Case study
24hrDoc
An urgent-care funnel that had been redesigned three times without moving, rebuilt on evidence — and measured against the baseline agreed before anything shipped.
- Client
- 24hrDoc
- Sector
- Urgent care · Telehealth
- Stage
- Scaling
- Scope
- Audit · Intake · Checkout
MEASURED
- +22.8%
- Overall site conversion
- +28%
- Paid search conversion
- +21%
- Organic conversion
- +11%
- Purchase funnel completion
- Dropped
- Cost per acquisition
- The stakes
- Three redesigns had already shipped. Each one changed how the site looked; none of them changed what qualified visitors did. With no baseline, nobody could say which had helped — so the next redesign would have been another guess, paid for at the same price.
- The read
- Two weeks of instrumentation and session review before anything was touched, then a step-by-step drop analysis with the cost of each leak attached. That turned an argument about taste into a ranked list. Two findings carried most of the loss: on the condition pages the process instructions were pulling more attention than the action, so visitors read the steps, kept scrolling, and drifted past checkout; and on the homepage a full navigation block sat level with the primary CTA, so the first screen offered several options and no clear next step.
- The move
- Intake rebuilt around what people can answer in one sitting. Pricing and checkout rebuilt mobile-first around one offer and one action. Released in stages, each against the same source of truth, so every change could be read on its own.
THE REFRAME
Not a design problem wearing a data costume. A data problem that earlier redesigns tried to solve with aesthetics.
So the work started with instrumentation, not a new visual direction.
Making the leak visible
The funnel was not short of traffic. Qualified people arrived, started, and left at a small number of identifiable steps — and the cost of each of those steps could be worked out to the dollar.
Once the drops were ranked by cost rather than by how they looked, the order of work chose itself. The most expensive step went first, shipped with a written hypothesis rather than a taste argument about how it should look.
APPROACH
Analytics went in first as the one source of truth, so every decision after that was argued from the same numbers rather than from personal judgement. Real drop-off was then tracked across the whole patient journey, from choosing a service through to checkout, which showed where qualified visitors were leaving instead of merely confirming that conversion was low. The full urgent-care redesign was then built to the framework the funnel produced, and launched on March 31, 2026.
STEP 01
Set performance metrics. Analytics are put in place as a single source of truth so that every decision is based on numbers, not opinions.
STEP 02
Map the funnel, find the leaks. Real drop-off analyzed across the patient journey, from landing through service selection to checkout, to isolate where qualified visitors abandoned rather than simply that conversion was low.
STEP 03
Redesign, informed, not decorative. A large-scale redesign of the urgent-care site built on what the funnel showed. Launched March 31, 2026.
STEP 04
Measure what actually worked. The same source of truth used after launch to separate the changes that earned their lift from the ones that were merely different.
STEP 05
Propagate the winners. Validated patterns rolled across the rest of the site, which is what turns a redesign into a system: prove a change on real behavior, then apply it everywhere it fits.
THE REDESIGN
Curated to tell the story, not to show every screen.
BEFORE AND AFTER
The hero was not the problem everyone believed it to be — the offer had simply never been part of it. A full navigation block sat level with the primary CTA, so a visitor’s first screen offered several options and an action at the same weight. That competition is what produced the hesitation: plenty of choices, no clear next step. The redesign resolves it by building the hero around one offer and one call to action, and moving the navigation below it, so the first thing anyone sees is the single thing to do.
Among the condition pages, the process instructions were drawing more attention than the action. Visitors read the steps, kept scrolling to understand them, and drifted away from the funnel instead of into checkout. The redesigned page holds a single focus: the hero is simplified so “Get treatment online” is unmissable, and the steps move down the page where they no longer compete. Beside the action sit the levers that ease hesitation at the moment of decision — benefit points on what you get, a “no insurance required” trust signal, and a clear time estimate, “prescription within 45 minutes”, so visitors know exactly what they are committing to before they commit.
THE SYSTEM / VALIDATED, THEN ROLLED OUT
Three elements recurred on every page that gained: the priced dark CTA with the price attached (“Get treatment”, “Get medication”), the Trustpilot badge anchored to the hero image, and the same hero structure — accented headline, benefit checklist, single action.
THE RECURRING ELEMENTS
Three, held identical across every page that gained: the dark, priced CTA that names the action and the price together — “Get treatment”, “Get medication”, “Start 5-minute assessment”, each at $39.99; the “$39.99 · no hidden fees” pill sitting above every headline; and the Trustpilot badge anchored into the hero image beside a short benefit checklist. Each one was kept because it was measured, not because it was assumed.
THE RESULT
Site conversion went from 7.54% to 9.26%, and held there for six weeks.
- Baseline
- 7.54%
- After
- 9.26%
- Change
- +22.8%
- Held
- Six weeks
Measured against the baseline agreed before anything shipped, on the same source of truth, over the same traffic mix. Changes that did not move it were written down too — you keep that record, and the instrumentation, after the engagement ends.
THE FULL READOUT
| +22.8% | Overall conversion | 7.54% → 9.26% |
|---|---|---|
| +28% | Paid search | 8.79% → 11.26% |
| +21% | Organic | 5.26% → 6.39% |
| +11% | Purchase funnel | 9.55% → 10.6% |
| Reduced | Cost per acquisition, as more visitors converted at a lower cost to acquire each | No figure released |
| +23s | Average engagement time | — |
HOW IT WAS MEASURED
The same source of truth used to diagnose the funnel was used to judge it afterwards, which is what separates the changes that earned their lift from the ones that were merely different.
Patterns that proved themselves were then rolled across the rest of the site. That is the difference between a redesign and a repeatable system: every pattern that scaled had already earned its lift on real users.
“Our website has undergone multiple redesigns, yet none of those versions led to a significant improvement in conversions. Jose recognized the problem as one related to the data rather than the design itself. He pinpointed the places where potential customers abandoned the process and redesigned those touchpoints accordingly. The result: conversion rates increased and cost of customer acquisition dropped.”
Joseph
Founder, 24hrDoc
WHAT THIS DEMONSTRATES
Conversion problems are often misdiagnosed as design problems. The value here wasn’t a new visual language. It was the discipline to measure the funnel first, locate the exact points of loss, redesign against real behavior, and validate what worked before rolling it across the rest of the site. That last step is the difference between a redesign and a repeatable system: every pattern that scaled had already earned its lift on real users.
The services behind this
Start a project
Think your funnel is leaking somewhere?
Send the numbers you already have. I’ll tell you where I’d look first, before you commit to anything.
Book a callOr email me