A rock kicked up on I-10 in Tempe at 7:45am. The windshield cracked from the passenger side to the driver's side in about four seconds. The driver pulled over, took a photo of the damage, and called her insurance company from the shoulder.
Fifteen minutes later, she'd filed a claim with State Farm, gotten a claim number, and was searching "windshield replacement near me" because she wanted to pick her own shop instead of using Safelite through the carrier's referral program. She called the first number that looked local.
That call is worth $35-45 to the shop that answers. The driver isn't price shopping — her insurance covers the repair. She's choosing based on convenience, reviews, and who can get it done today. Close rate on these calls: 85%+.
Same morning, different call. A driver in Mesa noticed a small chip in his windshield from a parking lot mishap two weeks ago. No insurance claim — deductible is higher than the repair cost. He's calling three shops to compare prices. "What do you charge for a chip repair?" If the first number is $20 higher than the second, he's gone.
Both calls cost roughly the same CPL. One closes at 85%. One closes at 30%. Unless you separate these? Your blended metrics will look mediocre and you'll never realize how profitable the insurance segment actually is. (I've watched shops run blended programs for six months wondering why margins feel thin — and yes, I spent way too long making the same mistake before I figured this out.)
Auto glass is an underserved vertical in pay-per-call. Steady demand — windshields crack year-round. Clear insurance vs. cash-pay split. Mobile service creates premium pricing. But there's no operator-facing guide on how to buy these calls properly. This is that guide.
What Auto Glass Calls Actually Cost
The CPL ranges operators pay — not what shows on rate cards.
Windshield replacement (insurance claim) — CPL: $30-45. The highest-value call in the vertical. Caller has already filed a claim. Carrier pays $250-600 for the replacement. The caller isn't price-sensitive — they're choosing based on speed, convenience, and reviews. Close rate: 80-90%. At a $38 CPL and 85% close, your cost per job is $45 against a $350+ paid repair. The math is good.
Windshield replacement (cash-pay) — CPL: $22-35. Caller is paying out of pocket. Maybe their deductible is $500 and the repair is $380. Maybe they don't have comprehensive coverage. These callers shop 3-5 quotes. Close rate: 25-40%. Tickets run $200-450 depending on vehicle make and glass type. You can make money here, but you're competing on price.
Chip repair — CPL: $15-25. Small ticket ($50-100) but fast turnaround. A tech can do 4-6 chip repairs in the time it takes to do one windshield replacement. Insurance often covers chip repair with no deductible — callers sometimes don't realize this. Close rate: 50-65% when you mention insurance coverage proactively.
Mobile service premium — add $5-12 to base CPL. Mobile-qualified callers expect the tech to come to them — home, office, parking lot. This is a feature for busy callers and a margin opportunity for shops. If you offer mobile, filter for it and pay the premium. If you don't, filter it out.
After-hours emergency — CPL: $35-50. Windshield shattered at 9pm, driver can't wait until morning. Limited supply of after-hours installers. Close rate: 70-85%. Premium pricing is expected and accepted.
The metric that matters: cost per closed job by call type. A $40 insurance-claim call closing at 85% costs $47 per job. A $25 cash-pay call closing at 30% costs $83 per job. The "cheaper" leads are often more expensive. For click fraud eating your budget before calls even come through, ClickzProtect handles detection on the paid search side.
Insurance vs. Cash-Pay: The Split That Defines Your Economics
Auto glass has the cleanest insurance/cash split of any home or auto service vertical. And it's binary — the caller either has a claim filed or they don't.
Insurance-claim callers aren't shopping on price. Their carrier covers the repair. They're choosing based on:
- Who can do it today (or tomorrow morning at latest)
- Who comes to them vs. requiring a shop visit
- Reviews and reputation
- Whether they trust you more than Safelite
These close at 80-90%. The conversation is short: confirm damage, confirm coverage, schedule. Ticket value is whatever the carrier pays (usually $300-500 for windshield replacement), and you're not negotiating it.
Cash-pay callers are shopping. They'll call 3-5 shops, ask "how much for a windshield on a 2019 Camry," and pick based on price and availability. Close rate: 25-40%. You're competing with every other shop and with mobile services that advertise "$199 windshield replacement."
If you route these identically, your blended close rate lands around 45-55% — which looks fine until you realize your insurance calls could be closing at 87% if you prioritized them.
IVR separation:
"Press 1 if you've filed an insurance claim for your windshield damage. Press 2 if you're paying out of pocket or haven't contacted your insurance yet."
Route Press 1 to your fastest dispatcher. These callers don't need a sales pitch — they need a time slot. Route Press 2 to your appointment setter who can sell on warranty, OEM glass, and convenience instead of just competing on price.
Some shops take it further: Press 1 gets the owner or senior tech. Press 2 gets the appointment desk. The thinking is that insurance callers are so high-value they deserve white-glove treatment.
And — honestly? — that thinking is correct. I resisted this for a long time because it felt elitist. But the numbers don't care about my feelings.
Qualifying Damage Type: What to Ask
Not all glass damage is equal. Different damage types need different techs, different inventory, different time slots.
Windshield (front) — the volume play. 70%+ of auto glass calls. Replacement tickets $250-600 depending on vehicle. Chip repair tickets $50-100. Most shops specialize here.
Side glass (door windows) — often insurance-covered if from a break-in. Tickets $150-350. Faster install than windshield. Lower volume but steady.
Rear glass — less common. $200-400 tickets. Defroster elements complicate some installs. Make sure your tech knows the vehicle before booking.
Sunroof/moonroof — specialty. $400-800 tickets. Not every tech handles these. If you take sunroof calls, route to a qualified tech or you'll schedule a job you can't complete.
Qualify in the IVR or early in the conversation:
"Is the damage to your windshield, a side window, rear window, or sunroof?"
This lets you route to the right tech with the right inventory. A caller needing a 2021 F-150 windshield shouldn't go to the tech who only stocks sedan glass. A sunroof job shouldn't go to a tech who's never done one.
For landing page attribution without the GDPR headaches, JustAnalytics handles that cleanly.
Location Qualification: The Mobile Service Factor
Auto glass is one of the few verticals where "we come to you" is a meaningful differentiator. Mobile service — the tech drives to the customer's home or office, does the install in the parking lot — commands premium pricing because it sells convenience.
But not every shop offers mobile. And callers who want mobile service will cancel if they learn they have to drive to a shop.
If you offer mobile service:
- Qualify for it: "Would you like us to come to your location, or would you prefer to bring your vehicle to our shop?"
- Route mobile requests to techs with mobile kits
- Bill at mobile CPL rates (15-25% premium over shop-based)
If you only do shop-based:
- Filter mobile callers early: "Just to confirm — we'd need you to bring your vehicle to our location at [address]. Does that work for you?"
- Don't pay premium CPLs for calls that won't convert to your service model
The waste is real. Shops that buy blended calls without filtering see 15-20% of their "mobile" callers cancel when they learn they have to drive in. That's $600-1,200/month on a $4,000 program going to calls that were never going to close.
Service area radius:
- Mobile service: 25-40 minute drive time from your tech's base location
- Shop-based: 15-30 minute drive time (customers won't drive far for commodity service)
Set geo-filters accordingly. Static zip codes work; real-time drive time is better if your routing supports it. VeloCalls supports both methods in its routing engine.
The Safelite Factor
Safelite has 30%+ market share in auto glass and is the preferred vendor for most major insurance carriers. When a policyholder files a windshield claim, the carrier often suggests Safelite by default.
Your pay-per-call callers are people who chose not to use Safelite. They searched for a local shop. They want an alternative. That's selection bias in your favor — but you need to understand why they're calling you.
Common reasons callers skip Safelite:
- Bad previous experience (Safelite's quality has become inconsistent as they've scaled)
- Want to support a local business
- Safelite's next available slot is 3-5 days out; they need it done today
- Heard that independent shops use better glass (this is sometimes true — OEM vs. aftermarket)
Use these in your conversion script:
- Mention OEM glass if you stock it ("We use factory-original glass, same as the dealership")
- Emphasize same-day or next-day availability
- Mention warranty terms ("Lifetime warranty on workmanship, not the 90-day you get from the big chains")
- Local ownership ("We've been in Phoenix for 14 years — you're not getting a call center")
This isn't about bashing Safelite. It's about understanding why your caller chose to search for an alternative and reinforcing that choice.
Publisher Landscape
Auto glass pay-per-call traffic runs through a smaller set of networks than HVAC or legal. The vertical is less competitive, which means fewer publishers but also less fraud.
Glass.net — The largest auto glass lead network. Decent quality, fair pricing. Works with insurance claim routing. Worth testing first.
Service Direct — Strong in home services, okay in auto glass. Lead quality acceptable. Support is slow — frustratingly so for time-sensitive glass calls. I've waited 72 hours to dispute a clearly fraudulent call. Infuriating.
HomeAdvisor / Angi — Volume exists but quality is inconsistent. Better for non-emergency glass than same-day windshield replacement. Treat as filler, not core.
Ringba marketplace — If you have your own publisher relationships, this is where you wire them up. Not where you discover new sources. For comparing call tracking platforms, see our VeloCalls pricing breakdown.
Direct SEO publishers — Search "windshield replacement [your city]" and find the non-Safelite sites ranking on page one. Contact the owners. Offer a direct PPCall deal at network-minus-20%. We've seen shops cut CPL by 30% this way while improving close rates.
Skip any network that won't show per-publisher source breakdown. If they're aggregating without transparency, at least one of those sources is underwater and dragging down your blended performance.
Seasonal Patterns
Auto glass has flatter seasonality than HVAC or roofing, but patterns exist.
Winter — Temperature swings crack windshields. Cold mornings followed by hot defrosters stress glass. Chips spread into cracks. Volume lifts 15-25% in cold-climate metros.
Spring — Construction season. More road debris. More rock chips. Steady volume.
Summer — Road trips. Highway miles. Rock damage peaks. Some operators report summer as their highest-volume quarter.
Fall — Transitional. Volume holds steady through October, tapers in November-December as driving decreases.
The mistake: flat-pacing your budget year-round when your metro has clear seasonal patterns. Phoenix is different from Minneapolis. Obviously. Track your historical volume by month and pace accordingly.
Common Mistakes That Cost Real Money
No insurance vs. cash-pay separation. Paying premium CPLs for price-shoppers while letting insurance claims blend into the same queue. These have completely different economics. Route them differently.
No damage-type qualification. Booking a sunroof job for a tech who's never done one. Taking a call for a vehicle you don't stock glass for. Qualify damage type and vehicle make/model early.
Buying mobile calls without offering mobile. Paying $38 for a caller who wants the tech to come to their office, then losing them when they learn they have to drive 25 minutes to your shop. Filter for your actual service model.
Trusting network QA. Networks score "qualified" based on duration, not conversion. A 3-minute call where the prospect is comparing prices and hangs up to call two more shops gets marked qualified. Sample your own calls. We've seen shops paying for 80-second "qualified" calls that were obviously price checks. This one drives me nuts — networks have every incentive to mark calls qualified, and you have every incentive not to pay for junk. Guess who wins if you're not listening?
No geo-filtering. Paying $35 for a caller 45 minutes outside your service area. For mobile service, that's a tech spending 90 minutes in the van for one job. For shop-based, that caller isn't driving to you anyway. Set radius filters.
Slow response on insurance claims. Insurance callers are the highest-value calls in the vertical. If they call at 8am and you call back at 2pm, they've already booked with Safelite. Answer live or lose the revenue.
For detecting repeat-caller fraud and duration-stuffing, we have a fraud filtering guide. Auto glass sees less fraud than insurance or legal, but it's not zero — especially on cash-pay calls where affiliate incentives can get weird.
TCPA Notes for Auto Glass
Auto glass is lower-risk than insurance or legal for TCPA, but not zero-risk.
One-to-one consent. The FCC's 2024 rule (enforced 2026) requires consent specific to your company, not blanket consent on a comparison site. If you're buying calls through an aggregator running a "get windshield quotes" form, the consent may not legally cover you. Get the records before paying invoices.
Callback autodialers. If your CRM auto-dials callbacks to mobile numbers, TCPA applies. Express written consent required. Audit your callback workflows.
Recording disclosure. Two-party consent states (California, Florida, Pennsylvania, others) require disclosure at call start. Make sure your whisper message includes it.
Our TCPA one-to-one consent guide covers the full requirements.
Where to Start
Standing up an auto glass pay-per-call program from zero:
-
Decide your service model. Mobile, shop-based, or both. This determines your geo-radius, CPL tolerance, and IVR scripting.
-
Build your IVR split. Insurance claim vs. cash-pay. Damage type. Mobile vs. shop visit. The more you qualify upfront, the better you route — and the more accurately you price. Our visual IVR builder makes this drag-and-drop simple.
-
Define your service area by drive time. 25-40 minutes for mobile service, 15-30 minutes for shop-based. Tighter is better until you have volume to justify expansion.
-
Sign one network and one direct publisher. Run 50 calls each. Compare close rate by call type, not blended CPL. The aggregate hides the problems.
-
Set qualification at 90 seconds. Under 60 seconds is almost always a price check. Don't pay for it.
-
Sample 20% of calls manually for the first 30 days. Listen for intent, not just duration. A 2-minute call where the caller says "I'm calling a few places" is different from a 90-second call where they say "I filed my claim with State Farm, can you do it today?"
-
Track cost per closed job by call type and publisher. Insurance claims should cost $40-55 per job. Cash-pay should cost $60-90 per job. If either bucket exceeds those ranges after 60 days, something's wrong with your routing or your source mix.
Auto glass is a cleaner vertical than most. Steady demand, clear insurance/cash split, mobile premium creates margin room.
But the same core rule applies: separate your call types or you'll blend away your profits. I'm probably beating this point to death. Don't care.
A $38 insurance-claim call closing at 85% is a $45 cost-per-job against a $400 ticket. A $25 cash-pay call closing at 30% is an $83 cost-per-job against a $250 ticket. The "cheap" calls aren't cheap. The math only works if you know which is which before you pay.
Frequently Asked Questions
What are typical CPLs for auto glass pay-per-call?
Windshield replacement calls run $25-45 depending on metro and whether the caller has insurance. Chip repair inquiries sit at $15-25 — lower ticket but faster turnaround. Insurance-claim calls (where the carrier covers the repair) command $30-45 CPLs because close rates hit 80-90%. Cash-pay shoppers — comparing quotes, price-sensitive — run $18-30 CPLs with 25-40% close rates. After-hours mobile service pushes CPLs to $35-50.
How do I separate insurance-claim callers from cash-pay shoppers?
IVR qualifying question: "Press 1 if you're filing an insurance claim for your windshield. Press 2 if you're paying out of pocket." Insurance callers close at 80-90% because they're not price-sensitive — the carrier pays. Cash-pay callers shop 3-5 quotes. Route insurance claims to your fastest dispatch. Route cash-pay to appointment setters who can compete on convenience and warranty, not just price.
What damage types should I qualify for in auto glass pay-per-call?
Windshield replacement (full crack, spiderweb, structural damage) — highest ticket, $250-600. Chip repair (small impact, no spreading) — $50-100 ticket but 15-minute job. Side and rear glass — variable ticket ($150-400), often insurance-covered. Sunroof glass — specialty, higher margin, lower volume. Qualify damage type and location in the IVR to route to techs with the right inventory and skills.
Should I buy calls for mobile auto glass service vs shop-based?
Mobile commands premium pricing — the tech comes to the customer's home or office, which is a selling point. CPLs run 15-25% higher for mobile-qualified calls. If you only offer shop-based service, filter out "come to me" callers in the IVR — they'll cancel when they learn they have to drive in. Match your routing to your service model or you'll burn 20% of spend on calls that don't fit.
Try VeloCalls for Your Vertical
AI calling + pay-per-call platform built for HVAC, plumbing, roofing, PI lawyers, Medicare brokers, and insurance. Smart routing, real-time bidding, visual IVR builder, AI conversation intelligence. Per-minute pricing — Managed starts at 4¢/min, BYOC at 2¢/min, both drop as you scale.