Process · 5 min read
A defence of spec work. Client briefs are invented too, and the skill is building constraints honest enough to be worth solving.
Every project in my portfolio is self-initiated. I’d rather say that in the first paragraph than have you find it in the fine print.
The usual response to spec work is a polite “yes, but it’s not real.” No client, no budget, nobody asking for a bigger logo. It’s easy to be brilliant when nobody pushes back.
I understand the suspicion. I just think it misreads what a brief is.
Somebody wrote the brief
I’ve read a lot of client briefs, and written plenty. Most of them have one thing in common: someone made them up.
Not dishonestly. A marketing lead sat down on a Tuesday and decided the audience was “urban professionals 25-40”, that the objective was “awareness and consideration” and that the tone should be “premium but approachable”. Those are choices, often good guesses, written down with confidence and then treated as physics by everyone downstream.
In architecture the brief is often called the programme: a list of spaces, areas and relationships the building has to hold. Architects learn quickly to treat it as a starting hypothesis. Half the job is pushing on it until you find what the client actually needs, which often differs from what’s on page two.
So client briefs and spec briefs are both fiction. What matters is whether the fiction is any good.
What makes a constraint honest
A lazy spec brief has no constraints at all: “imagine a campaign for Nike.” That’s a daydream with a logo.
A useful one is harder to write than most client briefs, because nobody else imposes the limits. You set them yourself and then don’t cheat. For every project here I set:
- A defined audience, specific enough to be wrong about
- A defined measure of success, written before the work starts
- Real production limits: cast size, locations, tool cost, days
- One uncomfortable constraint, the kind I’d have argued against if a client had given it to me
The Gardener had to happen in one flat, on one morning, with three visitors. Grip had to sell tyres while barely mentioning them. Didn’t Go couldn’t show a single feature of the app it was selling. DEW had to be something a small brand could keep consistent on its own. Each of those limits made the work harder, and better.
What spec work can’t prove
It can’t prove I’ll survive a stakeholder who changes their mind in round four, and it can’t show results, because there aren’t any.
What it can show is the thinking before anyone else’s opinion got layered on top, which is the hardest part to see in a client portfolio. Every Decisions section in these case studies is a trade-off I made and can defend. A client project needs the same judgment; here it’s easier to see because nobody else’s fingerprints are on it.
The actual skill
Landing a real client comes down to luck, timing and who you know. The skill is turning an open question into a problem worth solving, with constraints honest enough that a good solution has to earn its place.
Every brief you’ve ever received was someone’s attempt at that, and some were better than others.
Some of them, if we’re honest, were worse than mine.