Grade the traffic, not just count it.

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

The question people actually have is where the value came from. Those two lists routinely invert, and they invert most sharply on the channels carrying the largest budgets — because paid channels are optimised for volume against broad targeting, and volume is exactly what a sessions ranking rewards.

Anything a session can be attributed to, provided it carries enough sessions to be meaningful.

• Channels — paid social, paid search, organic, email, referral, direct.
• Campaigns, ad groups and individual creatives, where the tagging supports it.
• Referring domains, which is how partner and PR traffic gets a quality grade for the first time.
• Landing pages and entry points, which separates a traffic problem from a page problem.
• Device and geography, which is where a channel's average often hides two very different audiences.

Each gets an intent distribution rather than a single average. An average hides the shape, and the shape is usually the finding — a channel with a small tail of very high-intent sessions is a different proposition from one with a uniform middling score, even when the averages match.

In practice, grading traffic by intent surfaces the same three situations over and over.

• Volume without intent. A large channel with a flat, low distribution. It is buying attention from people who were never going to buy. This is the money-losing case and it is invisible on a sessions report.
• Intent without conversion. Good scores arriving and not converting. Almost always a page, offer or checkout failure rather than a traffic failure — and turning the channel off would be exactly the wrong response.
• Small and excellent. A low-volume source with a strong distribution, sitting at the bottom of every report you have ever run because it cannot compete on sessions. Usually a partner, a newsletter or a community. Usually underfunded.

The other use is longitudinal. Because intent is measured per session, it responds within days of a targeting change — long before enough conversions accumulate to show anything.

When you narrow an audience, change a creative, or rewrite a landing page, the intent distribution for that source starts shifting almost immediately. That gives you a feedback loop measured in days rather than in quarters, which is the difference between iterating on a campaign and guessing at one.

It also gives you an early warning in the other direction. A channel whose intent distribution is quietly degrading is a channel whose revenue will degrade later — often after a platform-side targeting change nobody told you about.