Automatically Follow Up Lapsed Customers Without Losing Control

What this service is

AiStaffo connects your customer and booking data to an AI worker that runs a lapsed-customer follow-up sequence from start to finish. It finds customers with no order or booking inside a window you define, sends approved messages, logs each response, routes interested replies to a person, and removes anyone who asks to stop. It does not set discounts, prices or offers.

Who it is for

The service suits any business where customers come back on a rhythm. Clinics and dental practices, salons, equipment servicing firms, training providers, schools, wholesale suppliers and repeat-order retailers all fit this pattern. A business with one-off purchases and no natural return point gets little from a reactivation sequence, and we say so during the audit.

The same design works for a five-person firm and for a group with thousands of staff. The difference is the number of record systems involved and the number of people who need to approve messages.

What is included

• Lapse rules for each customer group, based on the last order or booking date you choose
• A short, pre-approved message sequence, usually two or three steps spaced several days apart
• Email sends from your own domain and WhatsApp sends through your WhatsApp Business account
• Submission of WhatsApp message templates to Meta for review before first use
• A consent record for each contact showing where and when permission was given
• Logging of every reply, with time, channel and outcome, on the customer record
• Hand-off of interested replies to a named person, with a short summary of the history
• Opt-out handling that removes the contact from further follow-ups in the same run
• A monthly report covering messages sent, replies, opt-outs, bounces and WhatsApp service-message volume

How it runs

Audit

We review where customer, order and booking data live, how complete the last-activity dates are, and what consent each contact has given. We also check which lapse window makes sense for each product or service. Most of the audit is spent on data quality, because a lapse rule built on wrong dates sends the wrong messages.

Process map

We draw each step on one page: who is selected, which message goes out, what counts as a reply, who receives a warm reply, and what happens on an opt-out. Each decision point gets a named owner. You approve the map before anything is built.

Build

We connect the data source, write the message steps with you, and submit any WhatsApp templates to Meta. Email content is tested on a small internal group. Nothing goes to real customers until the test run matches the map.

Run

The worker runs on the schedule you approve. Replies are logged as they arrive, and interested people are passed on the same day. In the first month we review a sample of sends each week and adjust timing or wording where replies show a problem.

Support

After go-live, we handle template changes, new lapse groups, and fixes when a data source changes. You receive the monthly report and a short review of opt-outs and complaints.

WhatsApp costs from 1 October 2026

This is the part most owners have not yet priced in. Meta's charge for service messages starts on 1 October 2026. Each business number receives 1,000 free service messages a month, and anything above that is billed per message at the rate for the recipient's market. Utility templates sent inside an open 24-hour customer window are also charged again from the same date. Outreach to a lapsed customer outside that window must use a template, so the follow-ups themselves follow template billing rules. The monthly report separates these categories so you can see the cost of each kind of message before the invoice arrives.

Channel rules at a glance

Timeline

These are typical planning ranges for a single data source and one or two channels. Larger groups with several systems take longer.

What you provide

• Access to, or a regular export of, customer and order or booking records with dates
• Evidence of how each contact gave consent, or a decision on which contacts will be excluded
• A named person who approves message wording and another who receives warm replies
• Access to your email sending domain and your WhatsApp Business account
• A clear rule on whether any offer can be mentioned, and who decides it

What this does not cover

The service does not set discounts, prices or offers, and it does not decide whether a particular customer should be contacted at all. It cannot tell why someone stopped buying. A customer who left after a bad experience, moved away, or no longer needs the product may still receive a message unless your rules exclude them, so the lapse rules need human review before launch. The service also does not write your privacy notice, give legal advice on consent, guarantee any return rate, make cold outreach to people who never opted in, or take over complaints that need judgement. Interested replies are handed to a person, and that person still does the selling.

Is it worth doing?

A 2026 benchmark compilation by Eightx, drawing on Klaviyo data, puts the conversion of a well-run email win-back flow at 2 to 5 percent of lapsed recipients. That range is for online retail and will differ for services. The case for automation is not the reply rate alone. It is that the sequence runs on time for every lapsed customer, without a staff member needing to remember who to contact.