How Qured made its booking flow convert better

If your booking flow feels like it leaks, you are probably right, and you are probably not sure where. That was the position Qured was in when I started working with them. This is the story of how we fixed their booking flow UX, and the order we did it in matters more than any single change.

What makes a booking flow convert?

A booking flow is the set of screens a visitor goes through between choosing a service and getting a confirmed booking. It converts well when each of those screens is easy to understand and easy to finish, so very few people give up along the way.

That sounds simple, and in my experience every booking flow loses people at more than one step. The hard part is knowing which step is losing the most. Until you know that, every fix is a guess.

Qured sells at-home health testing, both directly to consumers and to employers. The booking flow was the whole business. When I arrived, it was leaking at every step, and nobody could say with confidence which leak mattered most.

Why I measured the booking flow before changing it

The brief had two parts. First, work out where the booking funnel was actually losing people. Then run a weekly A/B testing routine against it.

I kept those two parts in that order on purpose. It is tempting to start redesigning screens on day one, because that feels like progress. Without a funnel you can trust, though, you cannot tell whether a change helped, hurt or did nothing.

So the first job was the measurement. We worked with GA4, Google Tag Manager and Mixpanel, with reports in Looker Studio. I explain how I make GA4 and Mixpanel agree in my post comparing GA4 and PostHog, and Qured is one of the examples there.

The process was plain and careful. We named every event in the booking funnel and agreed the names and their meaning in writing. The developers then built them from that written tracking plan.

Before anyone looked at a chart, I checked the new events against real traffic. If the funnel said fifty people reached a step, I wanted to know that roughly fifty people really did. These are the same kind of checks I describe in my post on whether your GA4 is lying.

That agreement sounds small. In practice it ended a lot of arguments. When marketing, product and the founders all see the same number, the conversation moves from “is this right?” to “what do we do about it?”.

How we found the step that was losing people

With the funnel measured properly, the drop-off stopped being a guess. We could see each step of the booking flow and how many people left at each one.

This is the part people often skip. They look at a page with a high exit rate and decide that page is the problem. I find it far more useful to look at steps in the flow, because one page can hold several steps and one step can span several pages.

Research on checkouts shows why this matters. The Baymard Institute’s summary of 50 studies puts the average abandonment rate at about 70%, and lists many separate reasons people give. Extra costs, trust worries, forced account creation and a long or complicated process all appear. Each reason shows up at a different step, so you need to know which step is yours.

Once we could see which step lost the most people, the list of things worth testing became short and obvious. That short list was the plan for the weeks that followed.

Why we tested every week

The tests ran every week, on both the booking funnel and the lead funnel. A weekly rhythm keeps momentum, and it makes testing part of how the team works instead of a big project that happens once a quarter.

Every test followed the same simple rules:

• One clear question. Each test was based on a step we could see losing people.
• Closed when the result was clear. We did not stop a test early just because it looked good for a few days.
• Written up, win or lose. The tests that lost were written up just as carefully as the ones that won, because they taught us what not to try again.

I explain why this habit matters so much in my post on running experiments that build on each other. The short version is that a record of past tests makes each new test a little smarter than the last.

What happened to the numbers

Over the engagement, three numbers moved:

• The conversion rate from session to booking went up by 247%. This is the share of sessions that ended in a booking, compared with where it was when we started.
• The conversion rate from session to lead went up by 186%. This is the share of sessions that ended with someone leaving their details.
• Lead volume went up by 421% from the same testing routine.

All three numbers moved because one thing changed first. Once the whole company agreed what a booking was and could trust the funnel, the tests had something solid to stand on. You can see the full summary in the Qured case study.

When this approach is right, and when it is not

This way of working suits a health-tech business when:

• The booking flow is where most of your revenue comes from
• You have enough visits and bookings for weekly tests to give clear answers
• Your developers can make changes to the flow every week or two
• Nobody is sure which step is losing people, or different tools give different answers

It is probably the wrong approach when:

• You have very few bookings each month. Weekly tests will not reach a clear answer, so I would look at session recordings and fix the obvious problems instead.
• Your tracking has never been set up. Then the first step is a one-off build, which I run as Growth Foundations. I describe what I set up first in a separate post.

If you want a quick first look at where your own booking flow loses people, I also built an AI agent that walks a booking flow as your customer. It is a good way to spot the obvious problems before any tracking work begins.

How I would start with your booking flow

At Qured I worked as their fractional Head of Growth, running the weekly tests alongside their team. I would start your booking flow the same way. In the first weeks we agree what each step is called, check that the numbers are right, and find the step that loses the most people.

Health-tech booking flows have their own challenges, including consent and keeping health data out of tools that should not have it. You can read more about how I work with health-tech companies.

Read the full Qured case study to see the results in one place, or book a thirty-minute call below. I will tell you honestly whether your booking flow needs a weekly testing routine, a one-off fix or nothing just yet.