Private practice
Lumen Clinic — +63% online bookings
Reception was fielding calls all day for appointments that could have been booked online — if anyone had been able to find the booking page.
All case studies
The situation
Lumen Clinic is a private practice with four practitioners and one reception desk. The desk was the bottleneck: three to four hours a day were going on inbound calls, and most of those calls were a person asking for an appointment slot that was already visible in the scheduling software.
The existing site was not badly designed in a visual sense. It was pleasant, calm and entirely passive — a brochure that described the practice to people who had already decided to come. The booking function existed but was reachable only from a footer link, and the widget behind it had been embedded at a fixed 960px width, which made it effectively unusable for the 78% of visitors arriving on a phone.
We tracked what people actually did for three weeks. The single most common path was: land on the home page, scroll to the footer, look for a phone number, leave the site and call. People were not failing to book because they did not want to book online. They were failing because the site made calling the faster option.
What the audit found
The action was three clicks deep
Booking sat behind a footer menu item. Every click between intent and action loses a share of people, and burying the primary action below secondary content inverts the page's priorities against the business's own interests.
The first screen sold nothing
The home page opened with a stock photograph and a welcome paragraph. Neither answered the two questions a visitor actually arrives with: can you see me soon, and what will it cost.
Practical detail was hidden
Opening hours, parking, what to bring and pricing lived across three separate pages. Each missing answer produced a phone call, and each phone call took reception away from the people physically standing in front of them.
The booking widget broke on mobile
Fixed-width embed, tiny tap targets, no keyboard handling. On a 390px screen the confirm button sat outside the visible area entirely — a customer could complete the whole form and never see the way to submit it.
What was broken
- The booking link lived in a footer menu, three clicks from the home page.
- The home page opened with a stock photo and a welcome paragraph, not an action.
- Opening hours and location were buried on a separate contact page.
- The booking widget was unusable on a phone, where most visits came from.
What we changed
Booking is the first thing you see
A live availability panel sits in the hero, showing the next three open slots rather than a generic button.
Built mobile-first
The whole flow was designed at 390px wide first and expanded up, instead of the usual reverse.
Answers before questions
Hours, parking, what to bring and what it costs are on the same page as the booking, so nobody needs to call to check.
The technical detail
The specific decisions behind the numbers, in plain language.
Live availability in the hero
The first screen shows the next three genuinely open slots, pulled from the practice diary and cached for sixty seconds. Concrete availability converts far better than a generic button because it answers the question before it is asked.
Designed at 390px first
Every screen was designed and built for the smallest common phone and expanded upward, rather than the usual reverse. Tap targets are a minimum of 44px, and the confirm action is fixed to the bottom of the viewport throughout the flow.
Booking without a page change
Date, time and details are collected in a single sheet with state preserved. Someone interrupted mid-booking returns to the same step rather than starting again.
Structured data for local search
Hours, location, services and pricing published as machine-readable markup, so the practice appears correctly in map results and search panels — a meaningful share of new bookings never touch the home page at all.
Content managed in one place
Hours and closures are edited once and propagate to every page and the search listing together, which removes the class of problem where a public holiday is updated in one location and not the other.
How it ran
Audit
Three weeks of behavior data reviewed, reception call log categorized by reason, competitor booking flows walked end to end.
Design
Mobile-first screens, tested with six patients on their own phones rather than a lab device.
Build
Diary integration, structured data, content model and staff training.
Measure
Booking and call tracking, with two rounds of copy adjustments based on drop-off.
How we measured it
Bookings were tracked for eight weeks after launch against the same eight-week window the previous year, adjusted for a 9% year-on-year traffic increase so growth alone could not explain the result.
Online bookings rose 63%. Inbound calls fell 38%, which reception measured directly as roughly six hours a week returned to in-person work. The proportion of bookings completed on a phone went from 21% to 68%.
No-shows also fell by 14% — an unplanned effect we attribute to the confirmation email and reminder now containing parking instructions and what to bring, two of the things people previously called to ask.
online bookings
inbound phone calls
reception time saved weekly
“The phones went quiet and the diary filled up. That was the whole brief, really.”
What we took from it
- If the phone rings for something the website could answer, that is a design defect, not a staffing problem.
- Showing real availability beats a button that says Book now.
- For a local service business, the search listing is part of the website whether you treat it that way or not.