The No-Show Math That Decides If Reminders Pay

When do appointment reminders actually pay for themselves?

Almost always, and the formula proves it in one line. Reminders break even when reminder cost per appointment = no-show rate x recovery rate x value per appointment. Solve for the no-show rate and you get the threshold: break-even no-show rate = cost / (recovery rate x value). With text-message costs published at roughly a penny per segment by Twilio as of July 2026, that threshold lands under one percent for most service businesses. The interesting question is not whether reminders pay. It is how much, and which of your four inputs you are measuring wrong.

Notice what is missing from that formula: appointment volume. It cancels out. Volume decides the size of the prize, not whether the prize is positive.

What are the four inputs, exactly?

Four numbers and nothing else. No-show rate, the fraction of booked appointments where nobody arrives and nobody cancels in time to resell the slot. Value per appointment, which is contribution margin, not revenue. Recovery rate, the fraction of would-be no-shows the reminder converts into attendance. Cost per appointment, all-in: messaging fees, software allocated per appointment, and staff time.

Three of those you can pull from your own booking system this afternoon. The recovery rate is the only one you have to either borrow from published research or measure with a holdout test. Everything else is a lookup.

What is the formula in full?

Write monthly value recovered as appointments x no-show rate x recovery rate x value, and monthly cost as appointments x cost per appointment. Set them equal and appointments divides out of both sides. What remains is no-show rate x recovery rate x value = cost. Rearranged three ways, that single equation answers every version of the question an owner actually asks.

• Break-even no-show rate: cost / (recovery rate x value)
• Break-even appointment value: cost / (no-show rate x recovery rate)
• Break-even recovery rate: cost / (no-show rate x value)
• Monthly net: appointments x (no-show rate x recovery rate x value - cost)

What recovery rate should you assume before you have your own data?

The best-evidenced anchor is the Cochrane systematic review Mobile phone messaging reminders for attendance at healthcare appointments, published 5 December 2013, pooling 8 randomised trials and 6,615 participants. Attendance was 67.8 percent with no reminder and 78.6 percent with text reminders, a risk ratio of 1.14, 95 percent CI 1.03 to 1.26, across 7 studies and 5,841 participants.

Convert that into a recovery rate rather than quoting it as a lift. No-shows fell from 32.2 percent to 21.4 percent, so the reminder recovered 10.8 of the 32.2 percentage points that were going to walk. That is a recovery rate of 33.5 percent, and it is the number to plug in. Two caveats worth stating plainly: the evidence is healthcare appointments, the review graded it low to moderate quality, and 2013 predates the era when everyone ignores texts. Treat 33.5 percent as a starting prior, not as your result.

Does a phone call beat a text?

Barely, and not enough to justify the labor. The same Cochrane review found attendance of 80.3 percent for phone call reminders versus 78.6 percent for text, and text versus phone call came out statistically indistinguishable at a risk ratio of 0.99, 95 percent CI 0.95 to 1.02, across 3 studies and 2,509 participants. The review also reported cost per attendance for text was 55 and 65 percent lower than for phone calls.

Run that through the formula and the picture sharpens. Take three minutes of staff time at an illustrative $22 per hour fully loaded, and a reminder call costs $1.10 per appointment against about $0.05 in messaging, roughly twenty times more. It buys a recovery rate of 38.8 percent instead of 33.5. That is twenty times the cost for a sixteen percent relative improvement in effect.

What does a worked example look like?

Everything below is illustrative, not measured. These are plausible inputs chosen to demonstrate the arithmetic. They are not from any deployment, ours or anyone else's. Substitute your own.

• Appointments booked per month: 400
• No-show rate: 20 percent
• Value per completed appointment: $120 contribution margin
• Recovery rate: 33.5 percent, the Cochrane-derived prior
• Cost per appointment: $0.18, built up below

The cost figure is the only one worth showing the work on. Twilio publishes outbound US SMS at $0.0083 per segment plus a carrier fee running $0.0035 to $0.005 depending on network, so call it $0.013 per segment delivered. A reminder with a confirm link runs two segments. Send two reminders, one at 24 hours and one at two hours, and you are at four segments, about $0.052. Add a $50 per month scheduling platform spread across 400 appointments, $0.125 each. Total $0.177, rounded to $0.18.

Now the arithmetic. Appointments at risk: 400 x 0.20 = 80. Recovered: 80 x 0.335 = 26.8. Value recovered: 26.8 x $120 = $3,216. Cost: 400 x $0.18 = $72. Net: $3,144 per month. Break-even no-show rate: 0.18 / (0.335 x 120) = 0.45 percent.

A business with a 0.45 percent no-show rate does not have a no-show problem. Which is the real finding here: on direct margin math, texted reminders clear break-even by a factor of forty or more for any appointment worth more than pocket change. If someone is selling you a reminder system and leading with ROI, they are answering a question that was never close.

At what appointment value do reminders stop paying?

This is the table that matters. Holding cost at $0.18 per appointment and recovery at 33.5 percent, both illustrative, here is the appointment value you would need to fall below before reminders stop covering themselves, alongside the net monthly return per 100 booked appointments at a $120 margin.

Read the middle column as the kill switch. Unless your appointments are worth less than about $27 in margin and your no-show rate is under 2 percent, texted reminders pay. Redo the same column with a staff phone call at $1.10 per appointment and a 38.8 percent recovery rate and the break-even value at a 5 percent no-show rate jumps to about $57, and at 2 percent to about $142. Manual calling is the option that can actually go underwater, and it does so exactly where appointments are cheap and attendance is already good.

Which input do owners get wrong?

Value per appointment, almost every time, and it is usually wrong in the direction that flatters the decision. Three specific errors:

1. Using revenue instead of contribution margin. A recovered appointment costs you the labor and materials to deliver it. Only the margin is recovered value.
2. Ignoring capacity. If you are booked solid and already backfill empty slots with walk-ins or a waitlist, a recovered no-show does not add a sale, it reshuffles one. Your effective value per recovered appointment collapses toward the difference between the two bookings.
3. Counting cancellations as recoveries. A reminder that turns a silent no-show into a 20-hours-notice cancellation is genuinely valuable, but it is worth the resale value of the slot, not the full margin. Track cancellations and no-shows as separate outcomes or you will overstate both your problem and your fix.

There is a fourth, subtler one on the cost side: opt-outs and annoyance. Messaging costs are trivial, but a reminder cadence that irritates people carries a real cost that never appears on the Twilio invoice. Two messages is a defensible default. Five is a different experiment with a different answer.

How do you measure your own recovery rate?

Stop borrowing the number and run a holdout. Randomly withhold reminders from 10 to 20 percent of bookings for four to six weeks, then compare no-show rates between the two groups. The difference in percentage points, divided by the control group's no-show rate, is your recovery rate. Do not split by day of week, provider, or service type, because those correlate with attendance and will poison the comparison. Randomise at the booking, and log the assignment.

Two structural levers will show up in that data whether you look for them or not. Lead time is one: the 2024 systematic review in Health Science Reports, covering 16 studies, reports no-show rates of 8, 16, and 22 percent for appointment lead times of 0 to 3, 4 to 6, and 28 to 30 days. If you are booking a month out, shortening the funnel may move more attendance than any message ever will. The other is that no-show behavior is not static. An MGMA Stat poll of 265 medical practices on 12 August 2025 found 60 percent reporting rates about the same as the prior year, 13 percent decreased, and 27 percent increased. Whatever you measure, re-measure.

What should you actually do with this?

Run the four numbers before you buy anything. Pull no-show rate and appointment volume from your booking system, compute contribution margin honestly, use 33.5 percent as a placeholder recovery rate, and price the messaging at a penny a segment. If the net is positive by a wide margin, which it will be, the decision is settled and the only remaining questions are implementation quality: does the reminder arrive at a useful hour, can the customer confirm or reschedule in one tap, and does a reschedule write back to the calendar without anyone retyping it.

Those implementation questions are where appointment scheduling systems actually differ from one another, and they are worth more attention than the ROI slide. The same reasoning applies to every other automation on the list: the way to evaluate back-office and workflow automation is to write the break-even equation first, identify which input you are guessing at, and go measure that one. It is the same discipline behind our missed call statistics work, where the leak is bigger and the math is nearly identical.

The Cochrane and Health Science Reports figures above are from published healthcare research and are cited and dated. The messaging prices are Twilio's published US rates as of July 2026. Every other number in the worked example is illustrative and labeled as such. None of it is a measured client result, and no result here should be read as a promise about your business.