The middle of the freelance market is hollowing out

The first time I noticed the shift, it was a client of nine years. A small bakery I had built the brand for in 2019, then helped through a refresh in 2022. They emailed in February to say they would not be needing me for their spring menu redesign. They had found something called a packaging template generator, ran it through a few rounds of AI iteration, and it was, in their words, "honestly pretty good."

I looked at the output. It was honestly pretty good. Not great. Not what I would have made. But good enough that the bakery, which was not going to be able to pay full design rates this year anyway, felt fine moving on.

That client was not the canary. I knew the canary had landed. I just had not realized it was already singing.

What the data says

Freelance writing volume on the major platforms dropped 32% year over year in 2025. The $40 blog post has effectively ceased to exist. Generic logo work in the $200 to $500 range, basic photo editing, simple landing page copy, social media post writing: all collapsed in the last 18 months, and not because demand fell. Demand may have actually risen. Supply just got infinite and free.

What is interesting, and the part most coverage misses, is that the high end is doing fine. Designers, writers, and developers at the top of the market are charging more than they did three years ago, not less. Some of them are charging dramatically more. The work has not gotten easier or shorter for them. The price of the alternative has gone up because there isn't really an alternative for what they do.

The middle is where the damage is concentrated. Freelancers who were charging $80 to $150 an hour for work that, in retrospect, was somewhere between commodity and craft. Producers more than authors. Executors more than strategists.

If you are in the middle, the next eighteen months are about figuring out which way to go. You cannot stay where you are. Staying still in this market is the same as moving down.

The question that determines your position

Every freelance service breaks into roughly three components.

The decision about what to make. The strategic part. The judgment.
The making of it. The craft execution.
The packaging and delivery. The format, the file, the polish.

Until recently, all three of these were valuable enough that you could build a freelance practice on any of them. Pure executors could thrive. Pure strategists could thrive. People who lived in the middle could thrive.

AI changed the price of category two and three. The making and the packaging are now substantially automated for most outputs. Not perfectly, but well enough that a client with $5,000 to spend can get something that looks like what $25,000 would have bought five years ago.

What AI did not change is category one. The decision about what to make is still a human judgment problem. Sometimes it is a problem of taste. Sometimes a problem of strategy. Sometimes a problem of understanding the specific client well enough to know what will actually work for their specific situation. The thing that all of these have in common is that AI is not particularly good at any of them yet, and the gap between AI and a thoughtful human in these areas is not closing as fast as the gap in the executional areas is.

The question every freelancer needs to ask themselves, and answer honestly, is which of those categories you currently charge for. If you are charging mostly for executional craft, you are exposed. If you are charging mostly for judgment, taste, and strategic decisions, you are likely fine, possibly more in demand than ever.

How to move up

Moving up the value stack is harder than it sounds, but it is also more achievable than most freelancers assume. A few specific shifts that I have seen work.

Stop selling deliverables. Start selling decisions.

If you sell a logo, your competition is everyone who can produce a logo, which now includes AI. If you sell a brand identity decision (the visual direction, the positioning, the system that will work for this specific business as it grows), your competition is much narrower, and harder for a tool to replicate.

The deliverable becomes a byproduct. The thing you are actually billing for is the judgment that produced it.

In practical terms, this often means the conversation with the client shifts. Instead of starting with what they want made, you start with what they are actually trying to achieve and what will work in their context. You become more useful earlier in the project. You charge for that usefulness.

Specialize aggressively

Generalists are losing ground fastest. Specialists are holding or gaining. Not because specialization is inherently better, but because clients can no longer tell from a portfolio whether a generalist is actually competent. The signal that used to come from the work itself ("look how well executed this is") is muddied because everyone's work looks well executed now.

Specialization restores the signal. A designer who only does packaging for small batch food brands is legible to a small batch food brand in a way that a generalist designer is not. The specialization is the credibility, not the craft.

The specialization can be vertical (industry), horizontal (a specific deliverable type), or methodological (a specific approach). Any of them work. None work if you are trying to be everything to everyone.

Get closer to the business outcome

This is the hardest move, and the most valuable. If you can connect your work to a business outcome the client cares about ("this rebrand made it possible for them to charge 20% more"), you have moved out of the deliverable market entirely. You are now selling business impact, which is priced not by what your work costs to produce but by what it is worth.

I am not pretending this is easy. Most freelancers cannot honestly draw that line yet. The way to start is to pick a metric the client cares about and start tracking your impact on it, even if it is informal. Over a few projects, you start to have real numbers. Over more projects, those numbers become your sales argument. Over enough projects, you stop being a creative vendor and start being something closer to a consultant who happens to make things.

Charge by the decision, not the deliverable

A logo costs X. A brand identity decision costs ten X. Same designer, same skills, different pricing structure. The brand identity decision involves the same logo plus strategic thinking, positioning work, and the system around it. The price reflects the broader value being delivered, not the time spent.

If you are still pricing by the asset, you are competing on a metric that AI is dragging toward zero. If you are pricing by the decision, you are competing on a metric AI cannot easily touch.

What about going down

I want to say a word about the other direction too, because it is a real option for some people. There is still volume work to be had. It just pays poorly, and the trend line is bad.

If you go down the stack, your job is to be radically more efficient than your competition. That probably means heavy AI workflow integration, strict scope discipline, no customization, and very fast turnaround. Productized services on tight templates. Subscription based deliverables. Lots of small clients rather than a few big ones.

This can work. I know a few freelancers running productized services in the $3,000 to $8,000 a month range by being the most operationally efficient option in their niche. But the work is mostly about systems, not craft. If craft is what drew you to freelancing, this is not the direction that will keep you happy.

The honest assessment

Here is what I tell freelancers who come to me concerned about the market right now.

The work you did three years ago is unlikely to be enough to support you three years from now. That is not a comment on your skill. It is a comment on the price floor of the kind of work you were doing. The price floor moved.

The question is whether you adjust now, while you still have the time and the client base to make the transition, or whether you wait until the math forces it and the transition is harder. I have seen freelancers in their late forties and early fifties move up the stack successfully. It is not impossible. It is also not free. The earlier you start, the smoother it goes.

The bakery that left me last year? I miss them. I also recognized in retrospect that I had been a generalist designer to them, working at executional rates. The lesson was not that AI ate my business. The lesson was that the work I had been doing for them was already at the low end of the value I could offer, and I had not adjusted in time.

I am not making that mistake twice.