{"course":{"slug":"ai-for-small-business","title":"AI for a Small Business","url":"https://addaly.com/learn/ai-for-small-business","level":"beginner","track":"work"},"lesson":{"slug":"bookings-reminders-no-shows","title":"Bookings and no-shows: the case where AI is the wrong tool","url":"https://addaly.com/learn/ai-for-small-business/bookings-reminders-no-shows","position":34,"of":88,"minutes":9,"answer":"Reminders and confirmations are rules work that a booking system does free and reliably, and a deposit beats every message; the model earns its place only on the varying parts: filling a cancelled slot.","keyIdea":"Reminders and confirmations are rules work that a booking system does free and reliably, and a deposit beats every message; the model earns its place only on the varying parts: filling a cancelled slot, the awkward conversation, and finding the pattern in six months of bookings."},"body":{"format":"markdown","text":"## The most valuable automation in this whole course is not AI\n\nA no-show costs you the slot, and in a clinic, a salon or a workshop it is often the single biggest recoverable loss in the week. Ten appointments a week missed at an average of 1,500 is 720,000 a year in whatever currency you count in.\n\nThe fix is a reminder, and a reminder is a rules problem. Same message, fixed time before the appointment, no judgement, no variation. A booking system's built-in reminder is free or nearly so, sends reliably, and cannot invent a time.\n\nPutting a language model in this path adds cost, latency, a failure mode and nothing else. This is the clearest example in the course of the T/R/J classification earning its keep: the box is R, and R does not want a model.\n\n## What actually reduces no-shows\n\nIn rough order of effect, from what practitioners consistently report:\n\n1. **A deposit.** By some distance the strongest, and the one most owners resist. Even a small one changes behaviour, because it converts a free option into a purchase.\n2. **A reminder 24 hours before**, with a one-tap way to cancel or rebook. Making cancellation easy feels wrong and reduces no-shows, because the alternative to an easy cancellation is not attendance: it is silence.\n3. **A confirmation at booking**, with the appointment added to their calendar.\n4. **A second reminder** on the morning, for anything more than a week ahead.\n5. **Calling the ones who have missed before.** A person, not a system.\n\nNone of these needs AI. All of them are available in booking tools costing between nothing and about USD 15 a month, and most shops already pay for something that does them and has not switched them on.\n\n## Where AI does add something\n\nThree narrow places, and they are worth having once the boring parts are working.\n\n**Rebooking a gap.** A cancellation at 11am leaves a hole at 2pm. Drafting a short, personal message to three customers on the waiting list (*not* a broadcast) is a good use, and it is the kind of task that never gets done manually because it needs doing immediately.\n\n**The awkward message.** Third no-show from a regular customer. Writing that without either being a doormat or losing them is hard, and a draft to react to is easier than a blank screen.\n\n**Understanding the pattern.** Feed six months of booking data into a spreadsheet and ask what the no-shows have in common. Time of day, day of week, lead time between booking and appointment, first-timers versus regulars. The answer is often something specific and fixable (bookings made more than three weeks ahead, or the first slot after lunch) and you would not have noticed.\n\nThat third one is real analysis, and it is the sort of thing a small business almost never does because it is nobody's job.\n\n## Language and timing\n\nTwo details that affect outcomes more than the wording of the message.\n\n**Send in the language the customer used to book.** Obvious, frequently not done, and the reason some no-shows are people who did not read the reminder.\n\n**Send when it can be acted on.** A reminder at 11pm for a 9am appointment gives nobody a chance to rearrange. Mid-morning or early evening the day before is when a cancellation can still be filled.\n\n## The rule to take away\n\nAutomate the reliable, repetitive part with a reliable, repetitive tool. Use the model for the parts that vary: the message to the customer who has let you down twice, the analysis of why Tuesdays are bad.\n\nBusinesses that get this backwards end up with an expensive, occasionally creative system doing a job that a free scheduled message did perfectly.\n\n## The message itself\n\nSince the message is a template, write it once and write it well. Four things belong in it and most reminders are missing two.\n\n**What, when, where.** Including the address, because people book in one place and travel from another.\n\n**What to bring or do beforehand.** The form, the referral, the measurements, coming with clean hair. Missing preparation is a wasted appointment that shows up in your figures as an attendance.\n\n**A one-tap way out.** A link that cancels or rebooks without a conversation. This is the field most owners omit deliberately and it is the one that recovers slots.\n\n**Who it is from.** A message from an unfamiliar number about an appointment reads like a scam to a growing number of people, and gets ignored.\n\n## Watch what the automation does to the diary\n\nOne side effect worth expecting. When cancelling becomes easy, cancellations go up and no-shows go down. Total attendance usually improves, because a cancelled slot at 24 hours can be refilled and a no-show cannot.\n\nIf you only count cancellations, this looks like a change for the worse. Count filled slots instead, which is the number that pays the rent."},"check":{"question":"A clinic wants to reduce no-shows and is choosing between an AI assistant that writes personalised reminders and switching on the deposit feature in its existing booking software. What does the evidence in this lesson suggest?","options":[{"text":"The deposit, because it converts a free option into a purchase, and personalised wording addresses a problem that reminders already solve mechanically.","correct":true},{"text":"The AI reminders, since personalisation increases the chance a message is read and acted on rather than ignored as routine.","correct":false,"tempting":"Assumes engagement is the binding constraint, when the people who miss appointments have usually read the reminder and simply had nothing at stake."},{"text":"Both together, since the deposit addresses commitment while personalised reminders address forgetfulness, and no-shows have both causes in roughly equal measure.","correct":false,"tempting":"Sounds balanced but spends money and supervision on the weaker lever first, and asserts an equal split that the clinic has not measured."},{"text":"Neither, until the clinic analyses six months of booking data to establish which appointment types and time slots actually account for the missed appointments in the first place.","correct":false,"tempting":"Sensible analysis that should happen anyway, but it delays a change with a known large effect in order to refine where to apply it."}]},"prev":{"slug":"complaints-and-refunds","title":"The angry message, and what your reply commits you to","url":"https://addaly.com/learn/ai-for-small-business/complaints-and-refunds"},"next":{"slug":"reviews-and-reputation","title":"Reviews: asking, answering, and the line you must not cross","url":"https://addaly.com/learn/ai-for-small-business/reviews-and-reputation"},"authorship":"Founder and series editor: Dr Mustafa Mun. Written by AI models to a written standard. Exam keys checked blind by a second model: 3,665 of 3,666 agreed. Academic review invited.","review":{"status":"Draft, open for academic review"},"cite":"Addaly (2026) 'Bookings and no-shows: the case where AI is the wrong tool', in AI for a Small Business. Edited by M. Mun. Addaly. Available at: https://addaly.com/learn/ai-for-small-business/bookings-reminders-no-shows (Accessed: 13 October 2026).","licence":{"name":"CC BY-NC-SA 4.0","url":"https://creativecommons.org/licenses/by-nc-sa/4.0/"},"reader":{"schema":1,"title":"Bookings and no-shows: the case where AI is the wrong tool","courseTitle":"AI for a Small Business","module":{"index":4,"slug":"customers-and-the-front-desk","title":"Customers, from first message to resolved complaint"},"answer":null,"keyIdea":"Reminders and confirmations are rules work that a booking system does free and reliably, and a deposit beats every message; the model earns its place only on the varying parts: filling a cancelled slot, the awkward conversation, and finding the pattern in six months of bookings.","hero":{"kind":"plate","image":null,"alt":null,"caption":null},"blocks":[{"t":"h2","id":"the-most-valuable-automation-in-this-whole-course-is-not-ai","runs":[{"text":"The most valuable automation in this whole course is not AI"}]},{"t":"p","runs":[{"text":"A no-show costs you the slot, and in a clinic, a salon or a workshop it is often the single biggest recoverable loss in the week. Ten appointments a week missed at an average of 1,500 is 720,000 a year in whatever currency you count in."}]},{"t":"p","runs":[{"text":"The fix is a reminder, and a reminder is a rules problem. Same message, fixed time before the appointment, no judgement, no variation. A booking system’s built-in reminder is free or nearly so, sends reliably, and cannot invent a time."}]},{"t":"p","runs":[{"text":"Putting a language model in this path adds cost, latency, a failure mode and nothing else. This is the clearest example in the course of the T/R/J classification earning its keep: the box is R, and R does not want a model."}]},{"t":"h2","id":"what-actually-reduces-no-shows","runs":[{"text":"What actually reduces no-shows"}]},{"t":"p","runs":[{"text":"In rough order of effect, from what practitioners consistently report:"}]},{"t":"ol","start":1,"items":[[{"text":"A deposit.","b":true},{"text":" By some distance the strongest, and the one most owners resist. Even a small one changes behaviour, because it converts a free option into a purchase."}],[{"text":"A reminder 24 hours before","b":true},{"text":", with a one-tap way to cancel or rebook. Making cancellation easy feels wrong and reduces no-shows, because the alternative to an easy cancellation is not attendance: it is silence."}],[{"text":"A confirmation at booking","b":true},{"text":", with the appointment added to their calendar."}],[{"text":"A second reminder","b":true},{"text":" on the morning, for anything more than a week ahead."}],[{"text":"Calling the ones who have missed before.","b":true},{"text":" A person, not a system."}]]},{"t":"p","runs":[{"text":"None of these needs AI. All of them are available in booking tools costing between nothing and about USD 15 a month, and most shops already pay for something that does them and has not switched them on."}]},{"t":"h2","id":"where-ai-does-add-something","runs":[{"text":"Where AI does add something"}]},{"t":"p","runs":[{"text":"Three narrow places, and they are worth having once the boring parts are working."}]},{"t":"p","runs":[{"text":"Rebooking a gap.","b":true},{"text":" A cancellation at 11am leaves a hole at 2pm. Drafting a short, personal message to three customers on the waiting list ("},{"text":"not","i":true},{"text":" a broadcast) is a good use, and it is the kind of task that never gets done manually because it needs doing immediately."}]},{"t":"p","runs":[{"text":"The awkward message.","b":true},{"text":" Third no-show from a regular customer. Writing that without either being a doormat or losing them is hard, and a draft to react to is easier than a blank screen."}]},{"t":"p","runs":[{"text":"Understanding the pattern.","b":true},{"text":" Feed six months of booking data into a spreadsheet and ask what the no-shows have in common. Time of day, day of week, lead time between booking and appointment, first-timers versus regulars. The answer is often something specific and fixable (bookings made more than three weeks ahead, or the first slot after lunch) and you would not have noticed."}]},{"t":"p","runs":[{"text":"That third one is real analysis, and it is the sort of thing a small business almost never does because it is nobody’s job."}]},{"t":"h2","id":"language-and-timing","runs":[{"text":"Language and timing"}]},{"t":"p","runs":[{"text":"Two details that affect outcomes more than the wording of the message."}]},{"t":"p","runs":[{"text":"Send in the language the customer used to book.","b":true},{"text":" Obvious, frequently not done, and the reason some no-shows are people who did not read the reminder."}]},{"t":"p","runs":[{"text":"Send when it can be acted on.","b":true},{"text":" A reminder at 11pm for a 9am appointment gives nobody a chance to rearrange. Mid-morning or early evening the day before is when a cancellation can still be filled."}]},{"t":"h2","id":"the-rule-to-take-away","runs":[{"text":"The rule to take away"}]},{"t":"p","runs":[{"text":"Automate the reliable, repetitive part with a reliable, repetitive tool. Use the model for the parts that vary: the message to the customer who has let you down twice, the analysis of why Tuesdays are bad."}]},{"t":"p","runs":[{"text":"Businesses that get this backwards end up with an expensive, occasionally creative system doing a job that a free scheduled message did perfectly."}]},{"t":"h2","id":"the-message-itself","runs":[{"text":"The message itself"}]},{"t":"p","runs":[{"text":"Since the message is a template, write it once and write it well. Four things belong in it and most reminders are missing two."}]},{"t":"p","runs":[{"text":"What, when, where.","b":true},{"text":" Including the address, because people book in one place and travel from another."}]},{"t":"p","runs":[{"text":"What to bring or do beforehand.","b":true},{"text":" The form, the referral, the measurements, coming with clean hair. Missing preparation is a wasted appointment that shows up in your figures as an attendance."}]},{"t":"p","runs":[{"text":"A one-tap way out.","b":true},{"text":" A link that cancels or rebooks without a conversation. This is the field most owners omit deliberately and it is the one that recovers slots."}]},{"t":"p","runs":[{"text":"Who it is from.","b":true},{"text":" A message from an unfamiliar number about an appointment reads like a scam to a growing number of people, and gets ignored."}]},{"t":"h2","id":"watch-what-the-automation-does-to-the-diary","runs":[{"text":"Watch what the automation does to the diary"}]},{"t":"p","runs":[{"text":"One side effect worth expecting. When cancelling becomes easy, cancellations go up and no-shows go down. Total attendance usually improves, because a cancelled slot at 24 hours can be refilled and a no-show cannot."}]},{"t":"p","runs":[{"text":"If you only count cancellations, this looks like a change for the worse. Count filled slots instead, which is the number that pays the rent."}]}],"check":{"question":[{"text":"A clinic wants to reduce no-shows and is choosing between an AI assistant that writes personalised reminders and switching on the deposit feature in its existing booking software. What does the evidence in this lesson suggest?"}],"options":[{"text":[{"text":"The deposit, because it converts a free option into a purchase, and personalised wording addresses a problem that reminders already solve mechanically."}],"correct":true,"tempting":null},{"text":[{"text":"The AI reminders, since personalisation increases the chance a message is read and acted on rather than ignored as routine."}],"correct":false,"tempting":[{"text":"Assumes engagement is the binding constraint, when the people who miss appointments have usually read the reminder and simply had nothing at stake."}]},{"text":[{"text":"Both together, since the deposit addresses commitment while personalised reminders address forgetfulness, and no-shows have both causes in roughly equal measure."}],"correct":false,"tempting":[{"text":"Sounds balanced but spends money and supervision on the weaker lever first, and asserts an equal split that the clinic has not measured."}]},{"text":[{"text":"Neither, until the clinic analyses six months of booking data to establish which appointment types and time slots actually account for the missed appointments in the first place."}],"correct":false,"tempting":[{"text":"Sensible analysis that should happen anyway, but it delays a change with a known large effect in order to refine where to apply it."}]}],"order":[3,2,0,1],"why":null},"prev":{"slug":"complaints-and-refunds","title":"The angry message, and what your reply commits you to","minutes":9},"next":{"slug":"reviews-and-reputation","title":"Reviews: asking, answering, and the line you must not cross","minutes":9}}}