- For windows replacement lead generation, the Google LSA vs Search Ads question isn’t about cheaper raw leads. It’s about which channel produces an in-home consult that actually shows.
- The HVAC default of 60/40 LSA-to-Search by spend overpays for booked consults by 20 to 35% in windows. The right 2026 default for most metro installers is closer to 30/70 Search-heavy.
- LSA’s dispute taxonomy was built for emergency trades. It does not refund the four junk-lead types that quietly tax windows accounts: price-shoppers, renters, single-window repair callers, and insurance-claim chasers.
- Search-with-scheduler leads typically book consults at 55 to 70%. LSA-sourced calls and form-fills typically book at 35 to 45%. That show-rate gap inverts the apparent CPL advantage.
- The operational fix is offline conversion import (now via Google’s Data Manager API) passing consult-attended back to Smart Bidding on a 5 to 10 day lag, not raw form-fills.
If you run a windows replacement business spending $40K to $200K a month on Google, you’ve probably been told the same thing your HVAC neighbor was told: load up Local Service Ads, sprinkle some Search on top, ride the Google Guaranteed badge. That advice will quietly bleed you in 2026.
Windows is not an emergency trade. The sales cycle is 30 to 90 days, the close happens in a kitchen, and the lead-to-revenue path runs through an in-home consult that has to show up. The right windows replacement lead generation Google LSA vs Search Ads split isn’t 60/40 LSA-heavy. It’s closer to 30/70 Search-heavy. Here’s the math.
Windows Replacement Is a Planned Purchase, and LSA’s Economics Assume Otherwise
LSA was designed for verticals where someone with an active problem calls the first credible local result. HVAC at 4pm in July. Plumbing at 11pm with water on the floor. Locksmith in a parking lot. The lead is hot, the decision is fast, and the close happens on the same call.
Windows doesn’t work that way. The funnel has five stages, and only one of them matters for profit:
- Lead (form fill, LSA message, inbound call)
- Qualification call (homeowner status, scope, timeline, decision-makers)
- Scheduled consult (calendar booked, both spouses present)
- Consult-attended (rep in the home, measuring, quoting)
- Sold job (contract signed, deposit collected)
The conversion event that controls your P&L is stage 4: consult-attended. Stages 1, 2, and 3 are vanity metrics if your sales team can’t get inside the house. The gap between “lead” and “consult-attended” is where the HVAC playbook breaks for windows.
Why show rate is the metric that quietly controls profitability
A 40% show rate versus a 65% show rate doesn’t just change your CPL. It changes which channel is profitable at all.
If you’re paying $120 per LSA lead and 40% show, your cost per shown consult is $300. If you’re paying $200 per Search lead with a scheduler and 65% show, your cost per shown consult is $308. Compare close rates on those two pools, and the channel that looked cheaper on a CPL dashboard is the channel that’s underwater on booked-job ROAS.
This is the calculation almost no one runs at the channel level. It’s also the only one that matters.
LSA’s Dispute System Was Built for Emergencies, So Windows Contractors Absorb the Asymmetry
This is the operator-level point most articles skip. Google’s LSA lead dispute system was designed around binary failure modes that work cleanly for emergency trades. The categories you can actually win disputes on are things like “wrong service area,” “spam,” “not the service I requested,” and “duplicate.”
A plumber gets a leak call. The lead is either a plumbing emergency or it isn’t. Binary. Disputable.
Windows replacement is a planned purchase. Google’s dispute taxonomy doesn’t have clean categories for the failure modes that actually waste a windows contractor’s money:
- Price-shoppers gathering three quotes for a refinance-funded project they haven’t committed to
- Renters who can’t authorize the work and shouldn’t have been routed to a replacement contractor
- Single-window repair callers when you sell whole-house packages starting at $14K
- Insurance-claim chasers looking for someone to write up storm damage
None of these have a clean dispute category. Google does not refund them. An HVAC contractor running the same LSA spend never feels this tax because their failure modes sit inside Google’s taxonomy. A windows contractor eats every one of these as unrecoverable spend. From the planned-purchase accounts we’ve audited, the unrecoverable junk rate runs roughly 2 to 3x what an emergency-trade operator sees on the same channel.
How to calculate your effective junk-lead tax
Pull 90 days of LSA leads. Tag every lead with one of three labels: qualified, recoverable junk (disputed and credited), or unrecoverable junk. The formula:
Put that on your dashboard. It’s the reason raw LSA CPL lies to you.
Booked-Consult CPL Inverts the Apparent LSA Advantage
Here is the math the top SERP results skip. The ranges below reflect what we typically see across planned-purchase home-improvement verticals in 2026 mid-sized metros:
| Metric | LSA | Search + Landing Page w/ Scheduler |
|---|---|---|
| Raw lead CPL | $90 to $140 | $160 to $240 |
| Lead-to-consult-scheduled rate | 55 to 65% | 70 to 80% |
| Show rate on scheduled consults | 60 to 70% | 80 to 90% |
| Effective booked-consult CPL | $220 to $380 | $260 to $340 |
| Average ticket on closed jobs | $11K to $13K | $13K to $16K |
The headline LSA CPL is half the Search CPL. The booked-consult CPL is roughly tied, with Search edging ahead on the high end. The average ticket on Search-sourced jobs runs higher because the scheduler step on a landing page does qualification work LSA structurally cannot. A homeowner who picks a time, confirms their address, and reads your three-step process is a different buyer than someone who tapped your LSA card and said “are you guys local?”
The booked-consult CPL formula every windows operator should run monthly
Run this per channel, per month. If you’ve never seen the number, you’ve never actually compared LSA and Search.
The max profitable CPL works backwards from the same place:
Max profitable CPL = (Gross profit per job × close rate on attended consults) ÷ (show rate × lead-to-schedule rate)
At a $13K average ticket, 35% gross margin, and a 30% close rate on attended consults, your max profitable cost per attended consult is roughly $1,365. Divide by your show rate to get max profitable cost per scheduled consult, and divide again by your lead-to-schedule rate to get max profitable raw CPL. This is the only ceiling that matters.
For a deeper look at why two leads at very different prices can produce the same number of jobs, see Why a $75 Windows Lead and a $160 Windows Lead Book the Same Number of Jobs in 2026.
Pass Consult-Attended Back to Smart Bidding. Stop Optimizing on Form-Fills.
Once you accept the show-rate gap, the operational fix is straightforward: stop letting Smart Bidding learn on raw leads.
Most windows accounts pass form-fill or call-as-conversion back to Google Ads. That tells Smart Bidding to find more cheap leads, which is exactly the wrong instruction. The bidding model doesn’t know which leads showed up to a consult. It just knows what looked like a conversion at click time.
The fix is offline conversion import. Pass consult-attended back to Google Ads on a 5 to 10 day CRM lag. Deprecate form-fill as the primary conversion event, or downweight it to a secondary signal. Smart Bidding will rebalance keywords toward the ones that produce shown appointments. Raw-lead CPL will appear to rise. Booked-job ROAS will climb.
This matters more in 2026 because Google is migrating offline conversion imports to the Data Manager API. If your CRM-to-Ads pipe was wired to the legacy OCI endpoint, you’re rebuilding it this year anyway. Do it correctly. Read our breakdown of the Data Manager API migration for lead gen accounts before your dev team starts.
The 5 to 10 day CRM lag is a feature, not a bug
A windows consult doesn’t happen the day the lead comes in. The qualification call is day 1. The scheduled consult is day 3 to 7. The shown appointment hits your CRM on day 4 to 10. That lag used to scare marketers into optimizing on the faster signal.
Smart Bidding can handle it. Google’s bidding models are built to learn on delayed conversions when you set the conversion window correctly. Set the click-to-conversion window to 30 days. Set the attribution model to data-driven. Let the lag breathe.
Call duration thresholds and IVR pre-qualification
For the calls that do come in, set minimum billable thresholds correctly:
- LSA calls: 90 seconds minimum. HVAC defaults to 60. Windows needs longer because a real lead has to confirm homeowner status, rough scope, and timeline before the call is worth your sales team’s time.
- Search calls: 60 seconds is acceptable if your landing page has done qualification work upfront.
Layer an IVR pre-qual that asks two questions: “Are you the homeowner?” and “Are you looking to replace windows in the next six months?” That single layer cuts unrecoverable junk meaningfully and improves the signal Smart Bidding learns on. For the broader call-tracking architecture, see how a CallRail threshold set to 60 seconds misses real dispatch-grade calls.
When LSA Should Rise Above 30%, and When It Should Drop Below 20%
The 30/70 split is a default for metro mid-market installers. The right answer flexes by territory, review velocity, and competitor density.
When LSA should be 40 to 50% of spend
- Rural and exurban markets where Search CPCs are low and LSA competition is thin. The dispute asymmetry still applies, but the volume LSA delivers in thin auctions is hard to replace.
- Single-location installers with 200+ Google reviews and high velocity that earns prominent Google Guaranteed badge placement. The badge does real work in a sparse competitive set.
- Territories where mega-budget national competitors have bid Search CPCs to $45+ but can’t dominate LSA. Renewal by Andersen, Window Nation, and Champion have national Search budgets that distort regional auctions. They’re often less efficient on LSA because location eligibility and per-location review velocity work against multi-location operators.
When LSA should drop to 15 to 20%
- Dense metros with multiple LSA competitors at similar review counts. Click share on LSA shrinks fast in saturated markets.
- Multi-location installers diluting reviews across territories. The badge doesn’t carry the same weight without concentrated review velocity per location.
- Accounts where the in-home sales team’s show rate is structurally weak. If you can’t get above 50% show rate on Search, LSA’s lower-quality leads will be even worse. Fix the sales process before reallocating budget.
How National Windows Brands Distort Your Regional Search Auction
If you operate inside a market where Renewal by Andersen, Window Nation, Champion, or Power Home Remodeling are spending, your Search CPCs are not what they would be in an open auction. Top-of-funnel terms like “replacement windows near me” and “window installation [city]” can run $35 to $60 per click in distorted markets.
You cannot beat them on bid. You can beat them on three things:
- Long-tail intent. “Triple-pane window cost,” “vinyl window replacement quote,” “egress window installation” carry intent the broad terms don’t. CPCs are 40 to 60% lower.
- Local proof. Your landing page should lead with neighborhood-specific install photos, named local reviews, and a same-week consult promise. National brands can’t match this credibly.
- Smart Bidding trained on consult-attended. When your bidding model is optimizing on shown appointments and theirs is optimizing on form-fills, you win on every keyword where intent and qualification matter more than reach.
This is why the Search side of your account is the leverage point, not LSA. LSA is a fixed-format auction with limited ranking levers. Search is where account architecture, conversion signal, and landing-page qualification compound.
Related guides
- windows replacement lead generation — LSA vs Search split also driven by sit-rate economics
Frequently Asked Questions
What is the right LSA-to-Search budget split for a windows replacement contractor in 2026?
For most metro mid-market installers, roughly 30/70 LSA-to-Search by spend. The HVAC default of 60/40 LSA-heavy overpays for booked consults by 20 to 35% in windows because the in-home consult sales cycle exposes LSA’s dispute asymmetry and show-rate gap. Rural markets, single-location installers with 200+ reviews, and territories where national windows brands have bid Search to $45+ CPCs can flex LSA up to 40 to 50%.
Why do my LSA leads book consults at a lower rate than my Search leads even though LSA is cheaper per lead?
LSA leads come in as inbound calls and form-fills with no qualification step in front of them. Search leads coming through a landing page with a scheduler self-qualify by picking a time, confirming their address, and reading your process. That extra friction filters out price-shoppers, renters, and single-window repair callers before they reach your sales team. Typical gap: LSA books at 35 to 45%, Search-with-scheduler books at 55 to 70%.
Which conversion event should I pass back to Google Ads Smart Bidding for windows replacement?
Consult-attended, not form-fill or call-as-conversion. Form-fill optimization tells Smart Bidding to find more cheap leads, which is the wrong instruction when 40 to 50% of those leads no-show. Pass consult-attended via offline conversion import (now Data Manager API) on a 5 to 10 day CRM lag. Raw CPL will rise, booked-job ROAS will rise faster.
How do I dispute LSA leads for windows when Google’s system doesn’t recognize price-shoppers as a valid category?
You largely can’t, and that’s the structural problem. The realistic disputes are wrong service area, spam, duplicates, and not-the-service-requested (a single-window repair caller when you only do whole-house). Price-shoppers and renters typically don’t qualify. The operational answer isn’t a better dispute workflow, it’s reducing LSA’s share of your mix and putting more spend into Search where landing-page qualification does the filtering Google’s dispute system won’t.
What minimum call duration should I set for windows replacement LSA calls?
90 seconds minimum, not the HVAC default of 60. A real windows lead has to confirm homeowner status, rough scope, and timeline before the call is worth anything to your sales team. Anything shorter is usually a price-shopper or wrong-fit caller. Layer a two-question IVR pre-qual (homeowner status, six-month timeline) in front of the call to push that signal earlier.
How do I model the maximum profitable CPL for my windows business?
Start with average ticket × gross margin × close rate on attended consults. At $13K ticket, 35% margin, and 30% close rate, max profitable cost per attended consult is roughly $1,365. Divide by your show rate (say 60%) to get max profitable cost per scheduled consult (~$820). Divide again by your lead-to-schedule rate (say 65%) to get max profitable raw CPL (~$530). That’s your ceiling, not a target.
When does it make sense to shift more budget to LSA after my review count grows?
When sustained review velocity (15+ new reviews per month per location) pushes you into the top 3 LSA ranking slots in your service area and click share climbs above 25%. Below that threshold, you’re paying for impressions that go to the operator above you. The badge does real work in sparse markets and on mobile, but only after review density crosses a competitive threshold. Audit monthly.
Build the Right Split for Your Market
The 30/70 Search-heavy split is a starting point, not a prescription. The actual optimal mix for your business depends on territory density, average ticket, sales team show rate, review velocity per location, and which national brands are distorting your auction. The booked-consult CPL math is what tells you the answer, not anyone’s blog post.
If you’re spending $25K to $500K a month on Google for windows replacement and your dashboard still reports raw CPL as the primary KPI, you’re guessing. Talk to our pay-per-call and lead-buying team about your windows replacement economics. We’ll model booked-consult CPL against your actual close rate and average ticket, audit your current Smart Bidding conversion events, and recommend a custom LSA-to-Search split for your specific markets. Book a free strategy call and bring your last 90 days of Google Ads and LSA data.
