Where Your Website Leaks Money: The Case for a Conversion Audit
Most underperforming websites aren’t broken. The pages load, the buttons click, the orders go through. They just quietly lose money at a dozen small points that nobody is looking at — and because nothing is obviously wrong, nobody goes looking.
A conversion audit is how you go looking on purpose. Instead of guessing why a site underperforms, it works backwards from real behaviour — analytics, heatmaps, session recordings and a structured review of every step in the journey — to pinpoint exactly where visitors hesitate, get confused, or give up. The output isn’t a vague list of opinions. It’s a prioritised set of fixes, ranked by how much revenue each one is likely to recover.
Speed is the leak everyone ignores
Start with the least glamorous and most expensive issue: page speed. The relationship between load time and money is brutally direct. Every 100 milliseconds of additional load time costs roughly 1% in conversions, and a one-second delay reduces conversions by about 7%. The probability that a visitor bounces rises 32% as a page goes from one second to three.
This isn’t only a revenue problem; it’s a ranking problem too. Google lowered the “Good” threshold for Largest Contentful Paint from 2.5 seconds to 2.0 in its March 2026 update, which means a lot of sites that passed Core Web Vitals last year now quietly fail. And most do fail: only around 42% of mobile sites pass all three Core Web Vitals, against 63% on desktop — with mobile carrying the majority of traffic. Sites that do pass see roughly 24% lower bounce rates. Speed, in other words, is a conversion feature wearing an engineering costume.
The scale becomes obvious with a single worked example: for a store turning over £10 million a year, a 500-millisecond improvement in load time is worth in the region of £500,000 in recovered revenue. That is a number that justifies an audit on its own.
What an audit actually looks at
Speed is the opening move, not the whole game. A proper audit examines the entire journey from landing to conversion:
- Analytics review — where exactly people drop out of the funnel, and which pages and devices lose the most.
- Heatmaps — what people click, how far they scroll, and what they ignore entirely.
- Session recordings — watching real users get stuck, hunt for information, or rage-click a broken element.
- Heuristic review — a structured expert pass over clarity, trust, navigation and the strength of the value proposition.
- Funnel mapping — tracing the intended path against the one users actually take.
Together these turn a vague sense that “the site could do better” into a specific, evidenced map of where money is leaking and why.
Opinion is the enemy
The reason audits work is that they replace opinion with evidence. Most websites are shaped by internal preferences — a stakeholder liked a hero image, a designer wanted a carousel, someone insisted the form capture more fields. None of those decisions were tested against how customers behave, and any one of them could be costing conversions. An audit doesn’t care who was right in the meeting; it cares what the data shows. That is a far cheaper way to settle a debate than shipping a redesign and hoping.
Where to start
The point of an audit is not a 60-page report that gathers dust. It’s a short, ranked list of changes ordered by impact and effort — the handful of fixes that will move the numbers fastest, plus the longer-term work worth planning for. You fix the cheapest high-impact items first, measure the lift, and reinvest it. That is how conversion improves month after month rather than in one risky leap.
If your traffic is healthy but your conversion rate isn’t, the problem is almost never that you need more visitors. It’s that the visitors you already pay for are leaking out through gaps you can’t see. A conversion audit is how you find them — and it usually pays for itself many times over.