How to reduce no-shows and late cancellations

A missed appointment leaves a staffed hour with nobody in it, and that hour cannot be sold again. The fixes below can be running by Friday: two reminders, a way to answer them, a short list of bookings worth a personal call, a waitlist that refills what opens up, one written policy, and for home service crews, arrival windows you can actually keep.

What the research covers, and what it does not

Most measured evidence on no-shows comes from medical practices, so read every figure below as healthcare evidence unless it says otherwise. It is still the closest thing to a controlled test for any appointment business. The schedules, scripts and routines beside the studies are what we recommend, not tested methods, unless a study is named with them.

The size of the problem: a 2018 systematic review in Health Policy of 105 studies across medical specialties put the average healthcare no-show rate at about 23%, highest in Africa (43.0%) and lowest in Oceania (13.2%), with no US figure of its own. The predictors reported most often were a long lead time (days between booking and visit) and a history of missing before.

A physical therapy study gets misquoted constantly. In a 2021 PLoS One study of one US provider with 828 clinics in 26 states (444,995 patients from 697 of those clinics, treated for musculoskeletal conditions in 2016 and 2017), 73% of patients had at least one visit in their episode of care that ended canceled or unattended, because the study counted cancellations as no-shows. It does not mean 73% skipped an appointment. Rates across clinics, states and providers ranged from 15% to 31% under that definition, and more prior cancellations and a longer gap before the visit predicted more missed visits.

So count no-shows (nobody came, no notice) and late cancellations (notice inside your policy window) as two separate numbers. A reminder that makes canceling easier can lower one while raising the other.

Send two reminders: one 2 to 3 days out, one the day before

The best-tested change is a second reminder. In a Kaiser Permanente Colorado trial of 54,066 patients at 25 primary care clinics (2018), patients reminded both 3 days and 1 day before missed 4.4% of the time, against 5.8% with only the 3-day reminder and 5.3% with only the 1-day reminder. Among high-risk patients the rates were 20.5%, 25.0% and 24.2%. The extra reminder did not lower visit satisfaction.

A 2016 BMJ Open meta-analysis of 21 randomized studies agrees: notified patients were 23% more likely to attend and 25% less likely to no-show than patients sent nothing, and multiple notifications improved attendance more than one (though not no-shows). Only one of the 21 studies was American.

Text or phone? The 2013 Cochrane review found texts raised attendance against no reminder and performed about the same as phone calls, and two of its studies found texts cost 55% and 65% less per attendance. The authors rated that evidence low to moderate quality.

The schedule we set up for appointment businesses:

When	What goes out	Who handles replies
At booking	Confirmation text: date, time, place, what to bring, the policy in one line	Automatic
3 days before (2 for short-notice bookings)	First reminder, answerable	Automatic, with “R” replies routed to a person
1 day before	Second reminder, answerable	Automatic, same routing
24 hours before, still unconfirmed	A phone call from the front desk	A person
Morning of (optional)	Arrival nudge with parking or entrance notes	Automatic

The unconfirmed call and the same-day nudge are our additions to the trial’s schedule; the nudge never replaces the day-before message.

A first reminder for a physical therapy practice: “Hi Jordan, you’re booked with Priya for physical therapy Thursday at 9:30 a.m. Reply C to confirm or R to change the time. Reply STOP to opt out.” Start every text with your business name; the examples here leave it out.

Vet clients lean to text too: in a 2019 Frontiers in Veterinary Science survey of pet owners whose dog or cat had been hospitalized (an October 2018 online convenience sample), 56.5% got confirmations as phone messages but only 29.4% preferred that, while 51.5% preferred text.

Make every reminder answerable

A reminder people can answer tells you who is coming while there is still time to fill the gap.

In a 2024 randomized trial at Penn Medicine of 59,994 patients flagged as high risk of missing, published in 2026, adding an automated call that let patients confirm, cancel or reschedule on top of text reminders cut the no-show rate from 11.3% to 9.6% and raised completed appointments from 75.9% to 77.8%. In a separate text survey of 186 patients who had missed appointments, the top reason, given by 22%, was not knowing they had an appointment at all.

When a UK sexual health and HIV clinic that already sent text reminders added the option to cancel by text in early 2018, its monthly did-not-attend rate fell from 14.42% to 12.18% of bookable appointments, while cancellations rose from 24.4% to 38.68%. It was one site, before and after, not randomized, but it shows the trade-off: answerable reminders turn some silent no-shows into cancellations, which pay off only if someone refills the slot that day.

This is not legal advice; confirm with counsel before you set up reminder texts.

Pick reply words carefully. Under the FCC rule in force today (47 CFR 64.1200), a reply of stop, quit, end, revoke, opt out, cancel or unsubscribe is a valid request to stop texts, and it must be honored within a reasonable time not to exceed ten business days. So “reply CANCEL to cancel your appointment” invites people to opt out of future reminders by accident. Use letters such as C to confirm and R to reschedule, and offer canceling through a link or a call. When someone does reply “cancel”, stop texting that number as the rule requires, and have a person call to ask whether they also meant the appointment.

Keep reminders free of offers. CTIA’s messaging guidelines treat a text sent for the purpose a customer gave their number for as informational, and note that adding a call to action such as a coupon code may make it promotional. A “book your next facial and save” line turns a reminder into marketing, which carries stricter consent rules.

Clinics: know the healthcare exemption’s conditions. The same rule lets healthcare providers send certain reminders without prior consent, but only if the messages cost the patient nothing, go only to the number the patient provided, name the provider, carry no marketing, billing or collection content, comply with HIPAA, generally run 160 characters or less, number at most one a day and three a week, and include a STOP opt-out. Ordinary business texting is not free to the recipient, so many clinics cannot use it; the sounder footing is usually the consent the patient gave by providing the number, kept within the scope of their care. Med spas are not named among the covered providers.

Route every “R” reply to a person the same business day and release the old slot the moment it moves.

Call the bookings most likely to miss

A short daily list deserves a human voice: anyone with a prior no-show or late cancellation, anyone booked far ahead, and first visits. The first two are the predictors in the Health Policy review and the physical therapy study. Adding first visits, and calling rather than texting, goes beyond what either study tested.

What the call sounds like: “Hi Wes, it’s Marisol from the front desk. I’m checking in about your appointment Tuesday at 2 p.m. Does that still work, or would a different day be easier?” If the answer is a different day, move it on the spot and offer the old slot to the waitlist before hanging up.

Shorten lead times where you can: since a long gap before the visit predicts missing in both studies, a slot three weeks out is a weaker booking than one this week.

Refill canceled slots from a waitlist

At UCSF, an opt-in automated waitlist that texts or emails patients when an earlier slot opens sent 60,660 offers for 21,978 open appointments between June 2022 and March 2023. 11% of offers were accepted and 8.9% became completed visits, patients were seen a median 14 days sooner, and the tool added 2,576 patient service hours.

Offers went out in evening batches to several patients per slot and expired after 12 hours. Older and non-English-speaking patients accepted less often, so keep a phone path for them. With roughly one offer in ten taken, each slot needs several offers out at once.

Our setup:

• Ask at booking: “If an earlier time opens, would you like a text?” Record the answer.
• When a slot frees up, text three to five people at once, first reply wins.
• Give the offer a short expiry, and say so in the text.
• Text everyone else that the slot has gone, so the list stays trusted.

Sample offer: “An earlier physical therapy time just opened: Wednesday at 8 a.m. with Priya. Reply Y to take it. This went to a few people on the early list and closes at 8 a.m. tomorrow.”

Online booking: fewer no-shows, more cancellations

Across nearly two million appointments at Johns Hopkins Community Physicians from 2019 to mid-2021, self-scheduled patients missed fewer appointments (2.7% against 4.6% for agent-booked) but canceled more (37.6% against 27.0%) and kept fewer overall (59.6% against 68.3%). It was observational, and younger, commercially insured patients used it more.

In a July 2025 MGMA Stat poll of 244 medical practices, 71% said fewer than a quarter of their patients self-schedule, and groups with the most self-scheduling were more likely to say it reduced no-shows (self-reported, sponsored poll). Patients say they want it: in Experian Health’s 2025 vendor survey, 80% of 1,004 US adults who received care or oversaw a dependent’s care chose “I would like to schedule appointments anytime, from my home or mobile device”, while 55% of 213 provider revenue-cycle decision-makers said patients are unsure how to navigate self-scheduling.

For salons, salon software vendor Phorest claims online bookings no-show 49% less than phone or in-salon bookings, with no sample, period or definition. Online bookers may simply be more organized; it is a claim, not a benchmark.

Put booking, rescheduling and canceling online for routine visits, link all three from every reminder, and count it a win only if the waitlist refills what the extra cancellations free up.

Write one cancellation policy and apply it to everyone

This is not legal advice; confirm with counsel before charging any fee, and check your state’s rules for every payer you accept.

In our experience a policy works when people see it before booking and in each reminder, and see it applied the same way to everyone. A three-sentence version for a salon or massage studio that takes deposits:

“Cancel or move any time up to 24 hours before your appointment by replying to your reminder or using its link. With less notice, or if you miss the visit, the deposit you paid at booking is kept. We apply this to every client.”

Check that your booking software actually shows it on the booking screen and in the reminder; a policy nobody saw is hard to enforce.

Deposits. We found no measured figure for how much deposits cut salon no-shows; the percentages that circulate come from vendor blogs with no method. We recommend deposits or a card on file only for long or high-value slots (a color correction, a two-hour body treatment, a holiday boarding stay) and for clients who have missed before. For scale only, Zenoti’s 2025 benchmark figures (vendor data) put salons at an 8% cancellation rate and a 3% no-show rate and med spas at a 5% no-show rate, with no definitions, sample or geography, and Zenoti operates globally.

Medicare patients. CMS allows charging Medicare patients for missed appointments only if the same policy and the same amount apply to all patients, Medicare and non-Medicare alike. The fee is a charge for a missed business opportunity, must not be billed to Medicare, and Medicare pays nothing toward it (Medicare Claims Processing Manual, chapter 1, section 30.3.13). Waiving it for some patients while charging Medicare patients fails that rule.

Medicaid patients. Colorado’s Medicaid agency states that federal and state policy prohibit charging Medicaid members for missed or canceled appointments, or asking them to sign forms accepting that liability, citing CMS guidance that missed appointments are a cost of doing business. Members with Medicare or commercial primary coverage may be charged in some circumstances, with criteria disclosed in advance. It is one state’s statement of the federal position; check your own state.

Restaurants: let guests cancel without calling

The published restaurant data covers cancellations, not no-shows. Across same-store full-service restaurants on Toast’s reservation platform, 17% of reservations were canceled in Q3 2024, down from 19% a year earlier (vendor data). Toast also reports that 45% of reservations were made for the same day.

Some no-shows are failed cancellations. In Sheryl Kimes’s Cornell survey of 1,230 diners in 2007 (self-selected, mostly New York Times Diner’s Journal readers, and before online booking was common), respondents strongly agreed that guests should call to cancel, yet often found it hard to reach the restaurant, and on dining blogs the main complaint about card guarantees was a busy signal or no answer when trying to cancel. The report recommends a designated cancellation line or email.

The fix we set up is a cancel path that never depends on the host picking up: a one-tap link in the confirmation and the reminder. Sample confirmation: “You’re booked for four at 7:30 p.m. Saturday. Plans change? Use the link below to move or cancel, no call needed.” If you hold a card, state the free-cancellation window in the same message.

Home services: your no-show is the missed arrival window

For a plumber, electrician or appliance tech, the customer rarely fails to show. The business does, by arriving outside the window. None of this research is from those trades exactly; read it as the closest evidence.

Window length. In a 2018 Zion and Zion survey of 1,044 US homeowners by a marketing agency, first-impression satisfaction with a quoted plumber arrival window averaged 8.01 out of 10 for a one-hour window, 7.18 for two hours, 6.25 for three and 5.52 for four. City dwellers grew less satisfied 24 minutes sooner than suburban or rural homeowners. These were reactions to a hypothetical window, not real visits.

Arriving on time and calling ahead. In J.D. Power’s 2018 study of telecom in-home technician visits, satisfaction with scheduling was 138 points higher when the customer was contacted before arrival, and 49 points higher with windows of an hour or less than with two-hour windows. Overall satisfaction was 871 on time and 683 late, but some late arrivals were days late, and telecom is only an analogy.

What homeowners say they would do. In a 2022 Service Direct survey of 559 homeowners by a lead-generation company, 23% said they would stop doing business with a contractor who arrived outside the window without notice, while 48% would expect no compensation for a late arrival if the technician contacted them beforehand. In FIELDBOSS’s 2025 vendor survey of 1,000 homeowners who had hired HVAC service, 13.0% named late or out-of-window arrivals as their single biggest frustration. In a 2011 survey for field-service software vendor TOA Technologies, mostly about cable and utility visits, 70% said they would recommend a company that arrived on time, falling to 27% if the technician was 15 minutes late.

Two texts per job is our practice: one at booking with the window in writing, then one when the tech leaves the previous job, or, if that job overruns, one before the window is missed, with a new time.

• At booking: “You’re booked for Thursday between 10 a.m. and noon to look at the water heater. Keith is your technician and will text when he’s on the way.”
• On the way: “Hi Mr. Alvarez, it’s Keith. I’m leaving my last job now and should reach you around 10:40. Reply here if the gate code has changed.”
• Running late, sent instead of the on-the-way text: “Hi Mr. Alvarez, it’s Keith. The job before yours ran long and I’m going to miss your window. I can be there by 12:45, or first thing tomorrow at 8. Which is better for you?”

Quote the shortest window you can actually hit, tighter on city routes. In our experience a narrow window missed does more damage than a wide one kept.

California Civil Code 1722. This is not legal advice; confirm with counsel. The statute requires retailers with 25 or more employees to agree on a four-hour window for deliveries, and for service or repair of merchandise, when the customer must be home, and requires cable companies and utilities to offer four-hour windows. If the visit does not start in the window, the customer can sue in small claims court for lost wages, expenses or other actual damages up to $600, except for delays from unforeseen or unavoidable events beyond the business’s control or a customer who was not home. It is also a defense if a retailer made a diligent attempt to warn of such a delay and then came within two hours of a newly agreed time. Waivers are void. The law covers the business missing the window, not the customer missing the visit, and on its face it does not reach trades that service no merchandise, such as plumbing or roofing; that reading is ours, not the statute’s words.

The weekly review

Set aside twenty minutes every Monday and pull these from your booking software or CRM:

• No-show rate: visits missed with no notice, divided by visits booked.
• Late-cancellation rate: cancellations inside your policy window, divided by visits booked.
• Confirmation rate: bookings confirmed by reply before the day-before reminder.
• Refill rate: freed slots that the waitlist or a call filled.
• On-time rate (home services): visits started inside the quoted window.

Split each by new and returning customers, lead time, booking channel and weekday. If far-ahead bookings miss most, shorten lead times or add a reminder a week out; if late cancellations climb after you open online booking, work the waitlist harder. Never compare yourself with the physical therapy chain’s 73%, which counts advance cancellations.

Figures to leave out of your planning

The large national dollar cost of missed medical appointments, and the per-visit cost that usually comes with it, both trace to a 2017 commentary by a scheduling-software vendor’s executive with no data behind it. The claim that most no-shows simply forgot has no published source; the nearest measured figure is Penn Medicine’s survey, where 22% of 186 patients who had missed a visit said they did not know they had one. And we found no study measuring how far deposits cut salon no-shows. Plan with your own weekly numbers instead.

Questions about no-shows and late cancellations

How many appointment reminders should you send?

Send two: one 2 to 3 days before and one the day before. In the Kaiser Permanente Colorado trial that pair beat either reminder alone, and patients were no less satisfied.

What should a reminder text say?

It should give the day, time, place and provider, then a way to answer: C to confirm, R to reschedule, a link to cancel, and STOP to opt out. Leave out promotions.

Are text reminders as effective as phone calls?

They perform about the same, according to the 2013 Cochrane review, and in two of its studies texts cost less per attendance. Calls still earn their place for unconfirmed and high-risk bookings.

Can a practice charge Medicare patients a no-show fee?

Yes, but only if the same policy and the same amount apply to all patients. The fee is never billed to Medicare, and Medicare pays nothing toward it. This is not legal advice; confirm with counsel.

Can you charge Medicaid patients for missed appointments?

Generally no: Colorado’s Medicaid agency states that federal and state policy prohibit it. Check your own state’s Medicaid agency and confirm with counsel before writing a policy.

Do deposits reduce no-shows?

We found no study that measured it for salons or studios, and the percentages on vendor blogs come with no method. Use deposits for long or high-value slots and repeat no-shows, stated before booking.

Does online booking reduce no-shows?

It can cut outright no-shows while raising cancellations: at Johns Hopkins Community Physicians, self-scheduled patients missed fewer visits but kept fewer overall.

How long should an arrival window be for home service visits?

Use the shortest window you can reliably keep. Homeowners rated a quoted one-hour plumber window 8.01 out of 10 and a four-hour window 5.52 in a 2018 survey.

Sources

Checked on October 3, 2026. Rules and fees change, and many are set state by state or city by city: confirm the current requirements with the agency that issues them before you apply.

• Dantas et al., Health Policy, 2018: systematic review of no-show rates and predictors
• Bhavsar et al., PLoS One, 2021: missed visits at a US physical therapy provider
• Gurol-Urganci et al., Cochrane review, 2013: text message reminders for appointments
• Robotham et al., BMJ Open, 2016: meta-analysis of text-based appointment notifications
• Steiner et al., American Journal of Managed Care, 2018: reminder timing trial at Kaiser Permanente Colorado
• Kavanaugh et al., NEJM Catalyst, 2026: confirm, cancel or reschedule calls at Penn Medicine
• Tomkins, Naylor and Morgan, International Journal of STD and AIDS, 2019: cancel-by-text pilot
• Woodcock, Sen and Weiner, JAMIA, 2022: self-scheduling at Johns Hopkins Community Physicians
• Ganeshan et al., Journal of Medical Internet Research, 2024: automated waitlist at UCSF
• MGMA Stat poll, July 2025: patient digital self-scheduling
• Experian Health, 2025 State of Patient Access (vendor survey)
• Kogan, Schoenfeld and Santi, Frontiers in Veterinary Science, 2019: how pet owners want confirmations
• CMS Medicare Claims Processing Manual, chapter 1, section 30.3.13 (missed appointments)
• Colorado Department of Health Care Policy and Financing: charging members for missed appointments
• 47 CFR 64.1200 (FCC consent, opt-out and healthcare message rules)
• CTIA Messaging Principles and Best Practices, 2023
• Zenoti salon booking survey and 2025 benchmark figures (vendor)
• Phorest, salon no-shows (vendor claim)
• Toast, 2024 restaurant reservation trends (vendor data)
• Kimes, Cornell Hospitality Report, 2008: diner attitudes to reservation policies
• Zion and Zion, 2018: consumer satisfaction and plumber arrival windows
• J.D. Power via PR Newswire, 2018: telecom in-home technician punctuality
• Service Direct, 2022 homeowner survey (vendor)
• FIELDBOSS, 2025 HVAC homeowner survey (vendor)
• Multichannel News (Next TV), 2011: TOA Technologies waiting survey (vendor)
• California Civil Code 1722 (delivery and service windows)