A Landing Page Redesign That Looked Great and Tanked Conversion

The old landing page was, by any design critique standard, ugly. Cluttered, inconsistent spacing, a hero section that didn't match the brand refresh from two years earlier. The new one was clean, modern, and every stakeholder who reviewed it in the design file loved it. We shipped it. Conversion dropped 23% in the first week and didn't recover.

What the design review missed

The old, ugly page had a pricing table above the fold. The new, clean page moved pricing below a hero section and a testimonial carousel, because that's the standard pattern for a 'modern' landing page and it tested well in isolation during the design review. Except this product's visitors were almost entirely bottom-of-funnel traffic from branded search — people who already knew what the product was and just wanted the price. We'd redesigned for a visitor who doesn't actually make up most of the traffic.

A design review tells you whether stakeholders like how a page looks. It tells you nothing about whether it matches what your actual traffic came there to do.

How we found the cause, and fixed it

• Segmented the conversion drop by traffic source first — it was almost entirely concentrated in branded and direct traffic, not top-of-funnel paid traffic, which immediately pointed at bottom-of-funnel visitor intent.
• Watched session recordings of the new page and saw repeated scroll-down-and-back-up behavior right where the pricing table used to be, a strong signal people were searching for something they expected to already see.
• Shipped a fast-follow that moved pricing back above the fold while keeping the new visual design, recovering most of the lost conversion within a week.
• Added a standing rule to future redesigns: map the above-the-fold content to actual traffic intent by segment, not to a generic 'best practice' page template.

The redesign wasn't wrong to make the page look better — it was wrong to apply a generic content hierarchy without checking it against what this specific audience actually came to the page to do. A page can look objectively better in every design critique and still be measurably worse at the one thing it exists to do.