Your Old Customer List Isn't Worth $600,000. Here's the Honest Math.

There is a pitch going around home services right now, and it goes like this: the list sitting in your CRM is a goldmine. A few thousand old customers and dead quotes, nobody has touched them in years, send them a text and book two hundred grand in ninety days.

The pages selling it are specific. One puts a 3,000 person database at a $2,500 average ticket and arrives at $225,000 to $600,000. Another says 2,000 to 5,000 dormant records "often" recovers $50,000 to $200,000 in 60 to 90 days. Others promise 18 to 32 percent of annual revenue at a ten to twenty times return.

Reactivation is real, it is one of the cheapest things a contractor can do, and we build it for clients. But those revenue numbers are wrong by close to an order of magnitude, for a reason you can catch with a calculator. And the campaign most of these pages describe is one a lot of contractors cannot legally send as written.

So we rebuilt the math with every assumption visible, then fetched the pages currently ranking for the searches a contractor would actually run and checked what they say about consent.

The Arithmetic Tell You Can Check Yourself

Take the biggest published claim: $600,000 from 3,000 records at a $2,500 ticket. Divide it out. $600,000 divided by $2,500 is 240 signed jobs. Out of 3,000 records, that is eight percent of the entire list putting a signature on a contract. The bottom of the same range, $225,000, is 90 jobs, or three percent.

Now look at the response rates published on those same pages: two to eight percent for email, five to fifteen percent for text, twelve to twenty-five percent for a multichannel sequence.

Those are the same numbers. The model takes the percentage of people who reply and treats it as the percentage who buy.

A reply is not a sale. A reply is "who is this," "STOP," "we sold the house," "maybe in the spring," and occasionally, genuinely, "yes, come take a look." Only the last one has any money in it.

The Honest Math on a 3,000 Record List

Here is the same campaign built back up with every step showing. This is a model, not a measurement, which is the point: swap your own numbers in and watch what happens. Start with 3,000 records and a $2,500 average ticket.

1. Contact details that still work: about 80 percent. Contractor databases accumulate typos, disconnected numbers, landlines nobody answers, and customers who moved. Call it 2,400 reachable.
2. Records you can legally text: call it 30 percent. This is the step nobody models, and we come back to it below. If you never ran a consent checkbox at intake, documented written consent for marketing texts covers a minority of your list. That is roughly 720 textable, leaving about 1,680 you can email but not text.
3. Replies. Use the vendors' own midpoints: ten percent of 720 texts is 72 replies, five percent of 1,680 emails is 84. About 156 replies.
4. Replies that are actual interest: about 30 percent. Strip out the opt-outs, wrong numbers, and polite no-thank-yous. Roughly 47 real conversations.
5. Conversations that become a booked estimate: about half. Roughly 23 estimates on the calendar.
6. Close rate on a warm past customer: call it 45 percent. Higher than cold, because they already know you. About 10 jobs.

Ten jobs at $2,500 is $25,000. Not $225,000. Not $600,000.

Now the part it would be dishonest to leave out. $25,000 is a good campaign. Sending 3,000 records across email and text costs a few hundred dollars, plus a few hours of setup. The return multiple those vendors quote, ten to twenty times, survives the honest math almost intact. It is the absolute revenue figure that does not.

That distinction matters because contractors do not budget against multiples, they budget against the revenue number. A contractor told to expect $200,000 who books $25,000 fires the vendor and never runs it again. A contractor told to expect $25,000 for $600 in sends runs it every quarter for the rest of his career. Same campaign, same result, opposite conclusion, and the only variable was the promise. It is the same trick we took apart in cost per lead is lying to you: a real number quoted at the wrong stage of the funnel.

We Read 17 of These Guides. Six Never Mention Consent at All.

The math is a modelling problem. The next one is a legal problem, so we measured it.

We took the pages ranking for the searches a contractor would actually type, things like "database reactivation for contractors" and "text message marketing for home services," fetched each, stripped the HTML, and searched the full text for any consent language at all. A mention anywhere counted, including footers and cookie banners, so every figure below is the most generous possible reading of these pages.

We targeted 20 URLs and retrieved 17 pages with real article text, across 16 domains. Three were excluded for returning a 2,887 character shell with no article in it. Worth noting: those shells still contained the word "consent," from a cookie banner, which is exactly the false positive our control term was built to catch.

Of the 17 pages:

• 9 mention the TCPA. Eight never do.
• 6 contain no consent language whatsoever: not "consent," not "opt-in," not "TCPA," not "CASL." We re-checked two of those six with a separate raw search to make sure it was not a parsing artifact. It was not.
• 4 mention written consent, which is the actual legal standard for a marketing text in the US.
• 3 mention A2P 10DLC registration, which US carriers require before your texts reliably deliver at all.
• 1 mentions CASL, the Canadian rule, and it is a Canadian domain.
• 1 mentions the 10 business day deadline to process an opt-out, which has been in force since April 2025.

Read that last line again. One page in seventeen tells a contractor about a deadline that has been federal law for over a year, on the exact campaign the page is selling.

The Rules That Decide Who You Can Actually Contact

Email and text are not one channel with two delivery methods. They are two separate legal regimes, and the gap between them is the most useful thing on this page. What follows is a plain-language summary, not legal advice, and the penalties are large enough that your own counsel is worth the hour.

Email in the US runs on CAN-SPAM, and it does not require prior consent. You may email a former customer cold. What you must do is tell the truth in your headers and subject line, include your physical postal address, and give a working opt-out. The FTC's own CAN-SPAM compliance guide is specific: your opt-out mechanism has to keep working for at least 30 days after you send, you must honour a request within 10 business days, and each violating email carries a penalty of up to $53,088. The FTC's own illustrative example of a message that must comply is, word for word, "a message to former customers announcing a new product line." That is a reactivation email.

Text messages in the US run on the TCPA, and it does require prior express written consent. Prior business does not create it. Statutory damages run per message, which is what turns a careless 2,000 record blast into a genuinely dangerous afternoon. Two changes matter right now:

• Since April 11, 2025, a consumer can revoke consent "in any reasonable manner," which means you cannot insist they use your magic keyword, and you must process it within 10 business days.
• The FCC's "revoke all" rule, under which opting out of one kind of message kills your consent for every kind, was pushed back again. Order DA-26-12A, released January 6, 2026, moved it to January 31, 2027. You have a runway, and it has an end date on it.

One thing to un-learn: the FCC's one-to-one consent rule, which many 2024-era posts still describe as law, was vacated by the Eleventh Circuit in January 2025. If your vendor still quotes it, their compliance guidance is two years stale.

In Canada, CASL flips the default. Commercial electronic messages need consent, including email, the opposite of the US position, and it catches contractors out. The saving grace is implied consent from an existing business relationship: two years from the last purchase, six months from an inquiry. A workable window for most campaigns, and a hard wall for the deep archive. Penalties reach $10 million for a corporation.

Practically this shakes out to: email the whole reachable list, text only the segment whose consent you can actually produce, register for 10DLC before sending a single text, and put a consent checkbox on every intake form today so next year's textable segment is three times bigger.

The Four Segments Worth Contacting, in Order

Reactivation gets sold as one blast to one list. It is worth far more as four separate messages, in this order.

1. Quotes you gave in the last 12 months and never heard back on. The highest value segment by a distance. They had the problem, they let you in the house, they got a number. No offer needed, just a plain note asking whether it ever got done. This is also the group that should never have gone cold, which is the subject of our lead follow-up system.
2. Past customers due on a natural service cycle. A furnace at year one, a water heater around year eight, a roof at year twelve. A reminder, not a pitch, and it converts like one. Time it to the season ahead, which is the argument in why August decides your fall.
3. Past customers who are off cycle. Do not sell them anything. Ask for the review you never asked for, or make a referral easy. The money here is indirect and real.
4. Old inquiries that never became a quote. Lowest yield, worst contact data, most consent-fragile. Send last, send by email, and judge the segment on one test before spending more.

The lesson underneath all of this: a reactivation campaign is a one-time recovery of value your follow-up process leaked. The cheapest version is the one you never need, because the quote got chased on day three instead of month fourteen. That is what automated follow-up is for, and why we pair it with an AI voice agent for the calls nobody picked up.

Run the campaign. Just run it on real numbers, send it to people you are allowed to send it to, and expect $25,000 instead of $600,000. It is a much better business decision at $25,000, because at $25,000 it is actually true.

Frequently Asked Questions