The last recorded source wasn't the whole customer story.
Paid social was generating enquiries, yet the final customer-source report told a different story. I led an investigation into how leads, follow-up and enrollments were being counted—and what leadership could reasonably conclude from the evidence.
Enrolled families with an earlier Meta lead record
Monthly planned additional spend directed to a combined-channel test
Key areas investigated: source definitions, connected systems and family-level reconciliation
Was Meta underperforming—or was the report incomplete?
A US multi-location consumer-services business acquired leads through Google and Meta. Salesforce supported follow-up, another system tracked enrollments, and call tracking supplied another view of the customer journey.
Google appeared to account for far more final enrollments than Meta. Leadership had to decide whether to maintain, reduce or expand its investment.
The critical question wasn't which platform deserved more credit. It was whether the reports were describing the same events, using the same definitions and counting unique families consistently.
One customer. Multiple touchpoints. Different answers.
A family could first enquire through a Meta advertisement, receive calls and messages, and later return through Google or by phone before enrolling.
An initial-source report and a last-recorded-touch report might both be internally consistent while answering fundamentally different commercial questions.
Before comparing channels, fix what you're counting.
Clarify source definitions.
I worked through initial acquisition versus later recorded interactions so that reports weren't being treated as interchangeable when they described different points in the journey.
Connect the relevant evidence.
I identified and led a BigQuery-based reporting approach involving Salesforce and Nimbata call-tracking information. I defined business requirements, coordinated specialist implementation and checked whether the outputs reflected the intended metrics.
Reconcile unique enrolled families.
Enrollment histories contained repeated status records. I supervised deduplication and quality checks so the analysis counted families rather than treating every status event as another customer.
Better evidence changed the budget conversation.
The investigation identified approximately 460 enrolled families with an earlier Meta lead record whose journeys weren't adequately represented by a simple final-source view.
Rather than cutting Meta based on that view alone, the owners maintained their existing Meta investment for the quarter and directed $4,000 per month of planned additional spending toward a test involving both Meta and Google.
Important distinction: An earlier Meta lead record establishes a documented association, not proof that Meta advertising independently caused the enrollment or generated a measured return on ad spend.
Strategic ownership, technical coordination and quality assurance.
What I personally led
Diagnosing the commercial reporting question, identifying the integration approach, specifying requirements, coordinating technical specialists, supervising deduplication, reviewing customer-level reconciliations and interpreting the findings for decision-makers.
Technical specialists performed the underlying platform configuration. This was a coordinated technical and commercial effort, not a claim that I personally coded the integrations.
Your marketing reports leads. Can you confidently connect them to customers?
My three-week Marketing-to-Revenue Clarity Sprint investigates where measurement becomes unreliable and identifies what needs to improve before your next investment decision.