Lighthouse 100s, sub-second loads, beautiful animation. None of it matters if the site is the wrong asset for the business question. Speed is necessary, not sufficient.
The fastest agencies in 2026 are competing on production speed. Cut the timeline from 18 weeks to 6, ship perfect Lighthouse scores, win the pitch. It's a real efficiency gain. It's also where most agency thinking now stops.
Here's the trouble with that. The slowest part of a website project was never the production. It was the part nobody scoped properly: figuring out what the website was supposed to do for the business in the first place. Compressing the production half from 12 weeks to 4 doesn't help you if the strategy half is still scoped at "one paragraph, Phase 0" while silently wrong.
We've been in enough post-launch retrospectives to know what this looks like. The site ships on time. It looks beautiful. Lighthouse is green. Six months later the metric that mattered hasn't moved. Sometimes it's worse than before, because the new homepage broke an SEO pattern nobody mapped.
Our position is unfashionable: discovery is the deliverable. Production is the bit that follows from doing discovery properly. The site is downstream of the strategy, every time.
That's why we put a real two-week sprint at the front of every engagement, with the operator (not just the marketing lead) in the room. We're trying to find the question the website is supposed to answer. Once we have that, the rest of the build is comparatively easy. Without it, no amount of production speed saves you.
Fast is the floor, not the ceiling. The ceiling is whether the site does its job.