How a Boise med spa cut CAC 42% with the Acquisition System

The client had run Meta and Google ads for 18 months through two different freelancers, each managing their own channel in their own ad account. Meta CAC ran around $145; Google around $210. Blended CAC sat at $187 — high enough that the marketing team was getting weekly questions from the CFO about whether paid was even profitable.

Worse: nobody could connect ad spend to actual booked appointments. Meta reported its conversions, Google reported its conversions, the booking platform reported bookings, and the three numbers didn't reconcile. On a good week the three sources disagreed by 15%; on a bad week by 40%. Decisions were getting made on platform-reported metrics that everyone quietly suspected were inflated by view-through conversions and cross-channel double-counting.

The freelance-per-channel model also produced a subtler problem: neither freelancer was accountable for the outcome the business actually cared about — booked appointments — because each optimized their own channel's platform-reported KPI. The Meta freelancer optimized for cost-per-lead as Meta measured it. The Google freelancer optimized for conversions as Google measured it. Nobody was optimizing for a med spa consult that showed up on the calendar.

We deployed the Acquisition System over the standard 30-day onboarding. The Meta and Google accounts were restructured into a unified architecture sharing creative, audiences, and conversion events. Consent-mode was properly configured on both platforms so first-party conversion signal continued flowing even where users opted out of cross-site tracking.

A first-party UTM taxonomy and a GHL-side conversion tracking layer were added so every booked appointment mapped back to the specific ad, campaign, and creative that drove it. The booking platform's calendar was wired to fire a GHL webhook on every confirmed appointment, timestamped and tagged with source. That became the single source of truth against which both platforms' pixels were reconciled weekly.

The creative iteration workflow shifted from monthly batches (the freelancer model) to AI-assisted weekly variants tied to automated performance feedback. Underperforming creative got paused inside 7 days; winning creative got scaled inside 14. Bid and budget management moved to daily reallocation across campaigns based on the unified attribution layer — not weekly meetings looking at platform-reported metrics that disagreed with each other.

The root cause of the client's $187 blended CAC wasn't ad spend — it was that neither channel could be optimized against the metric that mattered. Meta and Google each reported conversions using their own last-touch logic, then the booking platform reported a third number based on its own calendar events. Reconciliation happened in a spreadsheet, monthly, in arrears.

We replaced the spreadsheet with a first-party pipeline. Every ad in every campaign was tagged with a structured UTM (source · medium · campaign · content · term). Every landing page fired a GHL contact event with those UTMs preserved on the contact record. Every calendar booking fired a second GHL event that carried the contact ID forward. The result: every booked appointment carried its original ad-touch UTMs, timestamped, in a single database.

That first-party pipeline became the ground-truth conversion signal fed back to both Meta (via CAPI) and Google (via Enhanced Conversions). Instead of platforms optimizing against their own pixels' inflated view-through conversions, they optimized against actual booked appointments. Reported CAC on both platforms converged toward true CAC inside 30 days, and reported ROAS became a metric the CFO trusted.

By day 90, the unified attribution layer was producing reconcilable numbers — every booked appointment had a clean source of truth, and the CFO stopped asking the weekly profitability question.

More importantly, blended CAC dropped from $187 to $108 — a 42% reduction — while monthly bookings grew 28% over the same window. Same ad spend, more bookings, lower CAC per booking. The largest single driver was the shift to first-party conversion signal on both platforms: Meta's algorithm began optimizing against booked-appointment quality rather than lead-form fills, which reduced spend on unqualified prospects.

The system continues to run on the same architecture today. The client added the Revenue Engine in month 5 to convert one-time visits into recurring memberships. Blended CAC has continued to compress modestly as the attribution model has accumulated more historical training data — an effect we expect on every engagement where the client stays committed to first-party measurement.