Before you redesign your website, find where people actually stop

"The website looks dated."

It's one of the most common things founders say when growth stalls, and one of the most expensive to act on first.

A redesign is visible, fixable and feels like progress. It has a brief, a budget and a launch date, which is more than most real problems offer. But it changes everything at once, so it teaches you little. If conversions improve, you won't know which change did it. If they don't, you won't know what to undo.

Before you brief a designer, run a count.

Follow the buyer, count every step

Follow the buyer from first visit to paying customer, and count how many make it through each step. Then ask four questions:

1. Can a new visitor understand who you help, what problem you solve and what to do next within about ten seconds? If people leave before they explore, the fix is usually the message, not the design.
2. Do people who ask for a demo hear from you quickly, and with context? If requests sit for days, the leak is in the handoff.
3. Do trial users reach a first real result? If they sign up and stall, the website has probably done its job. Onboarding hasn't.
4. Do new customers get what they were promised? If many leave early, check what the ads told them.

Start with the biggest drop. It's rarely the homepage.

Weigh each drop by volume and value, because the steepest percentage drop isn't always the most expensive one. And if your volumes are small, treat the counts as directional, not definitive.

When the website really is the problem

Sometimes it is. It's slow, broken on mobile, can't be updated without a developer, or describes a company you no longer are.

Then fix it. But fix the specific thing, agree in advance what improvement you expect, and measure the step it affects.

The redesign that got smaller

A SaaS founder I worked with was sure the website was holding the business back. It looked dated, competitors' sites were far more polished, and conversion was below target. The plan was a full redesign and rebrand.

We ran the count first. People were getting through the website at a reasonable rate. They found the right pages, requested demos and started trials.

The drop came afterwards. Demo requests sat in the CRM too long before anyone followed up. Trial users weren't reliably reaching the point where the product showed its value.

So the redesign went on hold. Demo requests got a fast reply that used the context from the enquiry. New trial users got clear first steps.

The redesign didn't disappear. It got smaller. When it happened, it solved specific problems instead of carrying the whole growth strategy.

The website is often the easiest thing to blame because it's the easiest thing to see.

Find the step first.