Retargeting Ads for Roofing Companies: A Practitioner's Operating Manual

Most roofing companies that try retargeting ads quietly turn them off after a quarter or two. The logic going in is sound: someone visited the site, didn't call, and a $7 ad nudge brings them back to convert a job worth $12,000. The logic coming out is grim: the campaign spent $900, served 40,000 impressions to the same 600 people, generated three form fills, and one of them was a competitor checking your pricing.

The gap between those two outcomes is almost never the ad creative and almost never the budget. It's that roofing demand is event-driven and geographically narrow, and retargeting was designed for e-commerce, where demand is steady and the audience is huge. A homeowner does not "abandon a cart" for a roof. They get a leak, a storm rolls through, a neighbor's house gets a dumpster and a crew, or the roof crosses an age threshold where insurers and buyers start asking questions. Retargeting can absolutely work for a roofer, but only when you treat it as the second touch in a system rather than a standalone lead source — and only when the first touch was aimed at homes that actually have a reason to need you.

Below is the operating manual: how retargeting actually functions in 2026 after the privacy changes that gutted the old pixel-based version, what to retarget and to whom, the budget math that tells you whether to even bother, the creative that converts a roofing visitor versus the creative that burns money, and how to wire retargeting to your CRM and your field team so a re-engaged visitor becomes a booked inspection instead of a vanity impression. We'll stay honest about where retargeting is the wrong tool, and where a sharper top-of-funnel targeting move beats any amount of ad-nudging.

What "retargeting" actually means now (and what broke)

Retargeting, also called remarketing, shows ads to people who already interacted with you — visited your site, watched a video, opened an email, engaged with your Google Business Profile, or sat in a customer list you uploaded. The promise is efficiency: you're paying to re-reach warm people instead of buying cold reach.

There are two mechanics, and the distinction matters more than most roofers realize:

• Pixel/site-based retargeting — you place a tag (the Meta Pixel, the Google Ads tag, the LinkedIn Insight Tag) on your website. When someone visits, they get dropped into an audience and chased with ads across that network. This is the version everyone means when they say "retargeting."
• List-based retargeting (Customer Match / Custom Audiences) — you upload a list of names, emails, phone numbers, and addresses. The platform matches them to accounts and shows ads to the match. This version survived the privacy apocalypse far better than the pixel version, and for roofers it's frequently the stronger play.

Here's what broke, because it changes the strategy. Apple's App Tracking Transparency prompt (rolled out in iOS 14.5 and tightened since) let users block the identifier apps used to follow them, and a large majority opted out. Browsers killed or sandboxed third-party cookies. The practical effect: pixel-based audiences shrank, attribution windows got fuzzier, and the platforms moved to modeled and aggregated reporting. For a national e-commerce brand with millions of monthly visitors, the audiences are still big enough to work. For a roofer in a metro of 400,000 people getting 1,500 site visits a month, a pixel audience that loses 40-60% of its members to opt-outs and cookie expiry can collapse to a few hundred reachable people — too small to run efficiently, and small enough that frequency caps get blown out and you annoy the same homeowners 15 times a week.

The fix is not to abandon retargeting. It's to (a) lean harder on first-party data and list-based audiences, which you own and which don't depend on a pixel surviving, (b) feed the funnel enough qualified top-of-funnel traffic that your retargeting pool is large enough to be worth running, and (c) measure on real outcomes (booked inspections, signed jobs) rather than platform-reported "conversions" that are increasingly modeled estimates.

Server-side and consent: the plumbing you can't skip in 2026

Two technical realities you have to account for:

1. Server-side tagging / Conversions API. Because browser-side pixels miss a growing share of events, the platforms now want you to send conversion data server-to-server — Meta's Conversions API (CAPI), Google's Enhanced Conversions, the equivalent on other networks. If you only run the browser pixel, you're under-reporting conversions, which makes the algorithm optimize poorly and makes your reports lie to you. A competent web developer or a tag-management setup can implement this; it is not optional anymore if you're spending real money.
2. Consent. Depending on the states you operate in, you may be subject to privacy laws (California's CCPA/CPRA, plus a growing list of state laws in Virginia, Colorado, Connecticut, Texas, and others) that require a consent or opt-out mechanism for tracking and for sharing data with ad platforms. A consent banner that actually controls whether tags fire is the compliant pattern. The platforms also have their own rules — Customer Match requires that you collected the data in a way that permits you to share it, and that you're not uploading lists you scraped or bought without consent. Read the platform's customer-list policy before you upload anything. The U.S. Federal Trade Commission has been explicit that deceptive data practices and misuse of sensitive data are enforcement priorities.

If that sounds like overhead, it is. It's also the reason the roofers who do it well have a moat: most of their competitors are running a busted browser pixel, mis-measuring, and slowly concluding "retargeting doesn't work for roofing." It works. They're just flying blind.

Which platforms to actually run on

You have four realistic venues, and they are not interchangeable for a roofer:

• Meta (Facebook + Instagram). The default for roofing retargeting because the placements are visual (job photos and video shine), Custom Audiences and Customer Match are mature, and the audience is in-feed where homeowners actually are. Downside: feed CPMs run higher than display, and iOS opt-outs hit Meta's tracking hard, so CAPI is mandatory. Best for: video proof, before-and-afters, past-customer reactivation.
• Google Display Network (GDN). Cheap CPMs and broad reach, which is its blessing and its curse — without aggressive placement exclusions, your ads end up on junk apps and made-for-advertising sites where impressions are worthless. Best for: low-cost presence retargeting, but only with a tight placement whitelist/blacklist.
• YouTube. Underused by roofers and genuinely strong, because a 15-30 second job-site video is the most trust-building format for a high-ticket purchase, and you can retarget people who watched your videos or visited your site. Best for: storm-season trust-building and re-warming a long consideration window.
• Google Search RLSA (Remarketing Lists for Search Ads). This is the sleeper. RLSA doesn't show display ads — it adjusts your search bids and messaging when a past visitor searches again. A homeowner who visited your site, then a week later searches "roof replacement near me," is the highest-intent retarget there is. You bid up and tailor copy for them specifically. For roofing, RLSA often outperforms display retargeting on cost-per-job because it catches the homeowner at the moment of active intent rather than passively in a feed.

The honest sequencing for most roofers: RLSA and Customer Match first (highest intent, privacy-resilient), Meta video and YouTube second (trust-building), broad GDN display last and only with tight exclusions.

Why roofing is a hard retargeting case (and how to win anyway)

Before spending a dollar, internalize the four structural problems, because every tactic later is a response to one of them.

1. The audience is tiny and local. You can only sell to homeowners in your service radius, ideally ones who own a detached single-family home of an age and type that needs roof work. A general visitor-retargeting pool includes renters, apartment dwellers, out-of-area researchers, other contractors, and people who already replaced their roof. You're paying to re-reach a lot of people who will never be customers.

2. The buying window is event-driven, not browse-driven. E-commerce retargeting works because the visitor wanted the product when they visited and still wants it three days later. A homeowner who looked at your site after a storm scare may have had the scare resolve, gotten a competing bid, or filed and moved on. The "intent" you captured is perishable in a way a sweater in a cart isn't.

3. Consideration is long and offline. Roofing is a high-ticket, low-frequency, trust-driven purchase. The decision happens over weeks, across spouses, often gated by an insurance process or a home sale. The conversion frequently happens by phone or at the kitchen table, not on a web form — so your on-platform "conversion" tracking misses the real event unless you connect offline outcomes back.

4. Frequency burns out fast. With a small pool, your impressions pile onto the same households. Without aggressive frequency caps, you go from "helpful reminder" to "why does this roofing company keep stalking me" in a week, which trains people to mentally filter (and report) your ads.

The winning posture, given all four: retargeting is a closer, not an opener. It re-warms people your other channels already heated up, and its job is to be present and credible during a long consideration window — not to manufacture demand. Pair it with a top-of-funnel motion that puts you in front of the right homes for the right reason, and retargeting compounds. Run it alone and it starves.

The five retargeting audiences worth building for a roofer

Not all warm traffic is equal. Build these as separate audiences with separate creative and separate budgets, because a homeowner who priced out a $14,000 replacement is worth a very different ad than someone who read your blog post about gutter cleaning.

1. High-intent site visitors (the page tells you the intent)

Segment by what page they hit, rather than lumping everyone under "visited site":

The single biggest waste in roofing retargeting is treating a blog reader the same as a quote-form abandoner. Separate them. The quote-abandoner audience will be small but should get most of your money.

2. Past customers and your house list (Customer Match)

This is the audience most roofers ignore and the one with the best economics, because you already earned the trust. Upload your customer and lead lists (with proper consent) as a Customer Match / Custom Audience:

• Past roof-replacement customers for gutter, attic ventilation, maintenance plans, and — years out — re-roofs and referral asks. A homeowner you replaced a roof for in 2019 is not a re-roof prospect, but they're a perfect referral and add-on prospect.
• Old quotes that never closed. Someone who got a bid 14 months ago and didn't move may be back in market now, especially after a storm. A list-based retarget to dead quotes is one of the cheapest "leads" you can generate.
• Lookalike seeds. Your closed-won customer list is the best seed for a lookalike/similar audience to expand cold targeting — that's prospecting, not retargeting, but it's the same upload powering it.

3. Engaged social and video viewers

People who watched 50%+ of your job-site videos, engaged with your Facebook/Instagram posts, or interacted with your Google Business Profile. These on-platform audiences don't depend on your website pixel surviving, and a homeowner who watched two minutes of you tearing off and replacing a roof is meaningfully warm. Build a "video viewers 25%+ in last 60 days" audience and retarget with a soft proof-and-CTA message.

4. Local + recently storm-exposed overlay

Layer geography and recency on top of everything. If a hail or wind event hit specific neighborhoods, your site visitors and list members in those neighborhoods in the days after the event are your most perishable, most valuable retargeting audience. Don't let a generic "all visitors, last 180 days" audience dilute them. (More on the compliance line around storm/claim messaging below — there is a right and a wrong way to say this.)

5. Cart-equivalent: started-but-didn't-finish actions

The closest roofing analog to an abandoned cart:

• Started a quote form, didn't submit
• Clicked "call" on mobile but the call didn't connect or was under 30 seconds
• Booked-then-cancelled an inspection
• Opened a personalized report or estimate link and didn't reply

These are your true "abandoners," and they convert far above any browse audience. Wire your forms and call tracking so these micro-events are captured as audience triggers.

Window length by audience: how long to keep chasing

A common error is one global retargeting window — "everyone, last 180 days." A roofing visitor's intent decays at very different rates depending on what they did, and a too-long window pads your pool with cold people who dilute frequency and waste budget. Set the membership duration per segment:

The principle: the hotter and more perishable the action, the shorter and more aggressive the window; the colder the action, the shorter you cut it (or skip it) to protect budget. Evergreen low-frequency treatment is reserved for people who already trust you — past customers.

Budget math: should you even run retargeting?

Run the numbers before the campaign, because for a lot of roofers the honest answer is "your pool is too small this quarter — fix the top of funnel first."

Step 1: Estimate your reachable pool

Take your monthly unique website visitors. Strip the obvious non-prospects (rough rule, but start here): subtract bounce-only sessions under 10 seconds, subtract out-of-area traffic, subtract renters/apartment traffic you can infer. Then assume privacy attrition: a realistic pixel-based reachable audience is often 30-50% of raw visitors after opt-outs and cookie expiry.

Worked example:

• 2,000 monthly site visitors
• Minus ~35% bounce/junk = 1,300 meaningful sessions
• Minus ~25% out-of-area/renter = ~975 in-market
• Times ~40% reachable after privacy attrition = ~390 reachable retargeting prospects/month

If that number is under ~300-500, pixel retargeting will be inefficient on its own (you'll blow frequency caps and the algorithm won't have enough to optimize). That's your signal to lean on Customer Match lists and to invest in top-of-funnel volume first.

Step 2: Set frequency and budget to match the pool

A sane frequency for a high-ticket consideration purchase is roughly 2-4 impressions per person per week, capped, over a 21-45 day window. Over-serving past that doesn't increase conversions; it increases irritation and ad fatigue.

Rough monthly impression need = reachable pool × impressions/person/week × ~4 weeks.

• 390 people × 3 impressions/week × 4 = ~4,680 impressions/month.

At a display CPM of roughly $5-12 (varies by network and geo), that's ~$25-55/month for display, or more on Meta/feed placements where CPMs run higher. The point of the exercise isn't the exact dollar — it's that a healthy retargeting budget is small relative to your top-of-funnel spend. If a vendor is pitching you $2,000/month of "retargeting" against a 400-person pool, they are either over-serving your audience into oblivion or quietly buying cold reach and calling it retargeting. Ask to see the audience size and the frequency report.

Step 3: Judge it on cost-per-booked-inspection, not CPC or CTR

Platform "conversions" for roofing are mostly form fills and clicks-to-call, and a meaningful share of those are junk (wrong area, price shoppers, competitors). The only numbers that matter:

• Cost per booked, qualified inspection
• Cost per signed job from retargeting-attributed leads
• Incrementality: did retargeting actually cause conversions, or take credit for people who'd have come back anyway?

That last point is the one roofers get fleeced on. Retargeting platforms claim credit for a homeowner who already typed your name into Google and would have called regardless. To check it honestly, run a holdout test: exclude a random ~10-20% of your eligible audience from seeing retargeting ads for a month, and compare conversion rates between the held-out group and the targeted group. If the targeted group doesn't convert measurably better, you're paying to reach people who'd convert anyway. This single test separates roofers who know their numbers from those who get billed for fog.

A full worked P&L for a single quarter

Numbers make this concrete. Take a roofer in a metro with a $14,000 average replacement job and a 35% gross margin (so ~$4,900 gross profit per job). They get ~2,000 monthly site visits, yielding the ~390 reachable retargeting prospects we calculated earlier, refreshed monthly.

Quarterly retargeting spend, run sanely:

• Meta + GDN + YouTube retargeting media: ~$450/month × 3 = $1,350
• RLSA bid uplift (incremental search spend): ~$300/month × 3 = $900
• Tooling/tracking allocation (call tracking, server-side tag share): $300
• Total quarterly spend: ~$2,550

Now the conservative outcome. Across the quarter, the retargeting + RLSA audiences produce 18 booked inspections. Your inspection-to-signed close rate is 30%, so 5.4 — call it 5 signed jobs.

• Revenue: 5 × $14,000 = $70,000
• Gross profit: 5 × $4,900 = $24,500
• Cost per booked inspection: $2,550 / 18 = $142
• Cost per signed job: $2,550 / 5 = $510
• Return on ad spend (revenue basis): $70,000 / $2,550 = 27×; on gross-profit basis: ~9.6×

That looks fantastic — and it's the number a vendor will show you. But the holdout test is where honesty lives. Suppose your holdout reveals that 2 of those 5 jobs would have come back and signed without any retargeting (they'd already decided and were going to call). The incremental picture:

• Incremental signed jobs: 3
• Incremental gross profit: 3 × $4,900 = $14,700
• True cost per incremental signed job: $2,550 / 3 = $850
• Incremental gross-profit ROAS: $14,700 / $2,550 = ~5.8×

Still a strong program — but the real return is roughly 40% lower than the dashboard claimed. A 5.8× incremental return is a keep; a program that pencils to 27× on paper but turns out to be 1.5× incremental is a cut, and you'd never know which one you have without the holdout. This is the discipline that separates roofers who scale retargeting profitably from those who pour money into a number they never validated.

Creative that converts a roofing visitor (and the stuff that wastes money)

A retargeting impression is a reminder, not a first impression — the homeowner already knows who you are. So the creative job is different from a cold ad. Three rules:

Rule 1: Match the message to the segment. The quote-abandoner sees "Still deciding? Your inspection slot is held — book in 60 seconds." The blog reader sees a trust/proof brand ad. The past customer sees a referral or add-on offer. One generic ad to all of them wastes the segmentation you built.

Rule 2: Lead with the specific reason they were on your site. If they viewed the storm/hail page, the ad references storm-damage inspection and documentation — not a generic "we do roofs." Specificity is what makes a reminder feel relevant instead of creepy.

Rule 3: Reduce the next step to something tiny. Not "buy a roof." The micro-conversion: book a free inspection, get a same-week estimate, see your home's roof report, text us a photo of the problem. High-ticket purchases convert on small next steps, not big ones.

Creative that works for roofers

• Real local job photos / before-and-afters from neighborhoods the viewer recognizes. Stock roofing photos read as "national lead company," which kills trust. A drone shot of a tear-off two miles away reads as "these are my neighbors' roofers."
• Short job-site video (15-30 seconds): crew, cleanup, the finished ridge line, a one-line homeowner quote. Video retargeting outperforms static for trust-driven purchases.
• Specific, honest proof: years in business, manufacturer certification (GAF Master Elite, Owens Corning Platinum, CertainTeed SELECT ShingleMaster), local review counts, warranty terms. Real credentials, not adjectives.
• A held slot / urgency that's true: "Booking storm inspections this week" works if it's true. Fake countdown timers don't, and they erode trust on a high-ticket purchase.
• Financing-forward creative for the price-sensitive segment: monthly payment framing, not five-figure totals.

Five retargeting ad examples, written out by segment

Concrete copy beats abstraction. Here is a real-world set you can adapt — note how each is anchored to what the homeowner already did and ends in a tiny next step:

1. Quote-form abandoner. "Still weighing your roof estimate? We held your inspection slot through Friday. Two minutes to confirm a time — no pressure, no obligation." Pair with a photo of a finished local job. CTA: Confirm my inspection.
2. Storm-page viewer (compliant). "Worried the [neighborhood] hail hit your roof? We'll inspect it, photograph any damage, and write you a detailed repair estimate to use with your insurer. You file; they decide coverage." CTA: Book a storm inspection. (Notice: no claim-handling, no payout or deductible promise.)
3. Past customer — referral/add-on. "Your roof's handled — how about your gutters and attic ventilation? Past clients get priority scheduling. Know a neighbor who needs us? We'd be grateful." CTA: See add-on options.
4. Video viewer (trust). "You watched us replace a roof on [street]. Here's what 200+ neighbors already know about working with us." Short testimonial cut. CTA: Get a free inspection.
5. Financing-sensitive viewer. "A new roof doesn't have to be a five-figure surprise. See monthly payment options before you decide — most homeowners qualify in minutes." CTA: Check my options.

The throughline: specificity (the neighborhood, the action they took, the proof) is what makes a second impression feel relevant instead of intrusive. Generic beats nobody.

Creative and messaging that wastes money — or gets you in legal trouble

• Generic "We're the best roofers in town!" with a stock photo. Says nothing, proves nothing.
• The same ad to everyone regardless of segment.
• Heavy discounting on a high-ticket job — trains people to wait for a better deal and attracts price shoppers, not the homeowners who value workmanship.
• Anything that crosses the insurance-claims line. This is where roofers get into real trouble, so it gets its own section.

The storm/claims compliance line in retargeting creative

A large share of roofing retargeting runs after storms, aimed at homeowners researching damage. There is genuine, valuable help you can offer — and a bright line you cannot cross without risking unlicensed-public-adjusting exposure and platform ad rejections. Get this right; it protects your license and your ad account.

What a contractor may do and may say in an ad: offer to inspect the roof, document damage thoroughly with photos, and prepare an accurate repair estimate aligned to standard estimating practices (Xactimate-style line items) for the homeowner to use. State facts about the work you would do and your own scope. Educate homeowners on what storm damage looks like and how to document it.

What you may NOT say or imply in an ad — the do-not-say list:

• "We handle your insurance claim for you" / "we deal with the adjuster for you." Negotiating or adjusting a claim for a fee, on the homeowner's behalf, is public adjusting and is licensed (and in many states a contractor is prohibited from doing it at all).
• "We'll get your claim approved" or any promise of a specific payout, approval, or coverage outcome. You don't decide coverage; the insurer does.
• "We'll waive / cover / eat your deductible" or "free roof." Absorbing or rebating a deductible is illegal in many states and is insurance fraud framing everywhere; "free roof" advertising is specifically restricted in numerous states' statutes.
• Interpreting the homeowner's policy or telling them what is and isn't covered.
• Representing the homeowner against their insurer.

The safe frame that still captures the intent: "Storm-damage inspection and documentation. We inspect your roof, photograph the damage, and write you a detailed repair estimate. You file with your insurer; your insurer decides coverage." That message converts the same scared homeowner without putting your license — or your ad account — at risk. Many state Departments of Insurance (Texas's TDI is a well-known example) and Attorneys General publish guidance on the contractor/public-adjuster line and on deductible/"free roof" prohibitions; read your state's before you write storm creative. When in doubt, keep your claims entirely on the documentation and estimate side of the line and let the homeowner and the insurer own the claim itself.

Wiring retargeting to the rest of the system

Retargeting that isn't connected to your CRM and your field team is a leaky bucket. The re-engaged homeowner clicks, fills a form, and then... sits in a Facebook lead form nobody checks until Thursday. Here's the connective tissue that turns an impression into a signed job.

1. Capture the source immutably

When a retargeting click becomes a lead, the lead's first-touch source should be recorded and never overwritten by a later touch. If a homeowner first found you via a postcard, then a Google search, then a retargeting ad finally got the form fill, naive attribution gives retargeting 100% credit and you over-invest in it. Immutable first-touch plus full-path tracking keeps you honest about what's actually originating demand versus what's just the last nudge.

2. Speed-to-lead under 5 minutes

A retargeting lead is a homeowner mid-consideration. Lead-response research across industries is brutally consistent: contacting a web lead within 5 minutes versus 30 minutes can swing your odds of connecting and qualifying by an order of magnitude. Route retargeting form fills and click-to-calls straight into the pipeline with an instant alert (text/Slack/CRM task), not a daily email digest.

3. Close the loop with offline conversions

The signed job happens at the kitchen table, days after the click. If you don't feed that outcome back to the ad platform (via offline conversion import / CAPI), the algorithm optimizes toward cheap form fills instead of real jobs, and your reports overstate junk leads. Pushing "booked inspection" and "signed job" back as offline conversions is what lets retargeting optimize toward revenue instead of clicks.

4. Suppression lists (stop paying to chase the won and the dead)

Continuously exclude:

• Customers who already signed (stop paying to retarget a closed job)
• Leads you disqualified (out of area, renter, not a real roof project)
• Anyone who opted out

Without suppression, you waste 20-40% of a small budget re-serving ads to people who can't or won't convert.

5. Sequence the touches — don't blast one ad forever

A homeowner in a 30-day consideration window shouldn't see the same creative 12 times. Sequence it the way a good salesperson would pace a conversation:

• Days 1-7 (re-warm): the proof/trust ad — local job photos, reviews, certification. You're re-establishing credibility, not closing.
• Days 8-21 (educate + de-risk): address the objection that probably stalled them — financing, the inspection process, what storm documentation actually involves, your warranty. This is where you answer the unspoken "why haven't I called yet."
• Days 22-30 (direct): the held-slot, book-now ad with a genuine reason to act. If they don't convert by the end of the window, drop them from active retargeting (move to the evergreen low-frequency pool only if they're a past customer).

Sequenced messaging consistently beats a single repeated ad for high-ticket, considered purchases, because it mirrors how the decision actually unfolds — trust first, objections second, action last. Most platforms let you build this with sequential audiences or rules; even a manual three-flight rotation beats one ad on a loop.

How RoofPredict makes retargeting actually pay — by fixing the top of the funnel

Retargeting's core constraint is the one we opened with: a pool that's too small and too unqualified to be efficient. You can't retarget your way out of a weak top of funnel. RoofPredict's job is to make the funnel that feeds your retargeting both bigger and pre-qualified, so the warm pool you're re-engaging is full of homes that actually have a reason to need a roof.

Targeting that builds a qualified pool, house by house. RoofPredict scores every home in your service area by roof-age band (recent / mid-life / due / overdue) plus per-roof storm exposure plus an opportunity score, and produces a ranked target audience — which roofs are due, address by address — with a "why this home" evidence chain. You draw your territory on a hex map, filter to storm-hit areas, or import your own address list. That ranked due-roof list is the raw material for a retargeting program that isn't chasing random blog readers — it's chasing homeowners whose roofs are genuinely in the replacement window. (Honest limit: roof age here is a range, not an exact install date, and a storm-exposure score is odds of exposure, not proof of damage — you still inspect to confirm.)

First touch that actually drives the site visits you'll later retarget. From that ranked list, RoofPredict turns due roofs into a tracked direct-mail campaign — personalized mail proofs checked for brand, copy, and address; vendor release; per-piece delivery and return tracking; cost quoted up front. Every targeted home also gets a personalized microsite and PDF report (roof profile, storm history, risk and cost-of-waiting) with a lead-capture form, plus per-home and lookup QR codes for the mail piece and the door hanger. That's the move most roofers miss: the mail piece and QR drive a homeowner to their own report page, and that visit is what seeds a clean, high-intent retargeting audience — a homeowner you already know is in a due-roof, storm-exposed home, not an anonymous bounce.

A lead pipeline and CRM sync so the re-engaged homeowner doesn't leak. When retargeting (or any channel) produces a lead, RoofPredict's pipeline tracks it new → contacting → appointment → inspected → won/lost with an immutable first-touch source, and syncs two-way to 13 CRMs — HubSpot, ServiceTitan, JobNimbus, AccuLynx, Jobber, Housecall Pro, Salesforce, Pipedrive, Leap, Roofr, SalesRabbit, CompanyCam, plus Zapier and CSV. So the retargeting form fill lands in the same system as your canvassing and mail leads, with the original source preserved — which is exactly the immutable-first-touch attribution that keeps you from over-crediting the last ad nudge.

The honest framing: RoofPredict is not a retargeting ad platform and won't run your Meta campaign. What it does is make retargeting worth running — by giving you a ranked, qualified audience to drive to your microsites and forms, the mail-and-QR first touch that seeds the warm pool, and the pipeline that captures the result without losing the source.

Measuring it so you can defend the spend

Retargeting attribution is where roofers get separated from their money. Build a measurement frame you can defend to yourself.

The funnel you actually track

Actual vs estimate vs benchmark

Don't just track what happened — track it against what you projected and against a sane industry benchmark. If your projected cost-per-win was $400 and the actual is $1,800, the campaign is failing even if the platform dashboard is green with cheap form fills. RoofPredict's results funnel reports delivered → views → form → calls → leads → wins with cost-per-lead and cost-per-win, and shows actual vs estimate vs industry benchmark with A/B campaign variants — which is the frame that tells you whether the retargeting tied to a given mail-and-microsite campaign actually beat its projection, instead of just looking busy.

The reports that catch the scam

• Frequency report. If a small audience is seeing 15+ impressions/week, you're burning money and goodwill. Cap it.
• Placement report. Display retargeting on junk apps and made-for-advertising sites is where budgets quietly die. Exclude aggressively; whitelist if you can.
• Search-terms / audience-overlap report. Make sure your "retargeting" isn't secretly buying cold reach because the pool was too small to spend the budget.
• Geo report. Confirm spend is inside your service radius. Out-of-area impressions are pure waste for a roofer.

When NOT to run retargeting (the honest part)

Retargeting is the wrong first move when:

• Your pool is under ~300-500 reachable people. Fix top-of-funnel volume first (mail, canvassing, qualified targeting, local search). Retargeting a tiny pool just over-serves the same households.
• You have no offline-conversion loop. Without feeding booked inspections and signed jobs back, you'll optimize toward junk form fills and conclude retargeting "works" while your close rate craters.
• Your site doesn't convert. Retargeting sends warm people back to the same site that already failed to convert them. If the site has no clear inspection CTA, weak proof, and a slow form, fix the site before paying to send people to it twice.
• You can't respond fast. If retargeting leads sit for hours, the warm window closes. Speed-to-lead first, then spend on driving more leads.
• The math says canvassing or mail wins. For a hyper-local, event-driven, high-ticket purchase, getting the right due-roof, storm-exposed homes onto a mail piece and a door route often beats chasing a thin pixel pool. Sometimes the better "retargeting" is a second postcard and a door knock to a ranked list — physical touches to homes you know are due.

The roofers who win with retargeting are unsentimental about this. They run it where the pool and the loop justify it, and they spend the rest on the top-of-funnel targeting and field motion that creates the warm audience in the first place.

A 30-day implementation plan

A concrete sequence you can run starting Monday.

Week 1 — Plumbing and audiences

1. Implement server-side conversion tracking (Meta CAPI / Google Enhanced Conversions). Verify events fire.
2. Stand up a consent mechanism that actually controls tag firing, sized to the states you operate in.
3. Build segmented audiences: quote-abandoners, service-page viewers, blog readers, video viewers 25%+, past customers (Customer Match upload, consent-checked).
4. Build suppression lists: signed customers, disqualified leads, opt-outs.

Week 2 — Creative and offers

1. Produce 3-4 creatives per priority segment using real local job photos and a short job-site video. No stock.
2. Write segment-matched copy with a tiny next step (book inspection / get your roof report / text a photo).
3. If running storm creative, scrub every line against the do-not-say list. Documentation-and-estimate framing only.

Week 3 — Launch small, measure real

1. Launch with strict frequency caps (2-4/week) and a 21-30 day window.
2. Carve out a 10-20% holdout from your eligible audience for incrementality testing.
3. Route all leads to instant alerts with sub-5-minute response. Push booked-inspection and signed-job events back as offline conversions.

Week 4 — Read the truth and reallocate

1. Pull cost-per-booked-inspection and cost-per-win, not CTR.
2. Compare holdout vs targeted conversion rates — keep only the lift that's real.
3. Compare actual vs your estimate vs benchmark. Cut placements and segments that lose; double the ones that win.
4. If the pool was too thin to spend efficiently, shift budget upstream to qualified targeting + mail to build the warm audience, then re-run.

Do that, and retargeting stops being the line item you quietly turn off and becomes the closer that quietly lifts every other channel — because the homeowner who got your postcard, scanned the QR to their own roof report, and didn't quite call, sees you one more time at the moment they're finally ready, and books.

FAQ

Do retargeting ads actually work for roofing companies?

Yes, but only as a closer, not a lead source on its own. Roofing demand is local, event-driven, and the consideration window is long, so retargeting's job is to stay present and credible while a warm homeowner decides — not to manufacture demand. It works when your pool is large enough (roughly 300-500+ reachable people), your creative is segmented and local, and you feed booked inspections and signed jobs back to the platform so it optimizes toward real revenue instead of junk form fills. Run alone against a thin pool, it just over-serves the same households and gets turned off.

How much should a roofing company spend on retargeting?

A healthy retargeting budget is small relative to top-of-funnel spend, because you're re-reaching a finite warm pool, not buying cold reach. Size it from your reachable audience times your frequency cap (2-4 impressions per person per week) over a 21-45 day window. For a typical local roofer that's often tens to low-hundreds of dollars a month for display, more on feed placements. If a vendor pitches $2,000/month against a 400-person pool, they're either over-serving your audience or quietly buying cold traffic and calling it retargeting — ask to see the audience size and frequency report.

Why did privacy changes hurt roofing retargeting more than e-commerce?

Apple's App Tracking Transparency opt-outs and the death of third-party cookies shrank pixel-based audiences by roughly 30-60%. A national e-commerce brand still has millions of visitors, so even a reduced audience is workable. A local roofer with 1,500-2,000 monthly visits can see a pixel pool collapse to a few hundred reachable people — too small to optimize and easy to over-serve. The fix is leaning on first-party Customer Match lists you own, server-side conversion tracking (CAPI / Enhanced Conversions), and building more qualified top-of-funnel traffic.

What's the difference between pixel retargeting and Customer Match for roofers?

Pixel/site-based retargeting drops website visitors into an audience via a tag — it's powerful but shrank badly under privacy changes. Customer Match (also called Custom Audiences) uses a list of names, emails, phones, and addresses you upload and the platform matches. For roofers, Customer Match is often the stronger play: it survived the privacy changes, it lets you retarget past customers for add-ons and referrals, and it lets you re-reach old quotes that never closed — frequently the cheapest leads you can generate. You must have collected the data in a way that lets you share it.

Can I run retargeting ads about storm or insurance claims?

You can run storm-damage inspection and documentation ads, but stay strictly on the documentation-and-estimate side. You may offer to inspect, photograph damage, and write an accurate repair estimate the homeowner uses. You may NOT say you'll handle the claim, deal with the adjuster, get the claim approved, promise a payout, waive or cover the deductible, advertise a free roof, or interpret the homeowner's policy — those cross into unlicensed public adjusting and, for deductibles and free-roof claims, are illegal in many states. Safe frame: you inspect and document, the homeowner files, and the insurer decides coverage. Read your state Department of Insurance guidance before writing storm creative.

How do I know if retargeting is actually causing conversions or just taking credit?

Run a holdout test. Randomly exclude about 10-20% of your eligible audience from seeing retargeting ads for a month, then compare conversion rates between the held-out group and the targeted group. If the targeted group doesn't convert measurably better, you're paying to reach homeowners who would have come back anyway — common when retargeting claims credit for someone who already typed your name into Google. Holdout-tested incrementality, measured on cost-per-booked-inspection and cost-per-win rather than clicks, is how you separate real lift from billed fog.

What metrics should I judge roofing retargeting on?

Not CTR or platform-reported conversions. Judge it on cost per booked, qualified inspection and cost per signed job, plus holdout-measured incremental lift. Platform conversions for roofing are mostly form fills and click-to-calls, a meaningful share of which are out-of-area, price shoppers, or competitors. Track the full funnel — reach and frequency, micro-conversions, booked inspections, signed jobs — and compare actual cost-per-win against your estimate and an industry benchmark. A dashboard that's green with cheap form fills while your cost-per-win is triple your projection is a failing campaign.

Why are my retargeting leads not turning into jobs?

Usually one of four breakdowns: no offline-conversion loop, so the platform optimizes toward cheap junk form fills instead of real jobs; slow speed-to-lead, so warm homeowners go cold before you call (aim for under 5 minutes); a site that didn't convert them the first time and still doesn't the second; or no segmentation, so blog readers get the same hard-sell ad as quote-abandoners. Fix the loop, route leads to instant alerts with immutable first-touch source tracking, segment your audiences, and make sure your site has a clear inspection CTA before you pay to send people back to it.

How does RoofPredict help with retargeting if it isn't an ad platform?

RoofPredict fixes the top-of-funnel constraint that makes most roofing retargeting fail — too small and too unqualified a pool. It scores every home in your area by roof-age band and storm exposure into a ranked due-roof audience, turns that list into tracked direct mail with personalized microsites, PDF reports, and per-home QR codes, and captures the resulting site visits and leads in a pipeline that preserves an immutable first-touch source and syncs two-way to 13 CRMs. So the warm audience you retarget is full of genuinely due, storm-exposed homes, the QR-driven microsite visit seeds a clean high-intent pool, and the results funnel shows actual-vs-estimate cost-per-win. It won't run your Meta campaign — it makes that campaign worth running.

Is retargeting better than direct mail or canvassing for roofers?

It's not either/or — they reinforce each other, but the right first move depends on your pool. For a hyper-local, event-driven, high-ticket purchase, getting the right due-roof and storm-exposed homes onto a mail piece and a door route often beats chasing a thin pixel pool. The strongest pattern is a ranked due-roof list driving mail and QR codes to personalized microsites (which seeds a clean retargeting audience), then retargeting re-warming those known-qualified homeowners during a long consideration window, with canvassing closing the highest-opportunity addresses. If your reachable retargeting pool is under a few hundred people, spend upstream on targeting and physical touches first.