IAB Standard Traffic Measurement
Quantcast Measure provides online publishers neutral third-party audience metrics based on direct measurement of their media consumption. We publish a variety of metrics as part of the program, from overnights showing basic event counts to demographic profiles to viewership rates by geography and organization.
We have identified a core set of traffic metrics for more formal treatment, including page views, session counts, video plays, and unique cookies. The industry, through the Internet Advertising Bureau (IAB), has published standards defining how to compute these metrics and what disclosures and procedural controls must be in place for the industry to rely on what we publish. Since 2008, we have been adapting our definitions and procedures to enable us to publish a set of metrics that adhere to the IAB’s standards.
About IAB Standard Traffic Report
The IAB Standard traffic report on our web site shows metrics for which we prepare additional checks and corrections to adhere to industry standards. We publish these monthly, covering the 30-day period leading up to the end of the most recent month. We calculate the end of the month based on midnight GMT. More about the methodology behind these metrics
Although we take particular care to detect and resolve data anomalies before releasing our IAB report results, it is possible that we may discover errors after data has already been published. Our policy is to restate prior results if we discover significant discrepancies in net impression, cookie or play counts and if the erroneous data is still viewable through our site. To be significant, the difference must be at least 10% and an absolute count of 1000 cookies, 10,000 impressions, or 10,000 plays. We will restate the results by publishing new results on our site, with a note explaining the data has been restated.
Below are the definitions of the IAB Standard metrics for a given property and publisher:
The number of pixel requests we received, based on a complete count of rows in our server logs.
The fraction of all events that appear to be duplicates–that is, followed within one second by another event from the same user visiting the same publisher. Common reasons include fast redirects (sending the user to a second page as soon as the first has loaded), multiple tagged iframes, or multiple media types being tracked. A Quantified ad network serving ads into multiple iframes on a single page will typically generate multiple pixel requests within a second, and these would be counted as multiple events. Similarly a publisher using separate tags for tracking pages and the ad placements on those pages will see a significant multi-event count.
The fraction of all events we estimate to be generated by non-human activity. In addition to the IAB’s bot and spider list, we use our own proprietary techniques to identify non-human activity.
The fraction of all events that appear to be triggered by an auto-refresh mechanism. We base our estimate on a frequency analysis of the publisher’s traffic, in which the regularity of auto-refreshed pages stands out from the background of human-requested pages.
The number of seconds between refreshes of the fastest auto-refreshing content discovered.
The number of full, human-requested web pages for which this publisher supplied some or all of the content. This is calculated from the raw HTML event count we observed, less correction factors for non-human traffic, multiple impressions, and auto-refresh.
The number of post-buffering Flash video play events observed.
The number of distinct visits of users to the property. A session is any period of activity (which could be a single page view or a multi-day marathon) separated by at least a half-hour of inactivity, per the IAB’s standard.
The number of distinct Quantcast cookies we received with pixel requests.