Scenario: A Fintech CMO Separates New-Logo and Expansion Spend to Reduce CAC

This scenario illustrates a typical pattern observed across our customer base. Specific numbers are representative ranges drawn from public benchmarks (cited inline), not from a single named customer.

Problem

A mid-market fintech team is buying media for two jobs inside the same account: acquire new customers and expand relationships with people who already use the product. The campaigns share conversion events, audiences, budgets, and reporting. That setup makes the account look efficient while hiding the decision the CMO actually needs to make: which spend creates new demand, and which spend merely harvests or expands demand that already exists?

The failure is not simply weak targeting. It is a measurement boundary problem. An existing customer who returns to compare a new product can look like a new acquisition. A high-intent prospect who converts after several exposures can be credited to the last campaign rather than the segment and message that created the opportunity. When both groups optimize toward the same event, the platform receives a blended signal and the finance team receives a blended CAC. Neither view is sufficient for budget allocation.

The team therefore treats “paid waste” as spend that cannot be assigned to a defined commercial job. The first job is new-logo acquisition: reach people with no qualifying relationship and optimize toward a verified first conversion. The second is expansion: reach eligible existing customers with a product, limit, or use case that is genuinely incremental. The boundary must exist in the data before it can exist in the campaign structure.

The broader benchmark supports making that boundary explicit, but does not prove a result for this scenario. McKinsey reports that granular personalization can be associated with a 5–15% (The New Growth Game) revenue lift and a 10–30% (The New Growth Game) improvement in marketing efficiency. Those are external reference ranges, not a forecast for this team.

Solution

The CMO starts with a customer-state contract rather than a new campaign build. The CRM becomes the authority for four states: prospect, new customer, eligible expansion customer, and suppressed customer. Each state has a definition, an owner, a refresh rule, and a reason code. “Existing customer” is not a permanent label: eligibility depends on product ownership, recency, consent, service status, and whether the next offer is incremental.

The team then sends those states into the ad platforms as separate first-party audiences and server-side conversion events. The new-logo lane excludes current customers and optimizes toward a verified first conversion, not a soft lead or a page view. The expansion lane excludes prospects and uses product-specific events such as an approved upgrade, funded account, or activated add-on. The event names, value fields, and customer-state fields are documented so that platform reporting and CRM reporting use the same vocabulary.

Campaign structure follows the data contract. New-logo campaigns receive acquisition creative, prospecting exclusions, and a new-customer outcome. Expansion campaigns receive product-specific creative, existing-customer eligibility rules, and an incremental outcome. Budgets are not moved because one lane has a cheaper surface-level conversion; they are moved only after the team checks qualified conversion rate, downstream revenue, and the share of conversions that would have happened without the campaign.

This is deliberately more controlled than simply opening targeting. Public Meta tests show why a cleaner signal can matter: Axis Max Life Insurance reported a 10–17% reduction in cost metrics and a 40% increase in quality leads after adding the Conversions API to its campaign setup (Meta case study); Catawiki reported a 35% lower cost per completed registration in an acquisition test (Meta case study); and Mosh reported a 44–53% reduction in acquisition-cost metrics after adding creator-led partnership ads to its usual creative (Meta case study). Together, those public results form a 10–53% (Meta case studies, Catawiki, Mosh) reference band across different businesses and interventions, not an expected outcome for this fintech scenario.

Measurement is designed before launch. The team records baseline spend, qualified conversions, first-product revenue, expansion revenue, assisted conversions, and suppression coverage for each lane. It keeps the conversion definition stable during the comparison window. The same counterfactual discipline appears in RevenueProven’s retargeting holdout playbook. Where volume allows, it holds out a small, pre-defined cohort from expansion messaging and compares incremental outcomes against an eligible control. Where a holdout is not practical, it uses geo, account, or time-based comparisons and labels the result as directional rather than causal.

The operating cadence is equally specific. Marketing owns audience logic and creative hypotheses. RevOps owns CRM state changes, event delivery, and exclusions. Finance validates revenue definitions. For the attribution failure mode that appears when those definitions drift, see the fintech last-touch rebuild. The CMO reviews the lanes separately, then reviews the combined account only as a reconciliation view. A weekly exception report flags conversions with contradictory states, stale eligibility, missing value, or a source campaign that cannot be mapped to a commercial job. Those exceptions are fixed before budget is reallocated.

Results

The scenario has no named customer and therefore no invented post-intervention result. The correct result is a measurement plan that can show a baseline-to-post gap without mixing acquisition and expansion. The primary scorecard should report new-customer CAC, qualified first conversions, incremental expansion conversions, revenue by customer state, suppression coverage, and the proportion of spend assigned to a defined job.

The external benchmark envelope gives the team a disciplined range for interpretation: McKinsey cites 5–15% (report) revenue lift and 10–30% (report) marketing-efficiency improvement for scaled personalization; the Meta examples above span 10–53% (Axis Max Life, Catawiki, Mosh) lower cost metrics across separate tests. These ranges are context for a decision, not evidence that this hypothetical team achieved them.

A credible internal result would show the new-logo and expansion lanes separately, include the comparison window and denominator, and state whether the measurement was experimental or directional. If new-logo CAC falls but first-product revenue does not, the team has not solved the problem. If expansion revenue rises but the holdout does not move, the campaign may be harvesting existing intent rather than creating incremental value. The conversion comes from making those distinctions visible: one problem, one data contract, two clearly defined commercial jobs, and a result that can survive finance review.