The person answering your phone is also checking somebody out
In most single location practices the largest leak is not a page element, it is that the front desk has three jobs at once.
A new patient call is a five to ten minute conversation involving insurance, scheduling and reassurance. It arrives while somebody is collecting a payment, answering a hygienist and signing for a delivery.
Start by measuring it rather than debating it. Almost every phone system will report answer rate, missed calls and time of day. Pull a month and look at where the misses cluster. Lunch and the last hour usually stand out.
Then choose a real fix. A dedicated new patient line, an overflow answering service with a booking script rather than a message pad, or an automatic text back on every missed call so the person who could not get through has somewhere to go.
Whichever you pick, listen to a sample of calls yourself. Owners are consistently surprised by how many callers ask one insurance question, get an uncertain answer, and hang up without being offered a time.
Most practices try a new inquiry once and quietly give up
One outbound call with no voicemail and no text is not follow up, and it is where the majority of web inquiries die.
Somebody submits a form at nine at night. The desk calls at eleven the next morning from a number the patient does not recognize, does not leave a message, and marks the inquiry as unreachable.
Write down a cadence and hold to it. Reply within minutes if the inquiry lands in business hours. Text if the call is not answered, because people who filled in a form at night will answer a text and not a call. Send two or three specific times rather than a request to call back. Try again the next morning, then close it out.
Name who does it and when. Inquiries that belong to everybody belong to nobody, and the ones that arrive on a Friday afternoon are the ones that vanish.
Speed matters more than polish. A household transferring from out of state is contacting several practices in one sitting and generally books with whoever gets to them first with an actual time.
The spouse is easiest to book before the patient leaves the chair
The cheapest appointment you will schedule this month is the second one from a household already sitting in your practice.
Families arriving from another state usually come in one at a time. One person tests the practice, everybody else waits to hear how it went. Nobody asks, so nobody books.
Make it part of checkout. Ask who else in the household needs to be seen, offer times while the patient is standing there, and schedule the next hygiene visit before they leave rather than adding them to a recall list.
Track it as a number. Appointments booked per household, measured monthly, tells you more about the health of the practice than most of what appears in a marketing report.
The same logic applies to treatment that has already been diagnosed and never scheduled. Working that list is a conversion activity, and it costs nothing in advertising.
A two hour sedation block is the one appointment worth a deposit
Long appointments carry the most production and the most risk, and most practices protect them with nothing at all.
A short notice cancellation on a two hour block is not a scheduling annoyance. It is the most expensive hour of the week, and it is very hard to refill on the day.
Options worth considering include a deposit against the appointment, a confirmation call from a clinical team member rather than an automated message, and a short standby list of patients who have said they can come in at short notice.
Policy details need care. What you can charge and how you communicate it can be affected by your payer contracts and by state rules, so the specific wording is worth confirming with your own counsel before it is introduced.
Whatever you adopt, apply it consistently and say it plainly at booking. Surprise fees generate complaints, and the review that follows costs more than the empty chair did.
Every Monday, match the inquiry list to the appointment book by name
Attribution in dentistry is a weekly reconciliation habit, not a dashboard you install once.
Take the week's calls, forms and online bookings and check each name against the schedule. Booked, not booked, or still open. Fifteen minutes of work, and it is the only source of numbers you will actually trust.
Record the source on the chart at the moment the appointment is made. Asking how did you hear about us weeks later produces a courteous guess, not attribution, and it is where most reporting goes wrong.
Carry the same names forward a few weeks so you can see who kept the appointment. Booking and showing up are different events, and the gap between them is where the real problem often sits.
Then the number that matters becomes available: cost per kept new patient, by source and by area. Metro averages hide a lot, and an inquiry from Buda behaves differently from one three minutes away.
Two weeks of intake data cannot settle any argument you care about
A single practice does not generate enough traffic for small split tests, so change large things deliberately and measure them over a full cycle.
Button colors and headline variants require volumes a one location practice will never produce. Running those tests yields noise, and acting on the noise makes things worse.
Change things big enough to see with your eyes. Adding an early morning block. Answering the phone through lunch. Replacing a request form with real online booking. Introducing a text back on missed calls.
Hold each change long enough to read it, and compare against the same weeks in a previous year rather than against last month. Your own calendar has patterns in it, and reading a two week window will lead you to the wrong conclusion with confidence.
In the meantime, use qualitative evidence properly. Recorded calls, front desk notes and the reasons patients give for choosing elsewhere will tell you where the leak is faster than any experiment you can afford to run.
Questions we actually get
- What should we fix first?
- Answering. Pull a month of phone system data, find where the missed calls cluster, and put real coverage there before touching anything on the website. In most practices the number of unanswered new patient calls is larger than any gain available from page changes, and it costs nothing in advertising to recover them.
- How do we track which marketing produced a patient without breaking our listings?
- Use a distinct tracking number for each paid channel while keeping your main published number consistent everywhere your practice is listed, so the listings do not disagree. Then record the source on the chart at booking and reconcile names to the appointment book weekly. The chart entry is the reliable record. The dashboards are supporting evidence.
- Is online booking going to reduce our control of the schedule?
- Not if you constrain it. Publish only the appointment types and lengths you want filled, restrict it to the blocks you are willing to give away, and require a phone conversation for complex cases. Reviewing the first month of bookings usually settles the concern one way or the other quickly.
- Should we run A/B tests on the website?
- Generally no, at a single practice. The traffic will not support detecting small differences, so the results will be noise. Make changes large enough to see, hold them for a full cycle, and compare against the equivalent period in a previous year. Reserve genuine split testing for cases where several locations feed the same site and volume is high.
- How long should we give this before judging it?
- Long enough to see kept appointments, not just bookings, which usually means at least a quarter. Booking rate can improve immediately after an answering change while the effect on production shows up later. Nobody can responsibly promise a percentage or a date, but the weekly reconciliation will show movement well before a quarter is out.
