Skip to content

Case study · Past-customer reactivation

Handyman case study: 10 jobs from past customers in 90 days

He had years of happy customers and a phone that rang while he was on a ladder. One thank-you text went to 575 of them, and every reply got an answer the same minute. Ten of them came back as paid jobs.

A one-crew handyman business in Southern California. Client name withheld for privacy. The figures are unchanged.

Text messageFrom the business's own number

Hi [first name], it's [owner] from [business]. Hope you've been doing well. As a thank-you for being a valued customer, we're offering 20% off eligible new work booked in the next 30 days. Is there anything around the house you've been wanting to get done? Just text me back here. Reply STOP to opt out.

Yes, actually. The back gate sticks and the patio stucco has a crack.

Happy to help. We can come by Thursday at 9 AM or Friday at 1 PM. Which works better?

Thursday.

You're booked for Thursday at 9 AM. See you then.

The first text as sent, with names removed. The replies are an example.

Results at 90 days

  • 10Paid jobsEvery one from a past customer who came back.
  • $10kNew work10 jobs at about $1,000 each.
  • 3.3×Return on fees$10,000 of work on $3,000 paid to us.
  • 575Past customers textedOne text each. No follow-ups.

Source: the client's job records and our campaign log, first 90 days.

What we found

Where the work was slipping away.

We look at how a business answers before we change anything. For a one-crew handyman, the same few things came up that we see across the trades.

  1. Calls went to voicemail during the workday.

    He is the crew. When he was cutting tile or pouring a walkway, the phone rang out, and a caller who reaches voicemail can just as easily call the next name on Google.

  2. Callbacks waited until evening.

    Every voicemail became a call to return after a long day on site.

  3. Scheduling lived in text threads.

    Visits were set by back-and-forth texts spread across a dozen conversations, with no single calendar holding the week.

  4. Hundreds of happy customers had gone quiet.

    He had years of past customers who had agreed to hear from him. None of them had heard from him since the job was done.

  5. Google wasn't working for him.

    There was no Google Business Profile showing his work, so a homeowner who searched had nothing to go on.

What we did

What we changed.

The quiet list was the biggest opportunity, but texting it only pays off if someone answers the replies. So the phone came first.

  1. Margot took the phone, on his own number.

    She answers in his business name, asks what the job is, where and how soon, and texts him a summary. Callers who need him are put through to his cell. See how Margot answers calls.

  2. Booking went straight to his calendar.

    Margot only offers times he is free, books the visit, texts the customer a confirmation and handles reschedules. See how booking works.

  3. We built the list carefully.

    575 past customers, and only people who had agreed to hear from the business. The texts went out from his own number under his own carrier texting registration.

  4. One text, in his words.

    A thank-you for being a customer, with 20% off eligible new work booked in the next 30 days. Enough of a reason to reply, without giving the work away. It ended with “Reply STOP to opt out.”

  5. Sent once. No chasing.

    No second text, no drip sequence, no reminders. Each person got one message.

  6. Every reply got an answer right away.

    Replies and call-backs went to Margot, who asked what they needed and booked the visit. No one waited until the evening for an answer.

  7. His Google profile got real proof.

    We set up his Google Business Profile with eight before-and-after pairs from his own jobs, his services and service area, and the number Margot answers. Review requests now go out by text after finished jobs.

The honest part

565 people didn't book. That's how lists work.

Ten jobs from 575 texts is about one in every 57 people. Most of a past-customer list doesn't need work this month, and no text changes that. If you expect half the list to call back, you'll call a good result a failure.

We also chose not to follow up. A second or third text might have booked a few more jobs. We'd rather a past customer remember one polite thank-you than a string of reminders from a business they liked.

What made the ten count was speed. A reply is a customer with the job in mind right now. If the answer comes that evening, the moment has often passed.

The results

The numbers, plainly.

Here is what the first 90 days added up to, from his job records and our campaign log.

Texted once (565) Became a paid job (10)Dot positions are illustrative. Counts are from the campaign log and job records.
  • Paid jobs from past customers: 10, from one text each to 575 people
  • New work booked: about $10,000, at an average of about $1,000 a job
  • Fees paid to ApexScale AI over the same 90 days: $3,000
  • Return on fees: 3.3× ($10,000 ÷ $3,000)
  • Calls to his business number answered by Margot: every one
  • Appointments booked onto his calendar and confirmed by text: every one

Source: the client's job records and our campaign log, first 90 days.

What it's worth

Three jobs covered us. Seven were his.

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
Jobs 1–3: about $3,000, covered the fees Jobs 4–10: about $7,000, his to keep

At about $1,000 a job, the first three jobs paid for three months of ApexScale AI. The other seven, about $7,000 of work, were his.

These were customers who already knew his work, so none of the ten came with an ad bill or a lead fee.

New work booked
$10,000
Paid to ApexScale AI, 90 days
$3,000
Return on fees
3.3×

This is job revenue from his records, not profit: materials and his time on site come out of it. Jobs are counted at an average of about $1,000.

What this comes down to

Check your own business in three questions.

He didn't need new customers to get new work. He needed a reason to reach the ones he had, and someone to answer when they wrote back.

If you've been in business a few years, check three things before you buy another lead:

  • Do you have a list of past customers who agreed to hear from you?
  • When one of them texts back at 2 PM while you're on a job, who answers?
  • Does that answer end with a time on your calendar, or with “let me check and get back to you”?

Where this lives in Margot

Full refund if you cancel within 14 days of Margot answering your first call.

Figures cover the client's first 90 days: 575 past customers were texted once each, and 10 became paid jobs at an average of about $1,000. Texts went only to customers who had agreed to hear from the business. Reactivation texts and booking are part of Front Desk + Growth. These are one client's results over their first 90 days with ApexScale AI. Your results depend on your trade, your customers and your prices, and we don't guarantee jobs or revenue.

More case studies

Same client, other angles.

All case studies

Full refund if you cancel within 14 days of Margot answering your first call.