Your Next 20 Jobs Are Already Sitting in Your Customer List

Every shop hits the same wall a couple times a year. The phone slows down, the schedule gets holes in it, and the first instinct is to spend money: boost the ads, buy more leads, drop the price on tune-up specials for strangers.

Meanwhile there's a list of people who already trust you, already paid you, and already own equipment that needs service again. They're sitting in your Jobber or Housecall Pro account right now. Nobody calls them because nobody can see them.

Marketing folks have repeated the same rough numbers for decades: winning a new customer costs somewhere around five times what it costs to sell to an existing one, and repeat customers say yes far more often than cold leads do. You don't need the exact figures to know it's true in your own shop. The Hendersons, who had you replace their water heater two years ago, will pick up when you call. A stranger from a lead service might not even remember filling out the form.

Why doesn't anyone actually work the customer list?

Because pulling that list is a project, and slow season is exactly when morale for projects is lowest.

Think about what "call our past customers" actually requires in most shops:

1. Export the customer list, or scroll it screen by screen
2. Figure out who had what work done, and when
3. Guess which ones are due again: the AC tune-up from last April, the gutter cleaning from last fall, the quarterly pest visit that quietly stopped
4. Cross-check that they don't already have something booked
5. Sort out who's worth calling first

That's an afternoon in spreadsheets, minimum. So it becomes one of those things everyone agrees would be smart and nobody ever does. The list keeps growing, and every year a slice of it quietly drifts to whoever mailed them a postcard.

What happens when you can just ask?

This is the part that changed. When your Jobber or Housecall Pro account is connected to ChatGPT or Claude through Inrconnect, that whole afternoon collapses into a question typed in plain English:

"Which customers had a job with us last spring but haven't booked anything since?"

The answer comes from your real account. Real names, real service history, real dates. Not a guess, not a template. Then you narrow it:

"Of those, which ones spent over $500 with us? List them with their last job and what they paid."

Now you're looking at maybe thirty names ranked by what they're worth to you. The customer who paid $3,850 for a full duct cleaning and a new thermostat last year goes to the top of the call sheet. The one-time $89 filter swap can get the email blast instead.

One more sentence gets you the outreach itself:

"Draft a short, friendly check-in for each of the top ten, mentioning the work we did for them."

Because the AI can see the actual job, the message says "it's been about a year since we serviced your furnace on Maple Street," not "Dear Valued Customer." That difference is why these messages get answered.

Which customers should you call first?

A simple order that works for most trades:

Maintenance that's overdue. HVAC tune-ups, backflow tests, dryer vent cleanings, seasonal lawn treatments. These customers expect to hear from you. Ask which ones are past their usual interval.

Big jobs with a natural follow-up. A roof you patched is a roof that eventually needs replacing. A panel you inspected is a panel with a quote waiting to happen. Ask for customers whose last job suggests a next one.

Quiet regulars. The every-quarter pest control or cleaning customer who skipped a cycle. They didn't fire you. They just fell off the schedule, and a friendly nudge usually puts them back on it.

Old declined quotes. Someone who said no to a $2,400 repair eighteen months ago is living with a problem that hasn't fixed itself. Circumstances change. So do budgets.

You could hunt down each of these groups with reports and filters. Or you could ask four questions and be dialing by lunch.

Is it safe to let an AI look at your customer data?

Fair question, and the answer should be on your terms, not the software's.

Inrconnect is a direct connection between the assistant and your account, not a robot clicking around your screen. You decide what it's allowed to do. Most owners start in read-only mode: the AI can look things up and answer questions, but it cannot create, edit, or send anything. Building a call list is a read-only job start to finish, so you can run this entire play without ever granting write access. Every question it asks of your account shows up in an activity log, so you can see exactly what was looked at and when.

The slow-week play

Next time the schedule looks thin, skip the ad budget for one week and run this instead:

1. Ask who hasn't been back since their last job
2. Rank them by what they've spent with you
3. Have the AI draft personal check-ins based on their real history
4. You review, tweak, and hit send, then start calling the top of the list

If two of those thirty names turn into booked work, that's likely more than the ad spend would have returned, and those customers came back warmer than any lead ever arrives.

Connect your Jobber or Housecall Pro account to Inrconnect, leave it in look-but-don't-touch mode, and start with one question: "Which customers haven't we seen in a year?" The Jobber tools page and Housecall Pro tools page show everything else your assistant can dig up once it can see your account.