Blog · November 22, 2026 · Updated November 22, 2026
A running-late text is not a no-show
When a customer texts that they are running late, many chatbots reply politely while the no-show rule fires anyway. Hive AI builds agents for dental practices, med spas and home services that treat a late message as a schedule change: hold the slot, apply your policy, tell the team and send a receipt.
BY HIVE FORENSICS AI

A patient in Boca Raton is booked for a 9:00 cleaning. At 8:52 she texts the practice number: "So sorry, stuck behind an accident on Glades, about 15 minutes late." The texting agent replies "Thanks for letting us know!" and does nothing else. At 9:10 the scheduling system flags the appointment as a no-show, releases the chair to the waitlist and sends her an automated note about the missed-appointment fee.
She walks in at 9:14 to find her slot given away and a fee notice on her phone. The front desk never saw her text. The hygienist would have taken her. Now the practice has an upset patient, a fee to reverse and a review waiting to be written, all because an agent treated a message as small talk instead of a change to the schedule.
A running-late text is not a no-show. It is a customer telling you they are still coming. What the business does next should follow its own late policy, not whatever the automation happens to do by default.
Why agents get this wrong
Most texting and chat agents are built to answer questions and book new appointments. A late message is neither, so it falls through. The agent sends a friendly acknowledgement, the message sits in a thread nobody is watching, and the no-show rule in the scheduling system fires on time, exactly as configured.
The same gap shows up across the businesses we work with:
- Dental practices. A patient running late for a cleaning or a crown seat gets marked absent and the chair goes to the waitlist, even though the late policy allows 15 minutes.
- Med spas. A client texts that she is 10 minutes out for a 60-minute treatment. Nobody decides whether the provider can still do the full service, a shortened one, or needs to rebook, so the client finds out at the desk.
- Home services. A homeowner replies to the "tech on the way" text saying she won't be home for 30 minutes. The tech arrives, nobody answers, and the visit is closed as "customer not home" with a trip charge.
In each case the customer did the right thing and the system punished them for it.
What a good agent does with a late message
The agents Hive AI builds for dental practices, med spas and home services companies treat a running-late message as a schedule event, and follow a short set of rules:
- Recognize it. Messages like "running late", "stuck in traffic", "10 minutes out" or "won't be home until 2" are classified as a late notice, not a general question.
- Find the right appointment. The agent matches the sender to today's booking and confirms the time, so it never pauses the wrong visit.
- Hold the slot. It pauses the automatic no-show or "customer not home" rule for that appointment and logs the expected arrival time.
- Apply the business's own policy. Inside the grace period the business set, the agent confirms the visit is still on. Past it, the agent offers only the options the business approved: a shortened service, the next open slot or a callback. It never invents a fee waiver or promises a full service the provider hasn't agreed to.
- Tell the people in the building. The front desk, provider or dispatcher gets the late notice with the new ETA, so the hygienist or tech knows before the clock runs out, not after.
- Send a receipt. The customer gets a short message saying what was recorded and what happens next. The team gets the same record on the appointment.
Questions to ask about your current setup
If customers can text, chat or call your business about an appointment, ask your vendor or your team:
- When a customer says they are running late, what exactly happens to the appointment?
- Does anything stop the no-show, missed-appointment fee or trip charge from firing?
- Who on the team sees the message, and how fast?
- What does the agent say when the customer is past the grace period, and who approved those words?
- Can you show the record of one late message from last week, start to finish?
If nobody can answer the last question, late customers are being handled by default settings.
Evidence over demos
A demo shows an agent booking an appointment. The real test is the morning a customer is stuck in traffic and your system has to decide whether she still has a chair. At Hive AI we build chatbots and voice agents for Florida dental practices, med spas and home services companies that follow your policies, keep your team informed and leave a record of every change, so a polite text never turns into a lost customer.
If you want an agent that keeps late customers instead of charging them, talk to Hive AI. If you want your own team to learn how to build agents with these guardrails, look at the Hive AI bootcamp.