Two weeks ago we looked at our own homepage and genuinely cringed. Not because it was broken — it loaded fast, it had all the right sections — but because it no longer represented who we are. It was the homepage we built when we were trying to prove ourselves. This is the story of how we scrapped it and shipped something better in 48 hours.

We started with a Hotjar session review and a quick heatmap. The data told an uncomfortable story: 68% of visitors scrolled less than 30% of the page. The hero was a paragraph-heavy explanation of our process that nobody read. The CTA was buried below three sections of social proof that, ironically, visitors were skipping to reach it.

We wrote one constraint on the whiteboard: the visitor should understand what we do and how to start within 10 seconds. Everything else was negotiable.

We spent a surprisingly long time on copy before touching Figma. The old headline was: “We build custom WordPress websites that convert.” Fine. Forgettable. The new one: “Your WordPress site, rebuilt in weeks — not months.” Same information, completely different energy. It answers the question every client actually has: how long will this take?

We cut the about section entirely and moved three testimonials to the hero. Nobody scrolls to the bottom to decide if they trust you — you have to earn it at the top.

We built in Impreza, which meant no context switching between Figma and code — everything was directly in the builder. The biggest time saving was reusing the section library we’d built for client projects: the testimonial layout, the pricing cards, the contact form were all adapted from existing components.

The most painful decision: killing the full-bleed hero video we’d spent two days shooting last spring. It looked great. It added 1.2MB to the initial load. It went in the trash. A gradient block replaced it. The page is now faster and, honestly, cleaner.

We pushed live at 11pm on day two, deliberately avoiding peak traffic hours. Within 24 hours: average scroll depth jumped from 30% to 54%. Contact form submissions in the first week were up 31% over the prior seven-day average. We’ll wait a full month before drawing conclusions, but the early signal is clear.

The most important lesson wasn’t about design — it was about speed. A decision made in 48 hours beats a perfect solution shipped in six weeks because the imperfect one actually exists.