I want to tell you about a list that most trades business owners have sitting in their CRM right now. It’s a list of past customers who called once, booked a job, paid the invoice, and then went completely quiet. No second call. No referral. No response to the last email you sent eight months ago. Using AI to follow up with dormant customers is one of the most practical things I’ve seen work for small service businesses, and I want to walk you through a real example of how it actually plays out.
In a trades context, the dormant customer threshold is usually somewhere between 90 and 180 days of zero contact. An HVAC customer who got a tune-up last spring and hasn’t called since fall. A plumber’s client who had a water heater replaced two years ago. A roofer’s customer who got an inspection done after a storm and then disappeared.
These aren’t bad customers. They didn’t leave angry. Research from Bain and Company shows the most common reason customers go quiet is perceived indifference, meaning they just figured you forgot about them. No follow-up, no check-in, no reminder that you exist. So they moved on.
A scan of 72 small business CRM systems found that, on average, 58% of contacts hadn’t engaged in over 120 days. That’s not a small problem. That’s most of your list.
Honestly? Because nobody has time. You’re running crews, dealing with dispatch, handling callbacks on active jobs. Going back through a spreadsheet of people who haven’t called in a year feels like a low-priority task that never makes it to the top of the pile.
And even if you wanted to, what do you say? A generic email blast gets ignored. A personal phone call from the owner to every lapsed customer isn’t scalable. So the list just sits there, getting older, while you spend money chasing brand new leads. Harvard Business Review puts the cost of acquiring a new customer at five to 25 times more than retaining an existing one. That dormant list has real value, and most owners never touch it.
A gym owner had a list of lapsed members, people who had cancelled or simply stopped showing up. He considered that list dead. No point chasing people who already said no, right? The list had hundreds of contacts, most of them untouched for six months or more.
He ran an AI reactivation campaign through an automated system that could segment the list by how long ago each person had last engaged, craft personalized outreach based on their history, and reach out across multiple channels including AI voice calls.
First, the system used purchase history segmentation to divide the list. Recently lapsed contacts got a different message than people who had been gone over a year. The re-engagement timing was built around what research shows works: lead with value and a genuine check-in on the first touch, not a discount.
Second, the AI sent personalized outreach via SMS and email. Not a blast that says “We miss you!” with a coupon code. A message that referenced the customer’s actual history and offered something relevant to where they were in their service cycle.
Third, for contacts who didn’t respond to text or email, the system made AI voice calls. Voice has 8 to 12 times higher engagement than email. The AI identified itself as an AI, kept the call short, and offered to book a follow-up if the person was interested. Anyone who asked to stop hearing from the business was immediately removed.
The campaign generated a 7% positive response rate and over $83,000 in revenue from a list the owner had written off completely. That’s the FetchFunnel case study, and it’s one I come back to often because it illustrates the point clearly: the leads were already paid for. The acquisition cost was zero. The AI just did the follow-up that nobody had gotten around to doing.
A similar campaign for a multi-location dental group reactivated 14% of a cold list, with 60% of those contacts booking within three weeks. The AI reactivation benchmark across industries is 15 to 30% for recently lapsed customers and 8 to 15% for long-dormant ones. Compare that to 3 to 8% for a generic email campaign.
It’s not that AI is more charming than a person. It’s that AI is consistent and fast in ways that humans can’t be at scale.
A good AI reactivation system analyzes customer history to figure out the best time to reach out, the channel most likely to get a response, and the tone that fits the individual. A customer who always responded to texts gets a text first. A customer who picked up the phone last time gets a call. That kind of lead prioritization is nearly impossible to do manually across a list of 500 people.
Timing matters too. Churn prevention research consistently shows that the window between 60 and 120 days of inactivity is when a customer is most reachable. Wait until 18 months and you’re working a much harder problem. AI can flag those contacts automatically and trigger outreach before the window closes.
Leading with a discount on the first touch. I see this constantly. The email goes out, it says “Come back, here’s 20% off,” and the customer either ignores it or takes the deal and disappears again. Worse, you’ve now trained your lapsed base that waiting pays off.
A well-built AI campaign holds the incentive back. Touches one and two are about value and relevance. “Your furnace is coming up on its annual service window” hits differently than “Come back for a discount.” The incentive, if you use one at all, shows up on touch three or four, after the customer has already re-engaged mentally.
This is where having a system built correctly from the start matters. If you’re manually sending emails and guessing at the sequence, you’ll almost always default to the discount too early.
Pull a segment from your CRM of contacts who haven’t booked in 90 to 180 days. That’s your small batch testing group. You want enough contacts to get meaningful data, but not so many that a bad message goes out to your whole list at once.
Set up a three-touch sequence: an SMS on day one, an email on day three, and a voice call on day seven for anyone who hasn’t responded. Keep the messaging specific to the service they last used. Listen to call recordings from the AI outreach and refine the script based on what’s working.
Even a 4% reactivation rate on 100 contacts is four booked jobs you didn’t have before, with zero acquisition cost attached. That’s a meaningful number for most trades businesses in Snohomish County where every booked job counts.
The obvious metric is reactivation rate, meaning what percentage of your lapsed contacts booked a job. But the more important number is the 90-day retention rate: how many of those reactivated customers are still active three months later. A customer who books once and disappears again isn’t really retained.
Watch your email open rate too. Below 8% means your messaging or your list quality needs work. A healthy reactivation campaign runs 12 to 15% open rates or better. And track customer lifetime value on reactivated contacts over time. That’s where the real return shows up.
If you have a CRM with more than a few dozen past customers and you’re not systematically following up with the ones who’ve gone quiet, you’re leaving real money on the table. This isn’t theoretical. The benchmarks are there. The case studies are there. The math on customer retention cost versus acquisition cost is not close.
The challenge is that building this correctly takes time and the right platform. You need CRM segmentation, automated SMS and email sequences, AI voice call capability, and the logic to tie it all together. Piecing that together from five different tools is a project, not an afternoon.
That’s exactly what I build for local service businesses through KIKK Ventures. One platform, set up and managed for you, so you’re not spending your evenings figuring out integrations. If you’re curious how HVAC contractors in Snohomish County are already using AI to stay ahead of leads, this post on answering leads faster is worth a read.
And if you want to know how visible your business actually is to AI-powered search before you start any campaign, run a quick check at aivisibilitycheck.online. It takes about two minutes and gives you a real starting point.