A 'website refresh' can drift for six months. The Launch Sprint method exists because most teams don't actually need six months — they need disciplined scope and weekly artifacts.
The shape of the sprint
- Week 1 — Discovery, narrative, and architecture
- Week 2 — Visual system and key page design
- Week 3 — Build, copy pass, and motion polish
- Week 4 — QA, performance audit, and launch
Week 1 produces three artifacts
A one-page narrative, a sitemap with section-level intent, and a measurable launch goal. If we can't write the goal in one sentence, we don't continue.
The fixed deadline is the feature. Everything else negotiates around it.
What we cut to make it fit
Sprints succeed by being honest about what is out. Custom illustration libraries, multi-language support, blog migrations — these become follow-on engagements, not sprint risk.
Where most sprints break
- Stakeholder reviews scheduled later than week 2
- Copy treated as a 'final week' task
- Performance audit deferred until after launch
Solve all three in the kickoff doc and the rest is mostly execution.
Noah leads product strategy and design at Growrix OS. Previously shipped two SaaS startups from zero to Series B.
Discussion (2)
Be kind. Be specific.- PSPriya S.Apr 7, 2026
We ran a version of this and the weekly artifact discipline was the unlock for our team.
- BRBen R.Apr 9, 2026
How do you handle clients who can't move at this pace? Do you decline?
