argentDIGITAL
← Back to Blog

Automation

A Days-Since-Last-Job Threshold Triggers Win-Back, Not Memory

Win-back campaigns that depend on someone remembering to run them don't run — here's how to automate the trigger, the sequence, and the reply routing.

Aircraft technician working inside the open engine cowling of a small plane in a hangar

Key takeaways

  • A win-back campaign only works if it's triggered automatically by a CRM date field, not run manually when someone remembers to export a list.
  • Lapsed should be defined by a specific days-since-last-job number pulled from your own job history, not a general sense that it's been a while.
  • The strongest win-back sequences run three to five touches over two to three weeks and switch channels — email, text, and a call task — to match how the customer already communicates.
  • Every win-back sequence needs a reply-routing rule before it launches, so a positive reply reaches a rep the same business day instead of going unanswered.
  • A residential brokerage saw a 28% increase in lead-to-tour rate from automated follow-up on past and lapsed leads, driven by consistency of contact rather than better copy.

A win-back campaign that depends on someone remembering to run it will not run. The businesses that actually recover lapsed customers treat win-back as a trigger-based workflow inside their CRM: a contact crosses a days-since-last-job threshold, a sequence fires automatically, and a human only gets involved when a reply needs a real answer. Below is the structure that makes that reliable, and the mistakes that quietly kill reply rates.

Define "lapsed" with a number, not a feeling

A past customer becomes a win-back target the moment their time-since-last-service crosses a threshold specific to your service cycle, not a vague sense that "it's been a while." An HVAC contractor's threshold might be 10–12 months since a tune-up; a pool builder's might be 18 months since a repair visit; a roofing contractor's might be 3–5 years tied to typical roof lifespan.

Pull this from your own job history: look at the gap between a customer's first and second service call across your last 50–100 jobs, and set the threshold just before the point where repeat rates drop off. That number is the trigger condition for everything downstream — get it wrong and you're either emailing people too early (annoying) or too late (they've already called a competitor).

Automate the trigger, not just the message

The mechanical failure point in most win-back attempts isn't bad copy — it's that nobody checks the calendar. Automating the message without automating the trigger just moves the bottleneck; you still need someone to remember to export a list. The fix is connecting your CRM's date fields to an automation layer that runs the check daily and enrolls contacts without a person touching it.

This is the same automation logic behind speed-to-lead response and renewal reminders — a date or status field changes, a sequence starts, a rep gets notified only when action is required. For a two-person sales team, this is the difference between win-back being a "someday" project and a channel that produces booked jobs every month.

A minimum viable win-back trigger

Days since last completed job ≥ threshold, no open quote, no job scheduled in the next 30 days, valid email or phone on file. Four conditions, checked once a day, enrolling contacts into a sequence automatically.

Sequence structure for service businesses

The sequence itself should be short — three to five touches over 2–3 weeks, not a 10-email drip. Touch one is a direct, specific reminder ("It's been about a year since your last tune-up — want to get on the schedule before the busy season?"). Touch two, sent 5–7 days later if there's no reply, adds a reason to act now: seasonal timing, a maintenance checklist, a review-driven prompt. Touch three is the close: a direct ask to book, sent by phone or text if you have it, not just email.

Each touch should route through the channel the customer actually used before — if they booked by phone, a text gets opened faster than an email. Mixing channels across the sequence (email, then text, then a call task assigned to a rep) consistently outperforms single-channel sequences because it matches how the customer actually communicates, not how your CRM defaults to sending.

Reply handling determines whether leads convert

An automated sequence that generates replies but routes them nowhere is worse than not running the campaign — the customer re-engaged, got no response, and now has a worse impression than before. Every win-back sequence needs a reply-routing rule before it launches: a positive reply creates a task for a specific rep within a set time window (same business day, not "sometime this week"), and a no-reply after the final touch moves the contact to a longer-interval nurture rather than dropping out of the system entirely.

This reply-routing step is the same mechanism as speed-to-lead response for new inquiries — the business doesn't care whether the lead is new or returning, it cares that someone replied and nothing happened for three days. Automated lead routing exists precisely to close that gap, whether the inquiry originated from a win-back text or a form fill on the website.

Segment by job value and churn reason before sending

A single generic win-back blast to every past customer wastes the highest-value segment on a message built for the average one. Split your list by two variables: job value (a customer whose last job was $8,000 deserves a different message than one whose last job was $200) and likely churn reason (moved away, used a different provider, just never got around to rebooking — each needs a different first line).

For the moved-away segment, don't send anything; it wastes a touch and skews your reply-rate metrics. For the "never got around to it" segment — usually your largest — the reminder-plus-urgency structure above works well. For the "used someone else" segment, a review-request or service-guarantee angle outperforms a generic reminder because it addresses the actual objection instead of ignoring it.

What results actually look like

A residential brokerage running automated follow-up sequences on past and lapsed leads saw a 28% increase in lead-to-tour rate compared to manual follow-up — the lift came from consistency of contact, not better copy. That's the realistic ceiling for what automation alone contributes: it doesn't invent demand, it captures demand that already exists in your past-customer list but currently leaks out because nobody follows up on schedule.

Ice rink technician resurfacing the ice of an empty community rink

Measure your own win-back performance the same way: reply rate by touch number, booked-job rate by segment, and revenue recovered per contact enrolled. If touch one consistently outperforms touches two and three, your list hygiene is good and your message is working — tighten the later touches. If reply rates are flat across all three touches, the trigger timing or segmentation is off before the copy is.

How do you know if a win-back campaign is worth building first?

Pull your customer list and count how many contacts have crossed your lapse threshold with no scheduled job and no open quote — if that number is in the hundreds, win-back is likely your highest-leverage automation project before anything else, because the list and the demand already exist. If it's a handful of names, your time is better spent on inbound lead routing or review requests, where volume is higher.

Run this calculation before building anything: multiply the lapsed-contact count by your average job value and by a conservative reactivation rate (3–5% is a reasonable planning number for a well-targeted sequence, not a promised outcome). If that projected revenue clears the cost of setting up the automation within a quarter, it's worth prioritizing.

CRM hygiene is the prerequisite, not an afterthought

None of the above works if your CRM's "last job date" field is inconsistently filled in, duplicate contacts split a customer's history across two records, or email addresses are five years stale. Win-back automation inherits whatever data quality exists in the CRM — a trigger built on a messy date field will misfire constantly, enrolling current customers into win-back sequences or missing lapsed ones entirely.

Before building the sequence, run a one-time cleanup pass: deduplicate contacts, backfill missing last-service dates from invoice history where possible, and standardize status fields so "completed," "closed," and "done" aren't three different values meaning the same thing. This is typically a few hours of work for a list under a few thousand contacts, and it determines whether the automation you build on top of it is trustworthy.

If you want a second set of eyes on whether your CRM and job history are clean enough to automate against, that's a reasonable thing to check in a free 30-minute audit — it's faster to diagnose from a screen share than to guess from a blog post. For a broader look at how this fits with lead routing, review requests and pipeline reporting, see our AI services or browse examples of completed client work.

Prefer it done for you? This playbook is our AI engine: see how we run it for clients →

Frequently asked questions.

How do I know if a past customer counts as lapsed?

Pull the gap between first and second service calls across your last 50-100 jobs and set your threshold just before repeat rates drop off. That number becomes the trigger condition for enrolling a contact in the win-back sequence, and it varies by service cycle — months for HVAC, years for roofing.

What's the right structure for a win-back email and text sequence?

Keep it short: three to five touches over two to three weeks, not a long drip. Open with a direct, specific reminder, follow with a reason to act now five to seven days later, and close with a direct ask to book sent through the channel the customer already uses.

Do I need a CRM to automate win-back campaigns?

Yes — the trigger has to be a date or status field inside your CRM that an automation layer checks daily, enrolling contacts without anyone exporting a list. Without that connection, you've only automated the message, not the workflow, and the bottleneck just moves to whoever forgets to run it.

How many lapsed customers justify building a win-back automation?

If your lapsed-contact count is in the hundreds, win-back is likely your highest-leverage automation project because the list and demand already exist. Multiply that count by average job value and a conservative 3-5% reactivation rate to see if the projected revenue clears the setup cost within a quarter.

WANT A HAND WITH THIS?

Tell us what you’re working on.

Let’s talk