Why your channel report and your revenue report disagree

Open almost any analytics tool and the default channel report ranks by sessions or users. It is the most-looked-at table in marketing, and it is answering a question almost nobody actually has: where did volume come from?

The question people have is where the revenue came from. Those two lists are not variations on a theme. They frequently invert.

Take a set of channels over a month. Paid social sends 18,400 sessions — comfortably the largest single source, and the one with the largest invoice attached. Organic search sends 12,900. Non-brand paid search sends 9,600. Direct sends 6,200. Brand paid search sends 3,100. Email sends 2,400. A partner referral sends 1,150.

Ranked that way, the story writes itself: paid social is the growth engine, the partner referral is a rounding error, and email is a legacy channel someone should probably sunset.

Now rank the same month by conversions. Organic produces 168. Email produces 139. Brand paid search produces 121. Direct produces 96. The partner referral produces 87. Non-brand paid search produces 41. Paid social produces 22.

The largest source of traffic is the smallest source of outcomes, by a factor of seven against a channel sending one-sixteenth the volume.

It is not that paid social is a bad channel in general. It is that traffic acquired against a broad interest signal converts at a rate one to two orders of magnitude below traffic acquired against an intent signal, and volume-based ranking has no way to express that.

Email and partner referrals sit at the other end: small audiences, pre-qualified, arriving with context. They will never top a sessions table and they routinely top a conversions table.

The uncomfortable case is brand paid search, which converts well and is partly harvesting demand that would have arrived anyway. Outcome ranking does not resolve that on its own — it tells you the channel produces conversions, not that it produced incremental ones. That question needs a holdout test, and it is worth knowing which of your top channels have that asterisk on them.

The obvious correction — rank by conversion rate instead of volume — swaps one distortion for another. Rate-ranked lists are dominated by tiny, high-intent sources: the partner referral at 7.6 percent looks like the best channel in the business until you notice it produced 87 conversions and cannot be scaled past about 1,200 sessions a month.

Volume answers "how much". Rate answers "how efficient". Neither answers "where did the outcomes come from", which is the ranking that maps to a budget decision. That one needs conversions in absolute terms, with rate visible alongside for context.

The disagreement between the two lists is itself the finding. When a channel sits high on sessions and low on conversions, one of a small number of things is true: the targeting is acquiring the wrong audience, the landing experience is failing a specific audience, or the conversion is not being measured properly for that path.

All three are actionable. None of them are visible on a sessions table, which is why the sessions table has quietly misdirected marketing budgets for twenty years.