What GA4 captured · all events lifetime
scrollEvery time a user scrolled down the pagepage_viewEvery page load, including refreshes and back-buttonfirst_visitEvery new visitor — including ad clickers, organic browsers, botsopen_formUser started filling a form. Doesn't mean they finished.leadbooster_openedPipedrive chat widget opened. Says nothing about lead quality.phone_clickUser clicked the phone number to call. Real intent.form_filledForm completed and submittedsubmit_formAlternate handler for form submissionsleadbooster_leadPipedrive chat actually qualified the conversation as a leademail_clickUser clicked the contact email linkCurate the conversion taxonomy first, then run the campaigns.
The account inherited 50+ conversion actions accumulated by previous setups — Universal Analytics imports, defunct landing-page goals, “Smart Goal” auto-imports that count any 5-page session as a conversion. We went through each one, kept the five lead events that actually represent intent (phone clicks, form fills, qualified Pipedrive chat leads, email clicks), and demoted the rest.
Without that curation, the account would have reported ~6,500 form-opens and 4,000 chat-opens as "conversions" in Google Ads — inflating the visible CPA improvement while quietly destroying every bid decision based on it.
Before we took over, the only conversion actions firing were two form-submission handlers. Phone calls, WhatsApp contacts, and chat-qualified leads were happening in the business but invisible to the ad platform. We added phone_click, WhatsApp contact, Pipedrive LeadBooster, and email_click as tracked conversions — so every channel that actually produced a lead got attributed correctly.
Result: Google’s bidding algorithm could finally see the full picture. CPA stopped being a half-truth.
First-month signal: +207% conv, +117% conv-rate, −17% CPA. What mattered: those held for 31 more months.
Once tracking was clean, the work shifted to maintenance: weekly search-term review, monthly Looker Studio reports, seasonality tuning around Chinese New Year and year-end mortgage cycles, periodic competitor-overlap analysis. We built dedicated landing pages for each ad cluster — tighter conversion paths than the broker’s own site, which kept the funnel performant independent of the main-site UX. We ran a US-targeted home-equity sub-campaign for three months in 2024 — killed it at SGD 244 spend with 0 conversions because the data said no.
CPA stayed in a SGD 30–60 band for 32 straight months. Year averages: 2023 SGD 43.81 · 2024 SGD 41.98 · 2025 SGD 41.81. Within SGD 2 of each other across three years.
I have witnessed firsthand his talent for managing my Google Ads account, where his strategic thinking and attention to detail played a pivotal role in the success of our team.
32 months, 1,602 qualified leads, no surprises.
From October 2022 to May 2025 — when Wayne wound down ad spend to evaluate his marketing — the account ran in a tight, predictable band. Same five conversion actions every month. Same SGD 40-something CPA every year.
Boring, predictable, attributable. That's what good PPC looks like.
Every engagement starts with a 30-minute audit — free, no slides, just a screen-share through your account.
Book the audit →