What happened when a mature window-replacement operator handed over a dormant digital program (paid search, LSA, Maps, organic, and the measurement pipeline underneath all of it) to Blue Glass Insights.
The client is an established residential window and exterior-door installer serving a multi-metro market across the mountain-west region. They install replacement windows, sliding glass doors, and exterior systems. A high-ticket category where one job averages $8,000–$30,000+ and the entire sales cycle hinges on an in-home demo and quote.
Before engaging Blue Glass Insights, the company's lead mix leaned on offline channels: home shows, canvass teams, referrals, and previous-customer rehash. Digital was a small slice of the pie, and the paid search account had been paused for 9 of the 12 months of 2025. A strong sales team. Tired infrastructure. No real visibility into what was working.
In the first two weeks we audited the existing digital stack end to end: Google Ads, LSA, Google Business Profile, the WordPress site, GA4, and the LeadPerfection CRM. What we found was less a marketing problem than an operational one.
A working channel was sitting at zero, while competitors collected the demand.
A "Flying Scripts" plugin deferred the gtag script in a way that hid three months of conversion data from Google Ads. Spend was flowing; signal wasn't returning.
Lighthouse mobile score of 61. Total Blocking Time over 5 seconds. 80% of main-thread work happening before the page was usable.
42% of all tracked calls came in through GBP and Maps. No review-response cadence, no posts, no service-area schema.
Customers who clearly found the company through Google were being entered as "Previous Customer" or "Rehash", washing out digital's contribution and undercutting the case for reinvestment.
Half of iOS prospects saw a 2.0★ reputation, while the real Google rating was 4.7★ across 800+ reviews.
Turn on the channels that were off. Fix the measurement under them. Make sure every dollar going out the door had a number coming back attached to it.
The Google Ads account had been dark for most of 2025. We restarted it from January 1, 2026 with a tightly-scoped campaign structure: replacement-window queries, service and warranty queries, and a separate brand-defense layer.
Landing-page-specific tracking went live week one. We ran a weekly 74-point audit against the account through April.
LSA was active but uncoached: wrong service categories, no review velocity, no proactive lead dispute discipline. We rebuilt the category mapping and set up weekly LSA lead reconciliation against the CRM to catch un-tagged sales.
We turned the GBP profile into a managed asset: schema markup across service-area pages, scheduled posts, automated review responses on the 4.7★ × 800+ Google profile, and an active dispute process targeting the Apple Maps Yelp drag.
We rebuilt the measurement stack from the ground up. Diagnosed and fixed the Flying Scripts gtag bug, set up GA4 + Google Search Console + Google Ads OAuth, deployed a fail-safe key event for conversion tracking, and built an end-to-end LP API → CallRail → GA4 → GSC → Google Ads pipeline.
In parallel on the site itself: WebP conversion across hero assets, a custom mu-plugin for inline CSS delivery, and CLS fixes on the homepage and service pages.
The last stream was the unglamorous one. We built a reconciliation report that cross-references every "Previous Customer" sale in the CRM against CallRail history, web-form submissions, and the channel-specific tracker registry.
Goal: surface every sale where a previous-customer tag was hiding a Google-sourced inbound call. Not a huge slice of total revenue, but it changed how the client credits digital channels internally and made the case for reinvestment defensible with line items, not impressions.
Period: January 1 – May 15, 2026. Revenue figures are verified contract amounts from the LeadPerfection CRM, cross-referenced against CallRail and the channel tracker registry.
“When we turned paid search back on after nine months of silence, I braced for the usual slow trickle, but seeing the revenue hit the books within the first forty-five days was the exact moment the pressure lifted.”
The hard part of home-services growth in 2026 isn't getting in front of homeowners. It's knowing, at the line-item level, which channels are actually producing the revenue, which leads were already going to convert, and which dollars are paying for themselves vs. quietly underwriting a stalled paid account.
What changed for the client isn't that they got new ideas about marketing. They got the channels they were already paying for, and the ones that had been turned off entirely, to actually show up as revenue in the CRM, with the math to defend reinvestment.
Most home-services operators we meet don't need more leads. They need someone to find the operational clogs (the broken tracking, the mis-tagged CRM entries, the channel that's been quietly off) and shorten the path from inquiry to signed contract.