds over time.

Cheap and fast describes the input. It says nothing about the output.

The agency owners who are struggling right now are the ones whose entire value proposition was the doing. It had everything to do with execution and deliverable. And when AI made the execution cheaper, they had nothing left to stand on.

The ones building durable businesses are the ones whose value was always in the judgment.

It’s the people whose eyes see the strategy behind the execution, the pattern recognition that comes from running hundreds of accounts, the ability to see what's working across a portfolio and apply it to a single client before they could ever figure it out on their own.

That judgment doesn't get cheaper. It gets more valuable as everything around it gets commoditized.

We ran our main content workflow through

Seven human steps two years ago → Seven handoffs → Seven points where someone had to stop what they were doing, pick it up, and move it forward.

Now it runs through three. The other four are handled by agents we built specifically for our process.

The humans in the loop are there for the 20% that still requires taste and context.

Everything else runs without them.

When the repetitive execution is handled reliably, the people doing the remaining work can spend their actual attention on the parts that matter like the judgment calls, the creative decisions, the client relationship.

Which brings up something most agency owners still underweight: clients don't leave because of bad results alone. They leave because of a pattern of small frictions that accumulate until they feel like they're being ignored.

- An unanswered message here.

- A missing report there.

- A repeated mistake that should have been caught.

None of those things are catastrophic on their own. But they stack. And by the time you notice the stack, the client has already made their decision and are waiting for a convenient moment to say it out loud.

No AI system fixes that. That's a communication problem, and communication is a choice.

The broader shift underneath all of this is that the line between agency and software is dissolving.

Services are getting productized. The value game is moving toward who can deliver the best outcome at the lowest cost, and that direction isn't going to reverse.

The agencies that are going to be in a genuinely strong position two years from now are the ones building proprietary systems like internal tools, custom workflows, data that compounds that a client simply cannot replicate by hiring someone with a laptop.

If you're spending 6 figures a year on software that half-does what you need, and you haven't explored what you could build with your own AI systems, you're subsidizing someone else's product roadmap instead of building your own moat.

The uncomfortable truth is that this era is going to be very good for a small number of agency owners and very difficult for the rest.

The difference isn't going to be who works harder. It's going to be who thought seriously about what they were selling and whether that thing was worth building a business around once AI made the easy version of it free.

— Wiz

Running a business and trying to figure out how to build a moat? Reply or book a call and I'll be honest with you.