Monday morning, and the appointment book is already full. Waiting underneath it are the weekend’s messages: a new client asking about color, a regular who needs to move a visit, and someone wondering whether their deposit arrived.
Our stylist is fictional, but the situation will be familiar to anyone who serves customers for a living. None of those messages is hard to find. What’s hard is finding a quiet moment between clients to understand each one, look up the details and reply properly. An automatic “thanks, we got your message” is kind to the customer, and I’m all for it. It just doesn’t do much for the stylist, whose to-do list is exactly as long as it was on Saturday.
Here is what more useful help could look like.
A briefing, not a pile
Instead of a stack of unread messages, the owner opens a short briefing like this:
- New color inquiry: the client has not said whether this is a first visit. A reply asking the approved consultation questions is ready to review.
- Appointment change: the customer requested next Tuesday afternoon. Two eligible openings were found; choose one to offer.
- Deposit question: the payment record could not be matched confidently. Check the original message and payment system before responding.
Every item ends in a decision rather than a summary, so nobody has to reread three conversations to work out what’s being asked. The original messages sit alongside for checking. The third item is my favorite, because it’s the one a less careful system would get wrong. “Could not confirm” is honest, useful information. A cheerful “Yes, we’ve got it, thanks!” would have been quicker to write and much harder to take back.
Who does what behind the scenes
AI earns its place in the reading: working out what a loosely worded request is really after, noticing what’s missing and drafting a reply that fits. Ordinary automation handles the plumbing, such as collecting messages from supported channels, retrieving booking details and setting a reminder when a conversation is still waiting. A simple form response doesn’t need AI just to say thank you.
Connecting those pieces is the real implementation work, and the options depend on the booking and messaging tools you already use. You might keep every reply in a review queue, or let approved routine answers go out on their own. Pricing exceptions and other decisions you reserve stay with you either way.
If a consultation question goes unanswered, a later follow-up can be prepared with the original conversation attached. Once the customer replies, the reminder should stop. Otherwise the automation becomes one more thing you have to supervise.
A reply you can use today
For a new inquiry that needs more detail before you can help, adapt this:
“Thanks for getting in touch about [service]. Before we suggest the next step, could you tell us [one detail needed to proceed]? That will help us give you a useful answer.”
Fill in the brackets and ask only for what you actually need. Don’t call a time confirmed until it’s booked, or a payment received until someone has checked. The FTC’s business data-security guide also recommends limiting personal information to what the business needs. A friendly follow-up shouldn’t quietly turn into a questionnaire about someone’s entire history with you.
Start with the question you keep asking
If there’s a preliminary question you answer most weeks, that’s a wonderfully concrete place to start. Bring one real incoming request and the reply you’d normally send. You don’t need to design a new customer-management system first.
The goal is a shorter path from “someone contacted us” to “this person knows what happens next,” and a Monday briefing that shows that progress while leaving the decisions you care about in your hands.
Sources
Sources support the reported facts in this article. Practical recommendations are analysis by Alasdair for New Kent Digital Works.