Key takeaways
- Your past customer list is usually the cheapest source of booked work you have.
- Segment by what they bought and when, then give each group a specific reason to call.
- Short, personal texts often beat polished newsletters for reactivation.
- Time campaigns to your slow weeks, not to the calendar.
- Only message customers who agreed to hear from you, and honor every opt-out.
Every business has slow weeks. The techs are on the schedule, the trucks are paid for, and the phone is quiet. Many owners respond by spending more on ads.
There's usually a faster, cheaper option sitting in your software: hundreds or thousands of people who already hired you, liked the work, and simply haven't thought about you lately. This post shows how to use AI to reactivate past customers, including how to segment your list and what to actually say.
Why should you reactivate past customers before buying more leads?
Because past customers already know you, trust you, and have your number saved, so it usually costs less to bring them back than to win someone new. You aren't convincing them you're legitimate. You're just reminding them you exist and giving them a reason to book.
A new lead from paid ads or a lead marketplace often costs real money before you even talk to them, and they're probably calling three other companies. A past customer you message directly costs you a small per-message fee and some setup time.
Past customers are also easier to schedule. You can target the exact week you need to fill, because you choose when to send.
What does AI actually do in a reactivation campaign?
AI handles the tedious parts: cleaning and sorting your customer list, writing personalized messages, sending them at the right time, and answering simple replies. Your team handles bookings and real conversations.
Without AI, reactivation usually looks like this: someone exports a spreadsheet, gives up halfway through cleaning it, and sends one generic "We miss you! 10% off!" email blast. It gets ignored.
With AI, the workflow looks like this:
- Pull customer history from your CRM, field service software, or invoicing tool.
- Clean the list: merge duplicates, flag bad numbers, and remove anyone who opted out.
- Segment each customer by what they bought, when, and what you recommended.
- Draft messages that mention their actual past job and a specific reason to book now.
- Send in waves sized to your open capacity, so you don't flood a week you can't serve.
- Handle replies: answer "what times do you have?" and route anything complex to a person.
- Track results: who replied, who booked, and how much revenue each segment produced.
Steps 2 through 6 are where AI saves the time. It's a classic workflow automation job: repetitive, rules-based, and painful to do by hand.
How should you segment past customers?
Segment by what they bought, when they last bought, and what's likely due next. A customer who had a furnace tune-up 13 months ago needs a different message than one who got a quote for a new roof and never booked.
Here's a segmentation table you can adapt. The examples lean toward home services, but the logic works for auto shops, clinics, cleaning companies, and more.
| Segment | Who's in it | Reason to reach out | Example offer |
|---|---|---|---|
| Due for maintenance | Last service 10–14 months ago | Routine service is due | Book a tune-up or annual service |
| Recommended but not done | Tech recommended work that wasn't booked | Follow up on a known issue | Revisit the recommended repair |
| One-time customers | Single job 12+ months ago | Remind them you're still here | Seasonal check or related service |
| Lapsed members | Maintenance plan expired | Renewal | Renew the plan before peak season |
| Quoted, never booked | Quote sent, no decision | Circumstances may have changed | Updated quote or smaller option |
| High-value regulars | Top customers by lifetime spend | Thank and prioritize | First pick of open slots |
The "recommended but not done" segment is often the most valuable. Your tech already diagnosed the problem. The customer already heard about it. They may simply have needed time or money. A friendly reminder months later often lands at the right moment.
If your data is messy (and almost everyone's is), AI can read job notes and invoices to infer segments. For example, it can scan tech notes for phrases like "recommend replacing" or "monitor next visit" and tag those customers automatically. That's the kind of document processing that would take a person days.
What should reactivation messages say?
Keep them short, personal, and specific: who you are, what you did for them, and one clear reason to book now. Sound like a person at your company, not a marketing department.
Here are sample messages for each segment. Swap in your details.
Due for maintenance (text):
Hi Maria, it's Jake at [Company]. We serviced your AC last June, so it's about time for its yearly check before summer. Want me to grab you a spot next week? Reply YES and I'll send times.
Recommended but not done (text):
Hi Tom, it's Sarah from [Company]. Last time you were in, Mike noted the front brakes were getting thin. Want us to take a quick look? We have openings Tuesday and Wednesday.
One-time customer (email):
Subject: Still in the neighborhood, [First name]
Hi [First name],
We cleaned your gutters last fall and wanted to check in. With the leaves coming down again, we're scheduling fall cleanings on [your street area] over the next two weeks. If you'd like us to swing by, just reply and we'll find a time that works.
Thanks for trusting us last year, [Name], [Company]
Lapsed maintenance plan (text):
Hi Dana, your pest control plan with [Company] ended in March. Want to restart it before spring activity picks up? Reply and I'll send the details. No pressure either way.
High-value regulars (text):
Hi Mr. Patel, Lisa at [Company] here. We have a few openings next week and wanted to offer them to our regulars first. Want one?
Notice what's missing: big discounts, all caps, exclamation points, "We miss you!" Discounts can help in slow weeks, but a specific reason to book often works better than a generic coupon, and it doesn't train your customers to wait for sales.
When should you send reactivation campaigns?
Send them two to three weeks before the slow period you want to fill, in small waves you can actually serve. Reactivation is a capacity tool, so tie it to your schedule, not to an arbitrary marketing calendar.
A practical approach:
- Look at your schedule three weeks out. Identify days with open capacity.
- Pick the segment most likely to book the kind of work you have room for.
- Send to a batch sized to fill the gap. If you need 10 jobs, and you assume (for example) one in ten recipients books, start with about 100 customers.
- Watch replies for 48 hours. Send the next wave only if you still have room.
Seasonal businesses can plan these ahead. An HVAC company might run maintenance campaigns in early spring and early fall. A pest control company might push perimeter treatments before peak season. A landscaper might sell fall cleanups in late September. If you run HVAC or pest control, your slow seasons are predictable enough to schedule a year of campaigns in advance.
How do you handle replies without drowning your office?
Let AI answer the simple replies instantly and hand off the rest with context. A good reactivation campaign creates a burst of replies, and that's exactly when your office gets overwhelmed.
Typical replies and how to route them:
- "Yes" or "What times do you have?" AI offers open slots from your calendar and books the job, or holds a slot for your dispatcher to confirm.
- "How much is it?" AI shares your standard published pricing if you have it, or says someone will confirm shortly.
- "We moved" or "Wrong number." AI updates the record and stops future messages.
- "Stop." Opt-out recorded immediately and permanently.
- Complaints or complicated questions. Routed straight to a person, with the customer's history attached.
This is where reactivation connects to the rest of your follow-up stack. The same system that answers reactivation replies can power quote follow-up and no-show reminders.
What rules should you follow when messaging past customers?
Only message people who agreed to hear from you, make opting out easy, and keep the frequency reasonable. Reactivation should feel like a helpful reminder, not spam.
A short checklist:
- Confirm how and when each phone number was collected, and whether the customer agreed to receive texts. Marketing texts generally require prior express written consent under TCPA rules.
- Register your business texting number for A2P 10DLC so carriers deliver your messages.
- Include clear opt-out instructions and honor them immediately.
- Use email for customers whose texting consent is unclear.
- Don't message the same customer more than a few times a year unless they ask for it.
- If you're in healthcare, keep protected health information out of messages and follow HIPAA rules.
None of this is legal advice. If you're unsure about consent on an older list, have your attorney review it before you send.
What should you do next?
Export your customer list this week and count how many people haven't booked in over a year. That number is your reactivation opportunity, and it's often bigger than owners expect.
If you want it done for you (list cleaned, segments built, messages written in your voice, replies handled, and results tracked), book a free AI audit. We'll look at your customer data and your slow weeks and show you what a reactivation system would look like for your business. You can see what our lead follow-up service includes, or explore what else AI can do for your business.
Free AI audit
Let us find the hours AI can save you.
We integrate AI into small and medium businesses — back office, phones, follow-up, custom tools and more. Done for you, built around the software you already use.
FAQ
Frequently asked questions
What is customer reactivation?
Customer reactivation is reaching out to people who bought from you before but haven't in a while, with a reason to come back. For a service business that might be a maintenance reminder, a seasonal check, or a follow-up on work you recommended last time. It is usually much cheaper than finding a brand-new customer.
How does AI help reactivate past customers?
AI can sort a messy customer list into useful groups, write personalized messages that reference each customer's past job, send them on a schedule, and handle replies like 'What times do you have next week?' Your team only steps in for real conversations and bookings that need a human.
How often should I contact past customers?
For most service businesses, a few well-timed, relevant messages per year is plenty, tied to seasons or service intervals. Sending generic blasts every week trains people to ignore you or opt out. Every message should have a clear, specific reason.
Is it legal to text my past customers?
It depends on how you collected their number and what consent they gave. Marketing texts generally require prior express written consent under TCPA rules, and your business texting number should be registered for A2P 10DLC. When in doubt, use email for older customers and ask your attorney to review your consent language.
What software do I need for a reactivation campaign?
Usually just your existing CRM, field service, or invoicing software plus a texting and email tool. Jobber, Housecall Pro, ServiceTitan, QuickBooks, Square, and HubSpot all hold customer history that can be exported or connected. The AI layer sits on top to segment, write, and respond.