Last year, our four-person studio shipped eight full brand and web identities for Series A startups. None of the projects ran late. None went over scope by more than fifteen percent. We thought we’d collapse halfway through. We didn’t. Here’s what made it possible — and what we’d change.

The non-negotiable: scope locks at week two

Every project has a discovery phase. Ours was three working sessions in week one, followed by a written scope document signed by the founder in week two. After that, anything new is a change request that adds time and cost. No exceptions, no goodwill creep.

We held the line on this for six of the eight projects. The two we softened on — both founders we adored — ran the longest and burned us out the most. The lesson held: scope discipline is kindness, not coldness.

The unexpected bottleneck: feedback loops

We expected design execution to be the constraint. It wasn’t. The constraint was feedback turnaround. A two-day delay in client review compounded across six rounds of work on each project, becoming weeks. We started scheduling weekly thirty-minute review calls instead of waiting for async comments. Cycle time dropped 40%.

The thing that broke: archive discipline

By month nine we couldn’t find anything. Project files lived across Figma, Notion, Dropbox, Slack threads, and individual desktops. We spent half a day every week looking for assets we’d shipped six weeks earlier. We rebuilt our internal asset architecture in December. The next year of project starts cost half as much in onboarding overhead.

What we’d do differently

Hire one more person. We treated four as a feature; it was actually a bottleneck. The fifth role — a junior brand designer who could own the long-tail asset production — would have given us another four projects in the same year without the burnout.