Improving website conversion through testing

Traffic was improving and enquiries were not. The funnel had been designed for a company that no longer existed: the messaging described services broadly, the forms asked for more than a first conversation needs, and nothing on the page told a visitor what happened after they clicked.

Ran the audit properly first using Hotjar and Clarity to watch what people actually did, not what the analytics suggested they did. Then worked the loop: one hypothesis, one test, one shipped change, repeat. Shortened forms, rewrote the value proposition around the problems buyers described in sales calls, and rebuilt the page structure so the proof arrived before the ask.

Conversion rate improved roughly six-fold end to end. No single test did it. The change that mattered structurally was owning implementation as well as analysis, so a winning variant reached production in days rather than dying in a backlog.