Yes. Your job software (Housecall Pro, Jobber, ServiceTitan and similar) can already send scheduled service reminders; AI adds messages that mention each customer's actual system and due date, sorts the replies into bookings, questions, new faults and cancellations, and flags members whose plan is about to lapse. Clean due dates come first.
That order matters because reminders fail on data, not wording. A typical export of plan members has customers with no last-service date, landlords listed once per property with the same phone number, people who moved house two years ago, and visits already booked that the reminder system doesn't know about. Send a beautifully written message to that list and you get complaints. Spend the first hour on the list.
The fields every reminder depends on
Whatever tool sends the messages, it needs these fields per plan member. If one is missing, the reminder is either wrong or generic.
| Field | Why the reminder needs it | Example value |
|---|---|---|
| Contact and preferred channel | Sends by the route the customer chose | Text, mobile number |
| Plan type and renewal date | Separates visit reminders from renewal reminders | Heating plan, renews 14 Nov |
| Equipment type and age | Makes the message specific | Gas combi boiler, installed 2014 |
| Last service date | Sets the next due date | 2 Oct last year |
| Next due date | Triggers the reminder sequence | 2 Oct this year |
| Visit already booked? | Stops reminders to people who've booked | Yes, 29 Sep |
| Owner or tenant | Sends to the person who can grant access | Tenant; landlord pays |
| Access notes | Carried into the booking | Key safe; dog in garden |
Cleaning the list with AI before the first send
Export the plan members to a spreadsheet and ask an AI model to find the problems, not fix them silently. Use a business plan whose terms exclude training on your data, or strip names and phone numbers first and keep just a customer ID.
This sheet lists our service plan members (one row per property).
Find and list, by customer ID:
1. Rows with no last-service date or no next-due date
2. Next-due dates earlier than the last-service date
3. Possible duplicates (same phone or email, different rows)
4. Plans marked active whose renewal date has passed
5. Next-due dates more than 14 months after the last service
Don't change the data. Give a table per problem with the reason.An illustrative result on a 600-member list: 14 members with no last-service date, 9 possible duplicates (mostly landlords with several properties, which are legitimate and need one contact across many rows), 6 rows where the next-due date came before the last service because the dates had been typed the wrong way round, and 3 plans still marked active after their renewal date. An office manager fixed all of that in about ninety minutes. Without the check, those 32 members would have received a wrong message, and the landlords would have received nine each. Cleaning up a messy CRM with AI covers this kind of job in more depth.
A reminder calendar built around the two seasons
Heating and cooling plans have a natural rhythm: heating checks before the cold months, cooling checks before the warm ones, and for heat pumps often both. Build the reminder sequence around the season, not just the anniversary of the last visit, or every customer who joined in December gets a service reminder in December, when your engineers are busiest with breakdowns.
| When | Message | Channel |
|---|---|---|
| 6 weeks before the season | Invitation to book, with a link and three suggested weeks | |
| 3 weeks before | Reminder with the next available slots | Text |
| 1 week after the due date | Overdue nudge: short, friendly, one link | Text |
| 3 weeks after the due date | Phone call from the office for anyone still unbooked | Phone |
| 30 days before plan renewal | Renewal notice with what's included and the price | |
| Day after a lapse | One message offering to restart the plan, then stop |
Three reminders and a phone call per due date is plenty. More than that and opt-outs climb.
Reminders that mention the customer's actual system
Most service reminders read like this: "It's time for your annual service! Book now." Compare a version written from the fields above:
Your combi boiler's annual service is due on 2 October. It's now eleven years old, so the yearly check is worth keeping up. Pick a slot here: [link]. The visit is included in your heating plan, so there's nothing extra to pay. Reply STOP to opt out of plan reminders.
AI writes these well from a template and the row's fields. A prompt that keeps it honest:
Write a service reminder text (under 320 characters) for each row.
Use only these fields: first name, equipment type, equipment age,
due date, plan name, booking link.
Rules: friendly and plain; no discounts or offers; no scare wording
about breakdowns; say the visit is included in the plan; end with
"Reply STOP to opt out of plan reminders."Illustrative outputs for three different members:
- Heat pump, two years old: "Hi [name], your air-source heat pump's annual check is due on 18 October. It's included in your plan. If your warranty terms ask for yearly servicing, this visit covers it. Choose a time: [link]. Reply STOP to opt out of plan reminders."
- Office air-conditioning, three split units: "Hi, the spring service for your three air-conditioning units is due in April, ahead of the warm months. It's included in your cooling plan. Book a visit outside office hours here: [link]. Reply STOP to opt out of plan reminders."
- Boiler, age unknown: the AI wrote "your boiler, which is getting on a bit". That's an invented detail from a blank field. Add a rule: "If a field is blank, leave that detail out."
Check a sample of twenty before any batch goes out. The third example is the typical mistake, and it only takes one to annoy a customer.
Sorting the replies without reading every one
A reminder campaign produces replies, and replies are where the value hides. Some are bookings. Some are new faults: "can you also look at the radiator upstairs, it's cold at the top". Those are repair jobs, and they shouldn't sit in a queue for three days. AI is good at sorting them. An illustrative batch:
| Reply | AI category | Action |
|---|---|---|
| "Tuesday afternoons are best" | Booking request | Propose a Tuesday slot; office confirms |
| "Is the flue check included?" | Plan question | Answer from plan terms; human checks |
| "Boiler's been losing pressure lately" | New fault | Flag to office today as a repair lead |
| "We've sold the house" | Moved | Stop reminders; ask for new owner's details if they're happy to share |
| "Stop sending these" | Opt-out | Opt out immediately; no reply needed |
If your job software's inbox can't sort replies, a small automation can. In Zapier, an "AI by Zapier" step uses 1, 3 or 5 tasks per run depending on the model tier chosen, and the Professional plan starts at $29.99 a month billed monthly for 750 tasks, which covers a few hundred replies a month with room to spare. Keep a person approving anything that books or cancels.
The renewal notice: the message that keeps the plan alive
Visit reminders fill the diary; renewal notices keep the recurring revenue. They deserve more care than they usually get, because a renewal that arrives as a bare payment notice invites the question "what am I actually paying for?"
Before: "Your Heating Care plan renews on 14 November. $14.50 will be taken monthly."
After, drafted by AI from the member's visit history:
Your heating plan renews on 14 November at the same price, $14.50 a month. This year it covered your boiler service in October and the call-out in February when the pressure dropped, which would otherwise have been charged separately. Next year's service will be due around October again. Nothing to do if you'd like to carry on; reply or call if you want to change or cancel.
The second version reminds the customer of value they've already had, using facts from their own record. Give the AI the visit and job history for the past year and one rule: mention only visits that appear in the record. The illustrative price here is a placeholder for your own plan price; the AI should read it from the plan field, never work it out.
Send renewal notices to members whose plan saw no visits at all with extra care. A member who paid for a year and never booked is the most likely to cancel, and the right message is an offer to book the missed visit before renewal, not a payment notice.
Landlords with several properties: one message, not nine
The list check found landlords with several plan properties. Sending each one a separate reminder per property is the quickest way to an opt-out. Group them: ask the AI to combine all due visits for one contact into a single message with a short table, and to book them as a run where the properties are close together. An illustrative version:
Hello, three of your properties have heating services due in the next six weeks: the ground-floor flat (due 4 Oct), the terrace on the hill (due 11 Oct) and the two-bed house by the park (due 2 Nov). We can do the first two on the same morning. Could you confirm tenant contact details, or should we arrange access with you? All three visits are included in your plans.
Landlords often prefer to hand over tenant numbers so you arrange access directly. Record that as a field, so next year's reminders go to the right person without asking again.
Winning back members whose plan lapsed
A lapsed member is a warm customer: they already trusted you with their system. One well-timed message a few months after the lapse, before the season, often brings some back. Pick them with a simple filter (plan lapsed in the last 12 months, equipment over eight years old, no complaint on the record) and let the AI draft from their history.
Before: "We miss you! Rejoin our maintenance plan today."
After: "Your heating plan ended in March, so your boiler's yearly check isn't booked this autumn. It was last serviced 14 months ago. You can book a one-off service, or restart the plan, which covers the service and priority call-outs. Either way, here's the link to pick a slot: [link]."
The second offers a choice without pressure and uses a fact from the record. Send it once. A lapsed customer who doesn't reply should be left alone until the next season.
Send to fifty before you send to six hundred
Before the first full send, run a pilot: fifty members, chosen across plan types and equipment. Check the messages arrive, read every reply, and look at what went wrong. In an illustrative pilot, three mobile numbers bounced, one customer replied that 7:30 in the morning was too early (the sending window was wrong), and one reminder went to a member who'd booked by phone the day before. Three fixes, made before they could reach 550 more people.
Setting it up in the software you probably already use
Housecall Pro: its Service Plans feature schedules plan visits automatically and can send scheduling reminders and renewal reminders, with renewals generated from your renewal settings. Set the reminder timing to your season calendar rather than the default anniversary.
Jobber: set plan visits up as recurring jobs. Visit reminders go by text and email with separate templates you can customise, on a schedule you choose. Jobber's built-in AI chat can draft the template wording, but check which plan includes it for your account.
ServiceTitan: memberships attach recurring services to each service location, which generate recurring service events for booking, and you can build SMS campaigns aimed at recurring-service customers.
In all three, the reminder engine is rule-based. The AI part is the wording, the reply sorting and the list checks around it. For how their AI features compare more broadly, see Jobber, Housecall Pro and ServiceTitan AI features compared.
One heating season with 600 plan members
Here is how the pieces add up for an illustrative heating-and-cooling firm with 600 heating-plan members, eight engineers and one office manager. Before, the office sent one generic email a month before each due date and phoned whoever hadn't booked, when there was time.
- Week 0: the list check finds 32 problem records; fixed in ninety minutes.
- Weeks 1-2: the season invitation goes to 600 members. 230 book through the link.
- Weeks 3-4: the text reminder brings in another 150. Replies: 88, sorted by AI into 41 booking requests, 19 plan questions, 16 new faults, 7 moves and 5 opt-outs.
- Week 5: the 16 new-fault replies become 11 repair visits, work that the old generic email never surfaced.
- Weeks 6-8: overdue nudges bring in 90 more. The office phones the remaining 130, prioritising older equipment and members up for renewal soon.
The office manager's estimate of time spent fell from roughly two days a month in season to half a day, most of it phone calls to the overdue list. The figures are illustrative, but the pattern is typical: the personalised wording helps, while the reply sorting and the clean list do most of the work.
Mistakes that turn reminders into complaints
- Reminding people who've already booked. If bookings made by phone don't update the "visit booked" field, the customer gets "please book" two days after booking. Sync that field before every send.
- Messaging tenants about the landlord's plan without context. Say whose plan it is and that access is needed.
- Carrying on after a sad reply. A reply saying the customer has died should stop every message to that record at once and go to a person, not an automated condolence.
- Sending at the wrong hour. Set a sending window (for example 9:00 to 19:00) so texts don't arrive at 6 a.m.
- AI inventing an offer. "10% off if you book this week" appears in drafts more often than you'd think. The "no discounts" rule and a sample check stop it.
What to measure after the first season
Track five numbers against last season: the share of members serviced by their due date, the plan renewal rate, repair jobs raised from reminder replies, opt-outs per thousand messages, and office hours spent. If opt-outs rise while bookings don't, you're sending too many messages. If repair leads appear, make sure they're reaching an engineer quickly. The reply-handling habits here work for other recall types too; service reminders for garage customers and cutting no-shows with AI reminders show the same approach in other businesses, and what missed calls cost a heating and cooling firm helps you value the extra calls a campaign produces.
Service plan reminder questions from heating and cooling firms
Do service reminders count as marketing messages?
A reminder about a visit the customer has already paid for under a plan is usually treated differently from a sales promotion, but the line varies with the law where your customers live and with what the message says. Keep plan reminders strictly about the plan, include an easy opt-out, and ask your data-protection adviser before adding offers or upsells to them.
Should reminders go by text or email?
Use the channel the customer chose when they joined the plan, and record it as a field. Many firms send the first invitation by email with a booking link, the follow-up by text because it is read faster, and save phone calls for the overdue list. Whatever mix you pick, cap the number of messages per due date so reminders don't become a nuisance.
Can AI book the visit on its own when a customer replies?
It can, if your job software exposes booking slots to an automation or AI receptionist, but start by having AI propose the slot and a person confirm it. Service visits depend on engineer skills, parts and travel, and an automated booking in the wrong slot creates a worse experience than a reply a few hours later. Move to automatic booking once the proposals are reliably right.
Further reads
- Setting Up an AI Answering Service That Never Misses an Emergency — Make sure the new-fault replies your reminders surface reach an engineer fast.
- AI vs Human Answering Service for HVAC Firms — Who should answer the calls your reminder campaign generates.
- Getting More Reviews After Every Job With AI Follow-Ups — Ask for a review after each plan visit while it's fresh.
- Chasing Unpaid Invoices With AI: Polite Reminders for Trades — Polite chasing for plan payments that fail or go overdue.
- Is Your Business Data Ready for AI? A Clean-Up Checklist — A wider clean-up checklist for customer data before automating.
- When Does Zapier Get Too Expensive? Finding the Tipping Point — When reply-sorting automations start costing more than they save.
- How a Plumber Can Save Five Hours a Week on Admin With AI — An illustrative time log for a sole plumber, six changes that add up to five hours a week, a four-week rollout and what it costs.
- Following Up Unapproved Repair Estimates With AI — A step-by-step way for repair trades to chase open estimates with AI-drafted messages that explain the fault honestly and allow partial approval.
- Can AI Handle Vaccination and Check-Up Reminders for a Vet? — What AI can and can't do for a vet practice's vaccination and check-up reminders, with a message sequence, a reply-triage prompt and worked numbers.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: Housecall Pro help pages (service plans, renewals); Jobber help pages (visit reminders, automated texts and emails); ServiceTitan help pages (memberships, recurring service events, SMS campaigns); Zapier pricing page.