The call that rings out at 12:40 is the most expensive leak in the practice
Missed calls during lunch, huddles and the hour after close account for more lost patients than any page element.
Pull your call log by hour and day of week. Count the unanswered calls and the ones under fifteen seconds, which are usually somebody hitting an automated greeting and giving up.
The pattern is almost always the same: a hole over lunch, a hole during the morning huddle, and everything after five o'clock. New-patient calls do not respect that schedule.
Fix the coverage before touching the website. Staggered lunches, an overflow service that can actually place an appointment rather than take a message, and a rule that missed calls get returned within minutes rather than the next day.
Speed of callback beats almost every other intervention available to you. A person shopping for a dentist is calling three practices, and the first human voice usually wins the case.
Grade twenty recorded calls before changing a single page element
The front desk conversation is the highest-leverage surface in the whole funnel and the least examined.
Build a simple rubric: was the phone number captured in the first thirty seconds, was an appointment actually offered, were two specific times proposed, was the insurance question handled without a dead end, was there a reason given to choose this practice.
Listen to twenty calls against it. The results are usually blunt. A large share of new-patient calls end with some version of "let me check and call you back", which is where the patient goes elsewhere.
Coach rather than script. Scripted desks sound like scripted desks, and dental patients are already nervous. What people need is a clear next step offered by a confident voice.
Repeat the exercise quarterly. Call handling degrades quietly whenever the team changes.
Booked, kept and treatment accepted are three different numbers
Measure through to the outcome that pays rather than the one that is easiest to log.
A booked new patient who does not show cost you the slot and the acquisition spend. Counting bookings alone flatters every report you will ever read.
Track three figures by source: appointments booked, appointments kept, and treatment accepted. The gaps between them tell you which problem to work on.
Big gap between booked and kept usually means a reminder problem or a booking made too far out. Confirmation by text with a real reply path closes most of it.
Big gap between kept and accepted is a consultation problem, not a marketing one, and no amount of traffic optimisation will fix it. Say so plainly rather than selling more clicks.
Insurance questions end more new-patient calls than price does
Verification friction quietly kills conversations that were ready to book.
The most common call-ending sentence in dentistry is a version of "we will check your coverage and get back to you". The patient's urgency does not survive that gap.
Publish an accurate, current list of the plans you accept and make sure the front desk has the same list open. Out-of-date lists produce calls you have to walk back, which is worse than no list.
Where verification can happen while the patient is still on the phone, do it. Where it cannot, book the appointment provisionally and verify afterwards, rather than leaving the caller holding the uncertainty.
Out-of-network is not automatically a lost call. A clear explanation of what the patient would actually pay, and how you handle the claim, converts better than avoiding the topic.
Online booking helps only when it writes into the real schedule
A request form dressed up as a booking button costs more trust than it earns.
If the button says Book Now and produces a request that gets confirmed hours later, patients count that as a broken promise and some of them do not answer the callback.
Real-time slots pulled from the practice management system are worth the integration work. Control which appointment types are bookable so a stranger cannot drop an implant consultation into a hygiene slot.
Send a confirmation immediately, then a reminder that can be replied to. Reminders that come from a no-reply number generate no-shows.
Keep the phone number equally prominent. A meaningful share of dental patients, especially for urgent or complex work, want to speak to a person and should not have to hunt for the option.
One practice in one county will never power a button-colour test
At single-location volumes, small-effect split testing burns a quarter and answers nothing.
Statistical power comes from traffic you do not have. Accept that and change what you test rather than pretending the maths works.
Test large swings: a rewritten page rather than a new heading, a form cut from twelve fields to five, a phone number moved into the first screen, an offer present or absent. Effects that big can be seen without a formal experiment.
Prefer changes with a mechanism behind them. Fixing an unanswered call at lunch has a known cause and a known effect; changing a button shade does not.
Where before-and-after comparison is your only option, be honest about the confound. Seasonality alone can move a practice's enquiry volume between quarters, so compare like periods, keep the window long, and treat the result as evidence rather than proof.
Questions we actually get
- What should we fix first?
- Whatever is losing enquiries you have already paid for. In practice that means the call log: find the hours when calls go unanswered and close that gap first. It costs a scheduling change rather than a project budget, and it usually returns more than a redesign. Website work comes after you know the phone is being answered.
- Is an answering service worth it?
- Only one that can book into your schedule. A service that takes messages moves the problem an hour down the line and adds a handoff the patient did not ask for. If you use one, listen to the recordings for the first month; the quality varies enormously and you are lending them your practice's first impression.
- How do we know which marketing source a call came from?
- Distinct tracking numbers per source, with dynamic insertion on the website so the number a visitor sees matches how they arrived. Keep your real number on the Business Profile and in print. Then reconcile the tracked calls against your appointment book monthly, because the platform's count of conversions and your count of booked patients will never match on their own.
- Should we text patients?
- Yes for confirmations, reminders and callback attempts, with consent captured properly and an easy opt-out. Text gets read when calls do not. Keep clinical detail out of it and use it to move people to a conversation rather than to conduct one, and have your own counsel confirm your consent language.
- How much traffic do we need before split testing is worth it?
- More than a single-location practice usually produces, which is why we would rather not sell it to you. The honest alternative is a small number of large, reasoned changes, measured over long enough windows against comparable periods, alongside relentless attention to the call and intake stages where the event volume actually is.
