Why Placement-Level Reporting Matters More Than Platform Averages
A campaign averaging a $12 CPM and a 0.8% click-through rate across 25 placements looks like it is working. But that average hides everything that matters. Some of those placements may be converting at twice the rate of others, while a handful produce clicks that never lead to a sale. The aggregate number is a blend of strong and weak performance, and it tells the advertiser nothing about where to spend more and where to cut.
This is a problem on any digital advertising platform that reports campaign-level totals without breaking the data down by individual placement. The advertiser sees one number. That number looks acceptable. The budget keeps being allocated evenly across all placements, including the ones dragging down the average. The loss is invisible because the top performers are subsidizing it.
How Averages Mislead
Simpson's paradox describes how aggregate data can tell the opposite story from the data underneath it. In a well-known example, a 1975 paper that examined UC Berkeley's graduate admissions found that overall numbers appeared to show bias against women. When the data was broken out by department, the pattern reversed. Women were admitted at equal or higher rates in most individual departments. The aggregate was misleading because women applied in larger numbers to more competitive departments with lower acceptance rates.
The same dynamic applies to ad campaigns. A campaign running across placements with very different traffic profiles will produce an average that misrepresents what is happening at the individual level. A placement with a low CPM and a high click rate pulls the average in one direction. A placement with expensive impressions and almost no engagement pulls it the other way. The average sits somewhere in the middle, where neither placement actually lives.
What Placement-Level Data Reveals
Viewability is one area where there are large placement-level differences. The IAS 21st Edition Media Quality Report, published in 2026, reported global display viewability at 67.9% and video viewability at 79.7%. But these are averages across environments and geographies. The report showed that video viewability by country ranged from 63.21% in Japan to 84.37% in Poland. That is a 21-percentage-point spread within the same format.
At the placement level, the spread is even wider because individual ad spots on different pages and positions have their own viewability profiles based on where the ad sits on the page, how fast the page loads, and how users scroll.
An advertiser paying a flat CPM across all placements is paying the same price for an impression that loads above the fold and stays on screen for five seconds as for one that renders below three scrolls of content and is never seen. Without placement-level viewability data, there is no way to separate the two.
The Cost of Delayed Granularity
Reporting speed compounds the problem. A campaign spending $3,000 a day across 20 placements allocates $150 per placement on average. If three of those placements have viewability rates under 40% and conversion rates near zero, that is roughly $450 per day going to impressions that are not producing results. Over a week, the waste reaches $3,150 before anyone catches it, assuming the data is even available at the placement level to begin with.
TrafficJunky provides placement-level performance data within its reporting dashboard, so advertisers buying across multiple spots on different sites can compare CPM, click-through rate, and conversion rate for each individual placement. That level of detail is what makes it possible to shift budget toward what is working without waiting for a weekly summary.
What to Look for in Reporting
The minimum requirement is reporting that breaks out impressions, clicks, cost, and conversions by placement, device, and geography independently. If the platform only reports at the campaign level, every optimization decision is based on an average that conceals the actual distribution of performance.
Time granularity matters too. Daily reporting is the baseline. Hourly is better for campaigns at higher spend levels. The shorter the reporting interval, the faster an advertiser can identify a placement that stopped converting or a time slot where costs spiked without a corresponding increase in results.
Data export is the other piece. Advertisers running campaigns across multiple platforms need to pull placement-level numbers into a single view, whether that is a spreadsheet, a BI tool, or a custom dashboard. A platform that locks data behind a proprietary interface with no export option slows down every comparison and every reallocation decision. The numbers have to be accessible to be useful.