Conversion Funnel: How to Fix the 5 Biggest Leaks

The average eCommerce funnel loses 97.5 percent of its visitors before conversion. That is not a disaster. It is normal. The question is not whether you have losses, but where the biggest leaks are and which one promises the highest ROI when fixed.

Funnel analyses show the same pattern again and again: the biggest leaks are not where the owners suspect. They sit in the transitions: between product page and basket, between basket and checkout, between checkout and payment confirmation. Fixing those transitions systematically is where the revenue sits, and it is worth more than any single page redesign.

Leak 1: The product page does not load

The most common cause of drop-offs on the product page is not the layout, not the price, not the images. It is load time. Every second of load time over three seconds reduces the conversion rate by 7 percent.

For a shop with 50,000 visitors a month and an average order value of 80 euros, that is 28,000 euros of lost revenue a month, for a single second.

Halving the product page load time is usually a small job with an outsized effect: image compression (WebP instead of JPEG, lazy loading), inline critical CSS and deferred JavaScript. Measure conversion before and after the change on your own traffic; that is the only number that counts here.

Worked example: an electronics shop with 80,000 sessions a month, a 2 percent conversion rate and a 400 euro basket cuts its product page load time from 4.2 to 2.1 seconds.

Conversion rises by 18 percent, from 1,600 orders to 1,888, and the extra 288 orders are worth about 115,000 euros a month. The work behind it: 12 hours of development time.

So before you work on layout or copy, measure load time. Use Google PageSpeed Insights, GTmetrix or WebPageTest. If your product page takes longer than three seconds to load, that is your leak number one, however good the rest looks.

Leak 2: The basket is a dead end

68 percent of customers who put an item in the basket do not buy. The reasons are well known: unexpected shipping costs, a complicated checkout, forced account creation, too many form fields. But the solution is not to fix every problem at once. The solution is to plug the biggest leak first.

A checkout analysis usually puts the biggest leak between the basket and the shipping information. A frequent cause is a long form with optional fields marked as required.

Cut the form to the fields the carrier and the payment provider genuinely need, make guest checkout prominent and show shipping costs in the basket. Those three fixes are what raises checkout completion most reliably, and they are cheap to test.

Worked example: a fashion shop with 120,000 sessions a month finds that 34 percent of its customers drop off between the basket and the shipping information. The form has 12 fields, 4 of them optional but marked as required.

Cutting it to 5 fields, making guest checkout prominent and showing shipping costs in the basket lifts checkout completion from 32 to 51 percent. On 36,000 baskets a month and a 50 euro order value that is about 340,000 euros of additional revenue.

Measure your checkout funnel step by step. Use Google Analytics 4 (event tracking for each checkout step) or Hotjar (funnel analysis). Identify the step with the highest drop-off. Fix only that step. Measure again. Repeat.

Leak 3: The mobile version is an afterthought

70 percent of eCommerce traffic comes from mobile devices. Yet the mobile conversion rate is on average 30 to 50 percent below desktop. That means most of your traffic does not convert because the mobile version does not work.

Where the mobile conversion rate is a fraction of the desktop rate, the analysis tends to find the same three problems: filter navigation that a thumb cannot use (touch targets too small), product images that cannot be zoomed, and a checkout button hidden below the fold.

The fixes are standard: touch targets of at least 48×48 pixels, pinch-to-zoom on product images, and a checkout button that stays sticky at the bottom of the screen.

Worked example: a furniture shop with 60,000 mobile sessions a month converts at 0.8 percent on mobile against 2.1 percent on desktop. Touch targets, pinch-to-zoom and a sticky checkout button lift the mobile rate to 1.4 percent: 360 extra orders a month, worth about 168,000 euros at this shop’s order value.

Test your shop on a real smartphone, not in a browser emulator. Try to buy a product. If at any point you have to zoom, tap twice or wait, you have a leak.

Leak 4: Trust is missing

Customers do not buy from shops they do not trust. Trust does not come from a big logo or a nice colour. It comes from specific signals in the right places: security seals at checkout, reviews on the product page, return conditions before purchase, legal notice and contact details that are easy to find.

A frequent finding: the product pages carry no reviews, the checkout shows no security seal, and the return conditions are only readable after the purchase.

The three fixes are cheap: reviews with photos on the product page, the seal next to the payment button, and the return conditions in the basket. Test them one at a time; a shop that ships all three at once never learns which one carried the result.

Worked example: a food supplements shop with 40,000 sessions a month converts at 1.2 percent against an industry average of 2.5 percent. Reviews with photos go on the product page, the seal moves next to the payment button and the return conditions move into the basket.

Conversion rises to 2.3 percent, which is 440 extra orders a month and, at a 320 euro basket, about 140,000 euros of additional revenue.

Trust signals are not a nice-to-have. In many shops they are a bigger conversion lever than any design change. Check every page of your funnel: where might a customer start to doubt? Which signal could resolve that doubt?

Leak 5: The data is there, but nobody reads it

The biggest mistake in funnel management is not missing data. It is missing follow-through. Most shops have Google Analytics installed, heatmaps running and A/B tests in progress. But the insights end up in reports nobody reads. Fixes get prioritised by “what feels right” instead of “what brings the highest ROI”.

Heatmaps and session recordings only pay off when they run long enough to show a pattern. A classic finding: customers click an element that is not clickable, often a navigation item that only responds to hover. The fix is trivial, the analysis that surfaces it is not. Keep the recordings running and read them.

Worked example: a sporting goods shop with 100,000 sessions a month and a 2 percent conversion rate runs heatmaps and session recordings for three months. The pattern: 40 percent of visitors click a navigation element that is not clickable, a dropdown that only responds to hover.

Making it clickable is a five minute decision and lifts conversion by 6 percent, from 2,000 orders to 2,120. At a 400 euro basket that is 48,000 euros of extra revenue a month. The fix took five minutes; the analysis that found it took three months.

The solution: a weekly funnel review. 30 minutes every week in which you look at last week’s data and make one decision. No report, no PowerPoint, no debate. Look at the data, form a hypothesis, start a test or make the fix.

Conclusion

The conversion funnel is not a mystical concept. It is a quantitative diagnostic tool that shows where your customers drop off. In most shops the five biggest leaks are the same: load time, basket abandonment, mobile optimisation, missing trust and unused data.

No months of waiting, no endless reports. We identify the biggest leak, test the fix and measure the result. If you want to know where your shop’s biggest leak is, Book a Dropoff Analysis with Jörg.