What actually happened to design roles in the last two years
By Allan Leone on
Adoption of AI design tools is near universal, but the headcount collapse people predicted did not arrive. Something else happened instead, and it shows up clearly if you look at which briefs are getting paid for.
88% of businesses now use AI design tools. Only 18% report that it reduced their need for designers. Those two numbers sit oddly together, and most of the argument online comes from picking one and ignoring the other.
The pattern I see in hiring conversations and in the briefs that come through the studio is not fewer designers. It is a split in what people are being paid to do.
The work that got cheaper
One tier of design work was always a rendering job. The brief arrived fully specified and the task was to produce the artefact. Resize this campaign into nine placements. Same layout, different market. Turn this deck outline into slides.
That work priced on time, and the time collapsed. A studio that used to quote three days for a campaign resize now quotes half a day, because that is what it takes. Clients found out at roughly the same speed everyone else did.
Nobody I know enjoyed that work. It was also, for a long time, the entry point into the profession and the way juniors built judgement by volume. That part has no obvious replacement yet, and I do not think the industry has been honest about it.
The work that got more expensive
The other tier starts before the brief exists. Someone has to decide what should be built, argue for cutting the feature the stakeholder asked for, and notice that the real problem is in onboarding rather than the dashboard everyone is staring at.
That work got more valuable, not less, and the reason is mechanical. Generation removed the cost of building the wrong thing. When a wrong direction took six weeks to surface, the market self-corrected slowly. When it takes two days, teams can now produce three wrong directions per sprint, and the constraint moves to whoever decides which direction is right.
How the split shows up in practice
- Briefs that arrive as a problem statement, not a feature list, are up. Briefs that arrive as a list of screens are down.
- Discovery is being paid for separately more often, because clients have learned that the build is no longer the expensive part.
- Rates for execution-only work are under pressure. Rates for the work that ends in a written recommendation are not.
None of this is a prediction. It is what the last eighteen months of proposals in one small studio looked like, and I would be careful generalising too far from that. But the direction matches what the survey numbers imply.
Where to start
- Look at your last ten pieces of work and mark each one as either "the brief was already decided" or "we helped decide the brief". The ratio is your exposure.
- In the next project, ask for the meeting before the brief. Even one hour of framing changes what you are being judged on later.
- Write down the recommendation you made and what happened. Not the screens. That record is what the second tier of work is bought on.
- If you manage juniors, replace the volume work they used to learn from with review work. Have them critique generated output against a brief. It builds the same judgement without the busywork.
Tags: ai, careers, design