How to find conversion leaks on your SaaS website

What is a conversion leak?

A conversion leak is a specific, fixable reason a visitor leaves the path to your goal. It is narrower than “low conversion rate”. A leak has a location, a cause and a fix: the pricing page does not show what the trial includes, the signup form asks for a phone number, the demo button sits below a wall of testimonials on mobile.

Thinking in leaks changes the work. Instead of redesigning a page because a number looks low, you list the reasons people stop, prove each one, and close them in order.

Leaks versus general UX problems

Not every usability issue costs conversions. A leak sits on the conversion path and affects a visitor who was ready to act. A slow blog archive is a UX problem; a slow checkout is a leak. The Leak library documents the 32 leaks Leakly checks for, each with the evidence it requires.

Where conversions typically break

On a SaaS site, the path to a signup or demo usually crosses four or five pages. Each page has a typical way of failing.

The largest drop is not always the most important one. A small drop at checkout can cost more than a large one on a blog post, because the visitors at checkout had already decided.

How to identify conversion leaks

This is the method Leakly runs automatically. You can run it by hand on any site in an afternoon.

1. 1

   Pick one conversion goal

   Trial started, demo booked or order placed. One goal per audit keeps findings comparable.

2. 2

   Map the path backwards

   List the pages a visitor crosses to reach that goal, starting from the conversion page and working back to the main entry points.

3. 3

   Walk it as a first-time visitor

   Use a fresh browser, on a phone and on a desktop. Note every point where you hesitate, scroll to find something, or have to guess. Most people scan pages rather than read them,^[2] so judge what is visible at a glance.

4. 4

   Collect evidence for each point

   A screenshot of the area, the exact copy, a measurement such as load time or field count. A finding without evidence is an opinion.

5. 5

   Name the leak precisely

   Write it as a diagnosis: “The signup form asks for company size before showing the product.” Not “improve the form”.

A first-screen checklist

• The headline names the outcome, not the category.
• The primary action is visible without scrolling, on mobile too.
• The action label says what happens next.
• There is one primary action, not three of equal weight.
• Something on screen reduces risk: price, trial terms or a credible proof point.
• The main content loads fast enough to read. Google’s Core Web Vitals give concrete thresholds.^[3]

Understanding the evidence

Evidence turns a finding into something a team can agree on and act on. A good finding answers four questions: where it is, why it matters, what to change, and how sure you are.

Before and after

The fix for this finding is mostly copy. The button keeps its place and its colour; the label and the line below it change.

Prioritizing conversion problems

Once you have a list, rank it. Two factors do most of the work: how close the leak is to the conversion, and how strong the evidence is. Effort breaks ties.

Fix one or two items, ship them, then audit again. Large batches make it impossible to tell which change mattered. Baymard Institute’s checkout research is a useful reference for how much friction small form decisions add.^[1]

Real-world examples

The fastest way to learn the pattern is to see it on a real report. We ran Leakly on our own site: the example shows the score, the finding with its annotated capture, and the fix it recommends.

Audit example · leakly.aiWe ran Leakly on leakly.aiSee the finding

Frequently asked questions