Why Your Design Team Looks Unprofitable (and Usually Isn't)

I've sat in the review where someone points at the design team's margin line and asks, gently, whether we have a design problem. It's a fair question to ask of the number on the screen. It's almost always the wrong question to ask of the team.

Design agency profitability has a reporting problem before it has a performance problem. The work designers do is harder to bound than the work developers do — a feature is "done" when it passes its tests, but a homepage is "done" when someone decides they like it — and the structures most agencies wrap around design don't account for that difference. So design absorbs cost that nobody priced, gives away work nobody scoped, and then reads as the unprofitable team on a report that was never set up to measure it fairly.

This is distinct from senior people doing junior work, which is a role-mismatch cost that hits any team. This is about the things that make design specifically look like it loses money: the open-ended nature of creative sign-off, and the way design cost leaks across the boundaries of a project. It's for the CFO or owner looking at a red design line and deciding whether to act on it. Usually you should — just not on the team.

Why design reads as the loss-making team

Four structural causes, in roughly the order they cost money.

Unbounded revision rounds

Development has a natural stopping point: the thing works or it doesn't. Design's stopping point is a judgement, and judgement can always change its mind. A landing page comes back for "one more pass" on the hero, then the colour, then the spacing, then the hero again because the new colour changed how it reads. Each pass is real hours. None of them were in the estimate, which assumed two rounds and got five.

Because the fee was fixed at two rounds, rounds three through five are delivered at zero marginal revenue and full marginal cost. The design team's hours climb; the design team's revenue doesn't. The margin line goes red. The designers did nothing wrong — they did exactly what was asked, more times than anyone priced for.

Subjective sign-off with no defined "done"

Closely related, but worth separating: even when revisions are notionally capped, design often has no defined approval gate. There's no moment where the client signs a specific artefact and the phase formally closes. So work that should have ended in week three trickles on into week six, picking up small unbilled changes, because nobody declared it finished. A developer's work closes when it merges. A designer's work closes when someone says so — and if no one is empowered to say so, it never closes on schedule.

Discovery and exploration given away

The most valuable thing a good design team does — the early exploration, the three directions before the one, the workshop that reframes the brief — is also the hardest to scope and the easiest to give away. It feels like "getting started" rather than billable work, so it gets folded into the project for free, or run as an unpriced pitch. The hours are large and senior, and they land entirely on the design team's cost line with no revenue beside them. Discovery isn't overhead. It's the most expensive thing design produces, and treating it as a free warm-up is most of why the line reads red.

Design hours bleeding into the build

The quietest cause. When a project moves into development, design doesn't stop — there are tweaks during implementation, responsive states that weren't in the comps, a component that needs rethinking once it's real. Those hours are design hours, but they're logged against the development phase, so they're costed against the build's margin and counted as the build's hours. The result distorts two lines at once: development looks slightly heavier than its design content implies, and design looks like it under-delivered relative to its hours because some of its real output got attributed elsewhere. On a report by team, design is paying for work the build is taking credit for.

How this distorts the apparent margin

Stack those four up and the mechanism is clear: design's cost line carries hours that were never priced (extra revisions, unscoped discovery) and is missing revenue for hours that got attributed to other phases (the bleed into build). Cost too high, revenue too low, both for structural reasons. The margin that falls out is not a measure of the design team's efficiency. It's a measure of how badly the agency scopes and attributes design work.

The danger is acting on the artefact. A leadership team that reads the red line literally will under-resource design, push back on rates, or treat the team as a cost to contain — which makes the craft worse without touching any of the four causes. The leak doesn't move. The team gets demoralised. And the next report looks the same, because nothing structural changed.

The fixes that don't crush craft

The point of fixing this is not to make designers work to a stopwatch. It's to make the work visible and bounded so the margin reads true. Four moves, matched to the four causes.

• Cost discovery as its own line. Scope discovery and exploration as a distinct, paid phase with its own deliverable and its own fee — a discovery sprint, a concept phase, however you frame it. It stops the most expensive design output being given away as a warm-up, and it gives the client something concrete they bought. The craft is unchanged; it's now priced.
• Cap or tier revisions. Two rounds included, further rounds at a defined rate, written into the proposal. This isn't about refusing the third round — it's about the third round being a known, costed thing rather than a silent cost. Most clients are entirely reasonable once "more rounds cost more" is stated up front rather than discovered at the end.
• Define a hard approval gate. A named decision-maker, a specific artefact, a formal sign-off that closes the phase. "Done" becomes a moment, not a vibe. The work that used to trickle on into week six now closes in week three, because someone is empowered to close it.
• Cost design at the real rate. Cost each designer's hours at their fully-loaded real cost, on whatever phase they actually worked — including the hours that bleed into the build. Attribute design hours to design, wherever they're logged. This is the move that fixes the report rather than the proposal, and it's the one that usually reveals the red line was never as red as it looked.

None of these ask the designers to be faster or more pliable. They ask the agency to scope, bound and attribute the work honestly. Craft is protected by definition — the better the discovery is priced and the cleaner the approval gate, the more room good design actually has, because it isn't being eroded by unbilled rounds.

A worked example: the brand-and-site project

Anonymised and scaled to round numbers.

A combined brand refresh and website, sold for £60k. On the team margin report it came in looking like the design portion lost money. Here's the report as first presented, then the same project once the structure is made visible.

As reported. Design was nominally allocated £24k of the £60k fee. Design hours, costed at real rate, came to £27k. Apparent design margin: (24 − 27) / 24 = −12.5%. Red line. "Do we have a design problem?"

Now make the four causes visible:

Re-cut the same numbers honestly. Attribute the £5k of bleed back to where it belongs (it leaves design's cost line; it was the build's content). Treat the £6k discovery as the unpriced phase it was, and the ~£4k of extra revisions as scope that should have been a tiered charge. Suddenly the picture is not a design team that overspent — it's an agency that gave away ~£10k of design work it never put a price on, and mis-attributed another £5k.

• Design hours that genuinely belong to design: £27k − £5k bleed = £22k
• Design's fair share of the fee, with discovery costed as its own £6k line and two extra revision rounds tiered at ~£4k: design should have been a £34k line, not £24k

Recomputed against a fair £34k of revenue for £22k of properly-attributed cost, design margin is (34 − 22) / 34 = 35.3% — one of the healthier lines on the project, not the loss-maker. The £60k didn't change. What changed is that the work design actually did got priced and attributed, instead of being absorbed silently and then blamed.

Before you act on the red line

The instinct when a team reads unprofitable is to manage the team. With design it's worth resisting that for one report cycle, because the line is so often an artefact of how design is scoped and costed rather than how the team performs. Different agencies run different shapes of work, and design behaves differently across the three project types — discovery-heavy builds, light support tweaks, open-ended retainers — so the same design team can look profitable on one and ruinous on another for purely structural reasons.

Ask four questions before you ask about the team. Are revisions bounded and tiered? Is discovery a priced phase or a free warm-up? Is there a real approval gate? Are design hours costed to design wherever they're logged? If the answer to any of those is no, you're looking at a scoping and attribution problem wearing the costume of a performance problem.

To see design costed at real rate and attributed to the right phase wherever the hours land — rather than reconstructed by hand each month — that's what the 15-minute demo shows on your own Jira data. The short version: the design team usually isn't losing money. The agency is giving design work away and then reading the gift back as a loss. Price the work, bound the rounds, attribute the hours honestly — and the red line tends to turn the colour it should have been all along.