Industry
Doctors & Clinics
A no-show is an hour nobody refills.
Practices and clinics lose time in three places: on the phone, on patients who do not turn up, and on documentation. The levers are known, the numbers behind them sourced - and none of them requires findings or patient data to land with a platform provider.
Calculator
How many appointments actually happen · Appointment utilisation
Same numbers, your business: pull the levers and watch what moves.
Appointment utilisation
Admin share of working time
Hebel
Simplified calculation for a practice with scheduled consultations. The admin share blends hospital and practice data, the effect of online booking is an assumption, that of AI assistance deliberately small - the way the studies measure it. Fees and specialty are not included.
Appointment utilisation
Admin share of working time
Simplified calculation for a practice with scheduled consultations. The admin share blends hospital and practice data, the effect of online booking is an assumption, that of AI assistance deliberately small - the way the studies measure it. Fees and specialty are not included.
Where the margin goes
Share of available appointments that are booked and kept. Every no-show is time that is not refilled - and admin that still has to be done.
The phone is the bottleneck - and the most expensive booking path.
44 % of insured patients cannot get through, 42 % say it got worse over five years. 64 % have already booked online, 58 % outside opening hours. A booking portal solves that - and takes patient data and a fee with it. Your own booking path on your website solves it too, and both stay with you. How much utilisation that brings, nobody has measured; the calculator is deliberately cautious about it.
Source: GKV-Versichertenbefragung 2024, n = 3.512; Bitkom 12/2025, n = 1.145
A no-show is an hour nobody refills.
In Vienna, 10 % of patients a day do not turn up on average, 20 % on peak days. A reminder the day before cuts no-shows by a quarter - from 21 to 15 % across 26 studies -, a second reminder lifts attendance to 78 % versus 62 %. That is not a project. That is an automation: SMS the day before, replying “no” frees the slot.
Source: ORF Wien 10/2024, Urologieberufsverband und Ärztekammer Wien; Robotham et al. 2016, BMJ Open - 26 Studien
Documentation eats half the day. AI gives back minutes, not hours.
Doctors in training in Austria spend 24.5 of 47.4 weekly hours on admin and documentation, employed doctors in Germany three hours a day on average. AI assistance measures honestly: 16 minutes less documentation per eight hours of patient time, three minutes per discharge letter - but the burnout score drops below the clinical threshold. And a local model is enough: 93.1 % usable documents, no byte leaves the building, the doctor signs off.
Source: ÖÄK Ausbildungsevaluierung 2024 mit ETH Zürich, n > 4.800; MB-Monitor 2022 (IQME), n = 8.464; JAMA 04/2026 - 5 Zentren, 1.809 Nutzer vs 6.772 Kontrolle; Heilmeyer et al., JMIR Medical Informatics 2024 (Uniklinik Freiburg)
The numbers behind it
- 10 %
- of patients a day do not turn up in Vienna - 20 % on peak daysSource: ORF Wien 10/2024, Urologieberufsverband und Ärztekammer Wien
- 21 % → 15 %
- no-shows without vs. with a reminder, 26 studiesSource: Robotham et al. 2016, BMJ Open - 26 Studien
- 51.6 %
- of the working time of doctors in training in Austria is admin and documentationSource: ÖÄK Ausbildungsevaluierung 2024 mit ETH Zürich, n > 4.800
- 44 %
- of insured patients are dissatisfied with reaching practices by phoneSource: GKV-Versichertenbefragung 2024, n = 3.512
- 16 min
- less documentation per eight hours of patient time with AI assistance - measured, not promisedSource: JAMA 04/2026 - 5 Zentren, 1.809 Nutzer vs 6.772 Kontrolle
- 93.1 %
- of documents from a local, non-commercial model usable after minor correctionsSource: Heilmeyer et al., JMIR Medical Informatics 2024 (Uniklinik Freiburg)
What we build for it
Practice website with its own booking path and automatic reminders, local AI assistance for findings and letters on your server, tracking with consent - and visibility for the services you actually want to offer. For clinics the same at scale: outpatient appointments, reminders, documentation.
Common questions
May a practice use AI for findings at all?
If the data does not leave the building and the doctor signs off every output - yes. That is exactly how we build it: local model on your own infrastructure, human in the loop, no byte going out. That a non-commercial model is enough, Freiburg University Hospital showed on 90,000 documents.Do I need a booking portal like Doctolib or DocFinder?
Not necessarily. Your own booking path on the website is enough for most practices - and keeps patient data and the relationship with you. Portals bring reach, but for a fee and with the data at the provider. We tell you beforehand what works for your practice.How much does an appointment reminder really bring?
Across 26 studies, no-shows drop from 21 to 15 % with one reminder. With a second one, attendance rises to 78 % versus 62 %. SMS or email the day before, replying “no” frees the slot - that is the whole mechanism.How much time does AI save in documentation?
Less than the vendors promise: the largest study measures 16 minutes per eight hours of patient time, three minutes per discharge letter. The burden, however, drops measurably. We calculate with what is measured - and tell you so.Does this apply to clinics too?
Yes, at other scales: outpatient departments lose millions of appointments to no-shows - 8.1 million a year in England alone -, and hospital doctors spend more than half their time on admin. Reminders, booking paths and local AI work there just the same, only with more leverage.
Next step
Where do you stand?
The analyzer reads your practice or clinic website in a minute - booking portal, tracking, speed, visibility. What of that matters to you, we think through together.