Yes.
Most articles answering this question say no, and most of them are written by people selling the services in question. I run one of Australia’s largest SEO agencies, so I have every incentive to say no too. But the honest answer is that AI is coming for the majority of service work, and a lot of service businesses, as they operate today, won’t survive the transition. Here’s exactly what I mean, what I’m seeing from the inside, and the way out.
Let me define the claim, because it’s the part everyone will argue with.
I believe AI will automate around 70% of the work service businesses currently sell. Not 70% of businesses disappearing overnight, 70% of the tasks inside them. But the second number follows from the first: the businesses whose core offer IS that 70% (the generic, plug-in, anyone-could-prompt-this work) don’t have a business once clients figure that out. The ones that adapt survive. The ones that don’t get replaced, not by AI directly, but by the competitors who adapted.
The pattern is simple: anything generic goes first. Work where the input is a template and the output is interchangeable. The grunt work clients always resented paying for.
In my industry that’s page titles, meta descriptions, basic optimisations, first-pass content. In others it’s the same shape: templated design, routine bookkeeping, tier-one support, boilerplate code, generic blog content. If a smart person with no context could do it from a prompt, it’s already commoditised.
Honest aside: nobody ever wanted to pay premium for that work anyway. AI didn’t kill something valuable, it called the bluff.
The work that survives clusters into a few families, and they’re the same in every industry:
This isn’t theory. In the last twelve months:
And at StudioHawk, we’ve stopped treating the commoditised layer as premium work, because it isn’t. It also hasn’t been a snap of the fingers. The rollout has been more gradual than most people predicted, which is exactly why the businesses moving now have an edge.
There’s an interesting asymmetry in who’s positioned to survive.
Big companies have the data and the moat, but they’re slow to pivot. Small businesses can build fast, but they don’t have the accumulated data. Which means right now is the single best window for a small team: build the software and the data lake before the big brands finish turning the ship. By the time they’ve pivoted, you want your system working and your moat filling.
They did. And about PPC, and social. Agencies adapted, and agencies will adapt to this too. It’s not the end of the world.
But notice what actually happened in each of those waves: the agencies that adapted obliterated the ones that didn’t. Nobody mourns the Yellow Pages agencies. The question was never “will the industry survive”, it’s “will you be in the cohort that does”. This wave is faster and cuts deeper, because it doesn’t change the channel, it changes the cost of the work itself.
Three moves, in order:
I think the shakeout starts showing in the next 6 to 18 months, as AI adoption crosses from experimentation into critical mass. That’s not a comfortable timeline, it’s a useful one: long enough to build, short enough that waiting is a decision.
AI didn’t kill service businesses. It killed average ones. Which 30% of your work would a client still pay for tomorrow? That’s your business. Go build the software and the data lake under it.