Conversion rate optimization: where the real leak usually is

You can't test your way out of the wrong traffic. The real CRO wins happen before the button.

Why page tests deliver diminishing returns

If you're investing in conversion rate optimization and testing button colors, headlines, and form fields without much to show for it, the reason is usually that you're optimizing the wrong part of the funnel. Real CRO isn't a pile of micro-tests on the page. It's finding where conversion actually leaks, and the biggest leaks are rarely the button.

Conversion rate optimization gets flattened into "test things on the page": try a green button, rewrite the headline, move the form. These tests aren't useless, but they're the smallest lever, and teams that start there usually grind out tiny gains while the real problem sits untouched. A page tweak can lift a healthy funnel a few percent. It can't rescue a funnel that's leaking somewhere bigger.

The three upstream leaks

The bigger leaks are almost always upstream of the button. The first is the traffic itself: if your ads bring people who were never a fit, the best page in the world won't convert them, because the problem arrived before they did. The second is the offer: if what you're asking people to do isn't compelling, no amount of layout polish fixes a weak deal. The third, and the most common, is message mismatch: the ad promises one thing, the page delivers another, and the visitor bounces in the gap between the two. None of those show up in a button test.

The right order of operations

This is why the order of operations matters more than the number of tests. Start by finding the single worst drop-off in your funnel, the step where the most people leave, and ask why they're leaving there specifically. Often it's not the checkout or the form, it's that the traffic was wrong or the promise didn't match. Fix the biggest leak first, then move to the next, and only then spend energy on the small page-level details that A/B tests are good at.

So before the next round of button tests, zoom out. Look at where people actually fall out, check whether the right people are even arriving, and make sure the page keeps the promise the ad made. CRO done in that order changes the number. CRO done as endless micro-tests just keeps you busy.

Keep reading

Related readingThe 4 silent leaks after the click that are draining your budgetRelated readingWhy am I getting the wrong leads (job seekers instead of customers)?Related readingWhy is my ROAS dropping even though I'm spending more?

Next step

The answer changes when we look at the entire operation.

In our first conversation, we look at the business, its goals and where value may be getting lost. No pressure to sign.

Book my diagnosis →Back to the Blog

Get the next article in your inbox.

Growth, paid media and sales operations. Only what helps you understand and decide.

I want the next one →