A few years ago, a good friend of mine decided to take the leap and start his own growth agency. Very sharp guy. Top-notch growth operator. And for a while, things were good. But we caught up last week, and he informed me that he was going to have to close up shop. We spent a while on the phone unpacking what happened, trying to make sense of the landscape right now.

I told him that while it may not mean much, he's not alone. I've talked to a dozen agency operators this year alone. Four are shutting down. Seven don't know if they'll make it through 2026. Only one said things are business as usual.

Of course, some of it’s just the nature of the beast. Growth and marketing agencies have some brutal structural dynamics. Things like the reverse retention paradox, where the better you perform for a client, the harder it can be to retain them (not too dissimilar to dynamics at play with dating apps). Or the growth expert paradox, where the skills that make you great at growing other companies are underutilized in terms of growing your own business. Both worth exploring another time.

But the thing that kept coming up, the thread running through almost every conversation I’ve had recently, isn’t specific to agencies. It’s hitting everyone.

Losing sight of what actually matters

Here’s a story that captures it perfectly. Another friend of mine does a lot of advisory/fractional work. He's been working with a client to bring on an SEO specialist. Had a great partner picked out, contract nearly signed.

Then the client sent him a LinkedIn post. Something about an AI-powered workflow that "could replace your whole SEO team." (It was a handful of basic AI prompts 🤦‍♀️.)

The client wanted to blow up the entire deal. “Why would we pay $9,000 a month for an agency if this guy's AI can do it virtually for free?”

My friend had to talk him off the ledge.

This is happening everywhere right now. And it points to something I think we need to be honest about: There's a massive difference between producing something and producing something better. And right now, almost nobody is asking which one they're actually getting.

Most people's thinking stops way too early. Can AI do the task? Yes. Does the output look decent? Kinda. Great, fire the team.

But since when is "decent output" how we evaluate anything? You already had good. You had experts. You had people who knew what they were doing. The question was never whether AI can spit something out. The question is whether it produces better results than the people you just laid off. And if you can't prove that, you didn't make a strategic decision. You made a vibes-based one at the expense of someone else's career.

Look, I could probably remove your appendix if you asked me to. Give me a couple YouTube videos and, well, something's coming out.

But that was never really the goal, was it? The goal wasn't "remove the appendix." The goal was to remove it safely, with minimal damage, fewest complications, and the highest chance of a good recovery. You know, the stuff a surgeon spends a decade learning how to do.

But that's what's happening in marketing right now. Companies are acting like the only goal is to get the appendix out.

Dial down the hype

Let me give you another example. I read a newsletter last week about a company claiming they have 40 employees, 20 human and 20 AI. The piece was sprinkled with lines like “the line between human and AI has never been more blurry.”

I'm sorry, but what the actual fuck are we talking about?

There's nothing blurry about it. You have 20 employees. You wrote 20 scripts, gave them names, and for some reason are calling them employees.

And this stuff has consequences. Because the pattern I keep seeing goes like this: someone on LinkedIn announces they’ve “automated their entire sales process” or “replaced their content team with AI.” Amazing. That’s the input. What’s the output? What happened to revenue? Pipeline quality? Brand perception?

I'm still waiting for someone to cover that part.

At this point, I imagine someone's reading this thinking I must be anti-AI.

I'm not. I use the shit out of AI. As a non-technical founder, I've never felt more empowered. It's genuinely incredible.

But I've spent my entire career helping operators cut through noise and focus on what actually moves the needle. That's the whole point of what we do at Demand Curve. Ignore the shiny objects. Think from first principles. Be ruthlessly evidence-based about what's working and what isn't.

And when I apply that same lens to what's happening right now, I'm not seeing the evidence. I'm seeing a lot of activity, a lot of hype, a lot of breathless claims. But I'm not seeing proof that the trade-offs being made are worth it.

The self-fulfilling prophecy

Here’s where it gets extra tricky.

A marketing executive reads one of these LinkedIn posts, or hears it secondhand from another executive who’s equally caught up in the hype. FOMO kicks in. They convince themselves a smaller team plus AI can do more. So they lay people off.

Those talented people, now unemployed, turn to freelancing. A lot of them. All at once.

Basic economics: when supply floods a market and demand stays flat (or in this case, decreases), prices drop.

So what happens next? These same executives who cut their teams can now hire talented freelancers for a fraction of what they used to cost.

And they celebrate. “See? I told you we could get more for less. AI is bringing down the cost of marketing.”

No. You made an emotional decision like a child. You flooded the talent market. Your fear-driven behavior helped suppress demand. And then you congratulated yourself for your role in creating a toxic positive feedback loop.

Where the market is actually shifting

So that's the cycle at play. Or my theory of the cycle, at least. Here's how it's reshaping the market:

SMBs and early-stage startups are stepping down-market (even more). They were already working with relatively affordable talent and scrappy solutions. Now they’re looking at AI and thinking, “Our output isn't great as it is. And every dollar counts. If there's a way to reduce costs without taking a big hit on quality, we may as well.” Fair enough. That’s rational for their situation.

Mid-market companies are the most interesting segment to me. And it seems they're also taking a step or two down market. And it's hard to blame them. All those marketers getting laid off from bigger companies? Falling right into the laps of these companies. That $30k/mo. agency that your investors pressured you to bring on isn't looking so appealing with so much talent on the market.

Large companies are mostly in holding patterns, waiting to see what happens from the cuts they have already made. They’re keeping their strategic talent and advisory relationships. The layoffs hit execution layers hardest.

Basically, that leaves a gap in the middle. Or at least what used to be the middle. And if you’re a marketer or agency operator, understanding where that gap is opening up matters a lot for what you do next.

What's the solution?

Alright. This has been a bit of a rant. We had a blizzard here in NYC, I haven't left my apartment in a while, so maybe I'm feeling a little extra spicy.

But ranting wasn't the whole point. My aim is to help shine some light on the market dynamics I believe are at play, so people can figure out how to respond.

If you're a marketer: The answer isn't just "get better at AI." You have to, obviously. But that's table stakes. The real move is the same thing the best marketers have always done — help decision-makers make better decisions. If your leadership team is making reactive, half-baked "strategic" decisions based on vibes, your job is to be the person in the room with actual data. Build the systems to quantify what AI is actually doing for your function. Not "we produced more content." What happened to pipeline? To revenue? To the metrics that matter? Don't accept bad decision-making. Challenge it. Respectfully, sure. But challenge it.

If you run an agency: What you're feeling right now is not a personal failure. The environment is genuinely brutal. But if there's a hole forming in the middle of the market, you've got two paths. Move upstream — get valued for how you think, how you solve problems, your strategic judgment. Or move downstream — figure out how to deliver real value at scale for a lower price point using the very tools that are disrupting you. Sitting in the middle and hoping things go back to normal is probably the worst option.

If you're a startup founder: Be honest with yourself about whether your AI investments are producing real improvements or just short-term cost savings. There's a difference. Cutting your marketing team might improve your bottom line this quarter. But are you maintaining the quality? Are you still investing in the few levers that have long-term, defensible value? Or are you being short-sighted and calling it efficiency?

Finally, to all the big co executives (and I may as well throw VCs into the mix too): I spent some time trying to figure out how to put this in a more eloquent way, but it's getting late, and I'm not one for mincing words anyway. So here goes: Get your head out of your ass. Learn what your teams actually do. Understand how the work actually gets done before you decide a prompt can replace it. Maybe even come on down from that ivory tower you love so much and stop making decisions based on what another executive told you at a dinner party. You are making choices that impact real people's livelihoods and careers. If you can show me data that your AI-augmented, leaner team is producing better results — genuinely better, not just cheaper — I'd love to see it. But if you can't, maybe sit with that for a minute before you greenlight the next round of cuts.

​Justin Setzer​
Demand Curve Co-Founder & CEO