Creativity pays, but not the way software promises
Topic:
Business
Year:
5 May 2026
he evidence base here is unusually good. The IPA’s The Long and the Short of It, by Les Binet and Peter Field, is built on a databank of effectiveness case studies and is explicit that the timescale of effect matters: activity aimed at immediate response behaves differently from work that builds a brand, and the balance between the two is what determines long-term growth. Creative work, in that framing, is not decoration on a media plan. It is the mechanism by which a business becomes easier to choose later.
But the same body of work carries a warning. Field’s later analysis, published by the IPA as The Crisis in Creative Effectiveness, found that creatively awarded campaigns had become no more effective than non-awarded ones, the weakest position in 24 years of that data. Creativity still pays. Creativity performed for other creatives has stopped paying. Those are different activities that happen to look alike.
Two different curves
Software sells a seductive shape: build once, sell infinitely, watch margin approach the sky. Creative work follows a different curve entirely, and pretending otherwise is how good studios go broke chasing a model that was never theirs. Creativity does pay, reliably and well, but it compounds through reputation and judgment, not through the near-zero marginal cost of code.
Why generative tooling is not the lever
If the constraint were volume, generative models would already have solved this. They have made volume nearly free. What they cannot do is supply the scarce input: a true insight about a specific business, in a specific category, told to a specific person who has a reason to be sceptical. A model can produce a thousand competent variations of what has already worked for someone else. Averaging the past is a strange way to become distinctive.
The pairing that actually moves numbers is less glamorous and harder to buy: unsuspecting craft, linked by true insight. Craft that nobody was told to admire, a kerning pass, a loading state, a sentence rewritten until it stops hedging, joined to a claim that is genuinely true about the business. The insight makes the work worth noticing. The craft makes it worth trusting. Remove either and you are left with volume, which the market already has in surplus.
Questions worth putting to your own work
What do we know about this business that a competitor could not credibly claim?
Which part of this is aimed at next quarter, and which part compounds?
If the logo were removed, would anyone know whose work this is?
Where did we accept a generated first draft as a decision rather than a sketch?
Where the leverage really is
The software fantasy applied to studios usually looks like productized templates, courses, or tools meant to detach income from hours. Some of that works at the margins. But the core asset of a creative practice is not a product; it is a reputation for judgment, and reputation does not duplicate. It is rebuilt, client by client, project by project, which is precisely why it is defensible.
Real leverage in creative work comes from three places. Pricing on value rather than time, so a decision that transforms a business is not billed like a commodity. Building systems and reusable thinking that make each project faster than the last. And being known for something specific enough that the right clients arrive already convinced, which collapses the cost of winning the work.
That is a wonderful business, durable, high-margin, and hard to copy, but it is a craft business, not a software one. The studios that thrive stop apologizing for selling judgment by the engagement and start compounding the thing software can never mint: a name that means something. Creativity pays. It just pays in trust, and trust has no free trial.
T
Selected work
Other articles
S.1
Have a project like this in mind?
Tell us what you’re building. We reply to every serious inquiry within two business days.