Your blended CAC looks healthy. That might be the most expensive illusion on your P&L.
Most e-commerce teams track customer acquisition cost one of two ways: blended CAC (total marketing spend divided by total new customers) or channel CAC (spend per channel divided by customers attributed to that channel). Both numbers are useful. Both are also routinely misread, and the misreading has a predictable cost: budget flows to channels that claim credit rather than channels that create customers. If you're deciding where next quarter's money goes based on only one of these numbers, you're almost certainly deciding wrong.
Blended CAC is the honest number. It doesn't care which platform's attribution model claimed the sale, whether the click was last or first, or how many touchpoints a customer wandered through. Total spend in, total new customers out. It's hard to game, which is exactly why it's valuable.
Its strengths are real:
Its weakness is just as real: blended CAC tells you nothing about why it moved. When it creeps up 20%, you can't tell if Meta got expensive, your affiliate mix shifted toward coupon publishers, or organic demand simply softened.
Channel CAC is the operational number. You need it to negotiate rates, set affiliate commissions, and decide which campaigns to scale. But it inherits every flaw of the attribution model behind it, and most attribution models are generous to the last click.
The predictable distortions:
Blended CAC tells you whether the machine is working. Channel CAC tells you which part to inspect. Confusing the two is how brands end up scaling their least incremental channel.
The two metrics are a system, not a choice. The practical workflow: treat blended CAC as the source of truth for budget-level decisions, and treat channel CAC as a diagnostic that must be validated before you act on it.
A simple discipline that works:
None of this requires a data science team. It requires a rhythm: weekly channel CAC reviews for operational tuning, monthly blended CAC reconciliation against contribution margin, and quarterly holdout or geo tests on your two largest channels. Brands that run this cadence consistently find at least one channel whose true acquisition cost is 30-50% worse than reported — usually a channel positioned at the bottom of the funnel, collecting credit for customers who were coming anyway.
That's the point of measuring acquisition cost in the first place. The goal was never a prettier dashboard; it's knowing which dollars create customers who wouldn't have arrived otherwise. Blended CAC keeps you honest about the total. Channel CAC — tested, reconciled, and treated with appropriate suspicion — tells you where the next incremental customer actually comes from. Get both right and budget allocation stops being a negotiation between platform dashboards and starts being a decision about real revenue.