01 · The Situation
This Malaysia automotive aftermarket operator runs four branches across Selangor — offering window tinting, paint protection film, ceramic coating, and vehicle wrap services. By August 2025, they had been running Google Ads for years with respectable but inconsistent results. The account averaged 3–4× ROAS depending on the month, but no one could explain why a given month was up or down.
The deeper problem wasn't spend efficiency. It was attribution. With four branches running ads against overlapping service-area keywords, conversions were getting credited unpredictably across campaigns. Performance Max was running but couldn't be evaluated cleanly against Search. Phone calls — the dominant booking channel for premium automotive work — were under-tracked. And the higher-margin services (PPF, ceramic coating) were getting lost in optimization signals dominated by entry-level window tinting.
The owner's instinct was right: the account was leaking. The dashboard just couldn't show where.
02 · The Diagnosis
We ran our standard 14-day audit across the four-branch account. Four issues surfaced:
Call tracking was incomplete. Phone bookings — roughly 40% of revenue for premium PPF — were tracked as anonymous calls without service-level attribution. Google Ads couldn't distinguish a RM 8,000 PPF inquiry from a RM 250 sun visor tint.
Performance Max was optimizing toward the wrong signal. Without offline conversion imports, PMax was treating low-value form fills the same as high-margin appointment bookings. The algorithm was getting smarter at attracting the wrong customers.
Branch-level structure was missing. All four branches shared a single campaign structure, which meant strong PPF conversion at one location was masking underperformance on the same service elsewhere. We couldn't see which branch needed help.
The package ladder wasn't reflected in bidding strategy. The highest-margin product tier was getting the same bid treatment as the entry-level tier. The algorithm wasn't incentivized to chase high-value leads.
03 · What We Did
Over the first 30 days, we rebuilt the tracking layer. Dynamic number insertion routed calls by branch and surfaced service-level intent. Offline conversion imports from the booking system fed revenue data back to Google Ads, so the algorithm started optimizing for revenue, not form fills. Server-side conversion tracking via Google Tag Manager closed the iOS attribution gap.
Then we restructured the campaigns by buyer intent rather than service category. PPF and ceramic coating — the considered-purchase services — got their own Search campaigns with dedicated landing pages and call-extension prominence. Window tinting, the lower-funnel service, stayed in Performance Max where its higher volume could feed the algorithm cleanly.
Branch-level segmentation gave each location its own asset group and budget pacing. Underperformance on PPF became visible within two weeks at one branch; strength at another became the template we cloned.
By March 2026, the system was compounding. February delivered 956 conversions at RM 37 CPA. March hit 1,402 conversions at RM 27 CPA — a 47% increase in volume alongside a 27% reduction in cost. Total March revenue: RM 294,459 on RM 38,000 spend.
04 · The Results
Selected outcomes from March 2026 — the operator's strongest month on record:
- 7.75× ROAS across all four branches
- 1,402 monthly conversions at RM 27 cost per conversion
- +47% conversion growth month-over-month from February
- -27% CPA month-over-month
- PPF and coating services — historically underweighted in the account — now represent 35% of monthly revenue, up from roughly 22% pre-restructure
- Branch-level reporting now identifies underperformance within two weeks, not two quarters
05 · What's Next
The current focus is two-fold. First, expanding the offline conversion feedback loop to include lifetime value signals — customers who return for ceramic coating after a tint job represent significantly higher LTV, and we want the algorithm to bid for them differently. Second, testing a brand-protection campaign structure to defend against competitors bidding on branded terms now that the account is producing visible results.
The system isn't finished. But it's measurable, and it's compounding.
Running multi-branch Google Ads in Malaysia? See Google Ads Agency Malaysia for the tracking-first playbook behind this engagement, or the automotive Google Ads insight for the calls / WhatsApp / package-tier operating system.
“Zenos rebuilt the way we measure success. Every ringgit we spend now ties back to a specific service, branch, and outcome. We're not guessing anymore.”