Roofing

    The Roofing Database Reactivation Playbook: What Your Dead Estimates Are Actually Worth

    Most roofing estimates lose to a competitor and then get filed as losses. Here's the math on what that pile is worth and the sequence that works it.

    Sonny Round, Co-Founder — Tailwinds Ops

    Key Takeaways

    • Only 16% of roofing and exterior contractors follow up the same day on unsold estimates (ServiceTitan, 1,000+ contractors surveyed).
    • 66% of homeowners gather three roofing quotes, so most estimates lose to a competitor rather than to a homeowner who never wanted the work.
    • Hatch's analysis of 163,212 estimate follow-up campaigns found a 60% average response rate; the best hit 90.06% using seven messages over seven days.
    • The worst campaign in that study pulled 17.82% using long messages and a voicemail-heavy mix.
    • Customer retention (53%) now outranks new customer acquisition (31%) as a growth priority for residential contractors.

    A roofing company that has been buying leads for five years is sitting on thousands of estimates that never closed. Every one cost money: the click, the call, the drive out, the hour on the roof, the write-up. That money is gone whether the job sold or not. Most roofers file those quotes under losses and go buy more leads. It is the cheapest pipeline in the business and almost nobody works it.

    What is roofing lead reactivation?

    Roofing lead reactivation is the practice of systematically re-contacting unsold estimates, old quotes and dormant inquiries already sitting in your CRM in order to book new jobs. It runs on text, email and voicemail across a fixed sequence, and it needs no new ad spend, because the cost of acquiring those homeowners is already sunk.

    A dead roofing estimate is a homeowner who met your rep, took your number in writing, and then bought from someone else or from nobody at all. It is the warmest cold lead in the trades.

    The general version is covered in our guide to database reactivation for service businesses. Roofing has a wrinkle most trades do not: your dead leads are not anonymous form fills. A rep stood on the roof, took photos, and handed the homeowner a number. That is a warm file sitting in a cold folder.

    What is a dead roofing estimate actually worth?

    Take the number of unsold estimates in your CRM and multiply it by what one estimate costs you all-in. That total is already spent. Reactivation is the only marketing motion where the acquisition cost is zero, because the receipt is already in last year's P&L.

    The sunk-cost math on 5,000 unsold estimates
    Input Value
    Unsold estimates in the CRM 5,000
    Your fully loaded cost per estimate Your number — ad spend, call handling, drive time, inspection, write-up
    Already spent, unrecoverable 5,000 x your number
    Response rate on a worked dormant list 2–5%
    Conversations created 100–250
    New ad spend required $0

    Run your own figure. If an estimate costs you $300 by the time a rep has driven out and written it up, 5,000 dead ones is $1.5 million already on the books. We are not going to hand you an industry-average cost per roofing lead — every candidate figure we found traces back to a marketing agency blog rather than primary research. Pull the number out of your own P&L. It usually lands higher than owners guess once drive time is counted.

    At a 2% response rate, 5,000 estimates is 100 conversations. Not 100 jobs — 100 homeowners who already met your rep, already have your number in writing, and are willing to talk again.

    Why do roofing estimates go dead in the first place?

    Two reasons, and neither one is that the homeowner was a tire-kicker. Roofing is a three-quote purchase, so most estimates lose by arithmetic. Then almost nobody follows up, which is what turns a competitive loss into a permanent one.

    66% of homeowners said they would gather three quotes before choosing a roofer, according to the 2026 homeowner survey run by Roofing Contractor with Owens Corning and myCLEARopinion. Another 27% said they would seek multiple quotes. Just 1% said they would accept a single one. A lost estimate tells you a competitor gave a better answer that week. It does not tell you the homeowner stopped needing a roof.

    Then the follow-up problem. ServiceTitan's 2026 Roofing & Exterior Market Report, a Thrive Analytics survey of more than 1,000 roofing-focused companies, found that only 16% of contractors follow up with homeowners the same day on unsold estimates. The same report put one-third of roofers at EBITDA margins between 6% and 15%. Thin margins and an unworked estimate pile tend to show up in the same business.

    The estimate that lost in March is a different lead by September. The homeowner who went with the cheaper roofer may have been ghosted. The one who needed to think about it has had two more storms since.

    How many times should you follow up on an unsold roofing estimate?

    Seven touches, weighted toward text, spread across about a week. Hatch analyzed 163,212 estimate follow-up campaigns and published the full breakdown. Average response rate was 60%. The best campaign hit 90.06%.

    • Seven messages total — five texts and two emails
    • Spread across seven days with two pause days built in
    • First text 200 characters, first email 134 characters
    • Every message ends with a question a homeowner can answer in one line

    The worst campaign pulled 17.82%. It sent six messages, nearly the same volume, but leaned on voicemail, ran long, and opened with a 434-character text. Same volume, opposite result. Length and channel mix did it.

    Hatch ran the study on HVAC campaigns and says the same templates are used across roofing, windows and remodeling, so treat the exact percentages as directional for a roof. The shape is what transfers: one long message loses, seven short ones win, and text carries the sequence.

    Their campaigns trigger two days after an estimate is given, and Hatch is explicit that the same structure should be run against 30-day and older open estimates. That second part is your backlog.

    Is working old estimates better than buying more roofing leads?

    For the first 90 days, yes, and it is not close. After that they are different tools. Buying leads scales with budget. Your database does not — it is finite, and once you have worked it, you have worked it. Start with the free pile, then go back to paid acquisition with a real cost-per-job number to beat.

    Working unsold estimates vs. buying new roofing leads
    Factor Working unsold estimates Buying new leads
    Cost to acquire the contact Already spent Paid again, per lead
    Prior relationship Homeowner met your rep and has your quote None
    Competitive position You were already one of the three quotes One of an unknown number
    Time to first booked job Days Weeks
    Scales with budget No — the list is finite Yes
    Main risk List hygiene and consent records Cost per lead climbing

    The market is moving this way regardless. ServiceTitan's 2026 Residential State of the Trades Report, a survey of 1,000 residential contractors, found customer retention (53%) now well ahead of new customer acquisition (31%) as a growth priority. Contractors are done paying full retail for demand they already own.

    Won't reaching out annoy homeowners who already said no?

    The homeowner data points the other way. In the Roofing Contractor survey, roughly one-third of homeowners named unclear or unexpected costs as their biggest problem with a roofer, and 23% named poor communication, missing updates and delays. Being contacted too often did not make the list. The complaint is silence.

    Two things keep it clean. Write like someone who remembers the job — reference the address, the date, the roof. And give people an obvious way out in the first message.

    The compliance side is not optional. In the US, marketing texts require prior express written consent, every message needs a working opt-out, and the standard sending window is 8am to 9pm in the recipient's local time. Carriers layer their own A2P registration rules on top of the federal ones. Suppress every opt-out permanently and keep the record. This is general information and not legal advice — have a telecom attorney review your consent language before a first send.

    How do you run this without hiring anyone?

    Four steps, roughly a week of setup, no new employee. The work is front-loaded into the list and the copy, which is where these campaigns are won or lost.

    • Pull and clean. Export every unsold estimate, no-show and dead inquiry. Strip duplicates, bad numbers, current customers, and anyone who ever opted out. Segment by age and job type — a 2023 replacement quote and a 2025 repair inquiry are different conversations.
    • Write for the file, not the list. Reference the specific property and the specific quote. Keep the first text near 200 characters and end it with a question a homeowner can answer in one word.
    • Sequence it. Five texts and two emails across seven days, with pause days. Kill the sequence the moment someone replies.
    • Route replies to a human fast. A reactivated homeowner behaves like a fresh lead and the window closes in minutes. If response time is already your weak point, the speed to lead research covers what the benchmarks actually are.

    The failure mode here is rarely the messaging. It is booking dozens of conversations in week one with nobody free to answer them. Do not start the sequence until someone owns the inbox.

    One adjacent point: half the homeowners in the Roofing Contractor survey said they use search engines to find a roofer. If buyers are asking an AI which roofer to call, your site has to be answerable by one. That is what our answer engine optimization work handles, and it runs in parallel with reactivation.

    Working the backlog itself is what our lead reactivation service does. We pull the list, build the sequence, and hand your team booked appointments.

    Frequently Asked Questions

    What is roofing lead reactivation?

    Roofing lead reactivation is the process of systematically re-contacting unsold estimates and dormant leads already in a roofing company's CRM in order to convert them into booked jobs. It runs on text, email and voicemail sequences and requires no new ad spend, because those homeowners were already paid for.

    How long can a roofing estimate sit before it is worthless?

    There is no hard expiry date. Hatch's estimate follow-up data uses a two-day trigger, but the same study recommends running the identical multi-touch structure against 30-day and older open estimates. Roofs fail on their own schedule, so a quote that lost two years ago can become urgent after one hailstorm.

    Why do so many roofing estimates go unsold?

    Mostly because homeowners shop. 66% said they gather three quotes before deciding, so two of every three estimates lose by default. Compounding that, ServiceTitan found only 16% of roofing and exterior contractors follow up the same day on an unsold estimate.

    How many follow-up messages should a roofing estimate get?

    Around seven across a week, weighted toward text. In Hatch's analysis of 163,212 campaigns, the top performer used five texts and two emails over seven days for a 90.06% response rate, while the worst used long, voicemail-heavy messaging and got 17.82%.

    Is texting old roofing leads legal?

    It is regulated, and it depends on your consent records. US marketing texts require prior express written consent, a working opt-out in every message, and sending inside the 8am to 9pm local window, with carrier A2P rules on top. This is general information and not legal advice — have a telecom attorney review your list before you send. The broader approach is covered in our database reactivation guide.