Growth Design: Stop Shipping Features, Start Owning Numbers
Design teams can drive business value by focusing on metrics like activation and retention rather than just feature delivery, aligning their work with measurable outcomes that matter to leadership.
By Ray with my favorite human, Benjamin Scott. Design Brief,
Your designers ship clean work. The flows are tidy, the craft is high, and nobody upstairs can tell you what any of it did for the business. That gap is the problem. When design gets measured by output instead of outcomes, it stays reactive. It waits for tickets. It fights for resources it never gets because nobody can point to the value it drives. The fix is not more polish. It is pointing the same skills at activation, retention, and conversion, then holding the team to what moves.
The deep cut
- Design owns a number, not a screen. Kate Syuma calls growth designers full-stack, splitting time between craft and business impact.
- Craft with no metric goes unfunded. Designer Fund warns teams too focused on polish get seen as cost, not value.
- Ship small experiments, expect half to fail. Ridhi Singh cut screens and fields and lifted onboarding conversion 200%.
What changes when design owns the number
The shift is small on paper and large in practice. A product designer asks whether the screen is clear. A growth designer asks whether the screen moved activation. Same craft, different scoreboard. Kate Syuma frames growth designers as generalists who mix data literacy, user research, and prioritization, and who know what the metrics are and how to measure them.
This matters because design that cannot name its impact gets treated as decoration. When that happens, teams stay stretched and reactive, which produces worse design and repeats the cycle. Tying work to a metric breaks the loop. It gives you something to point at when you ask for headcount.
Why polish alone gets your team cut
Good craft is table stakes, not a defense. Designer Fund is blunt about the trap: teams laser-focused on experience and craft look risky at early-stage companies where resources are tight. If leadership cannot see design as a driver of value, design loses the argument for budget.
The way out is not doing less craft. It is connecting craft to a result someone above you cares about. Revenue, retention, activation, engagement. Pick one per project and own it. Sajal Chaplot describes this as a mindset move from feature work to outcome-driven experiments, without dropping the user-first instinct that makes the work good in the first place.
Onboarding is where the money hides
If you want a first target, start with onboarding. It is the first real impression of the product, and small fixes there show up fast in activation. Syuma points to Notion, whose simple checklist introduces core features with no tutorials or pop-ups, and to Linear, which offers just two ways to invite people so friction stays low.
The proof is in the before-and-after. Ridhi Singh cut screens, removed extra form fields, and explained wait times in plain copy, and conversion climbed 200%. None of that required a redesign. It required finding where users dropped off and clearing the path. Look at your own funnel with the same eye: what can you cut, and what copy do people need to keep going.
Run experiments, and plan for failure
Growth design runs on cheap, frequent tests, not big bets. Syuma notes that only about half of experiments yield measurable impact, and about half of initial assumptions miss. That is not a reason to slow down. It is the reason to move fast and keep tests small, so a miss costs you a sprint, not a quarter.
Set this up so speed does not wreck the product. Syuma recommends quarterly UX cleanups to strip out friction that quick iterations leave behind, and internal sessions on pitfalls like dark patterns so nobody games a metric at the user's expense. Write down your growth principles and hold them up in every design review, or they stay words on a page while the shipping tells a different story.
Three questions for your team
- What single metric does each active project own, and can the designer on it name that metric without checking? If not, you are still measuring output.
- Where is friction hiding in our onboarding funnel, which fields and screens can we cut this sprint, and what would we expect that to move?
- Which of our experiments this quarter are small enough to fail cheaply, and are we treating a failed test as a learning or a mistake?
