Delivery of advertising creatives: Measuring and reporting data transfer and savings

SeenThis uses two different methods to measure and evaluate data transfer, in addition to the amount of data that's been reduced using our streaming technology compared to traditional technology.

A/B testing

A/B tests have been (and still are) the foundation for the continuous optimization of streaming technology when it comes to content delivery, while transferring even less data without compromising quality or load time. Since clients are keen to reduce as much unnecessary data waste as possible while maintaining high quality, A/B tests are set up on a per-request basis, as testing requires the use of traditional technology for campaigns.

The process for setting up an A/B test is as follows:
Two ads identical in appearance are created. One is set up to be delivered using traditional technology (which transfers more data than necessary and reduces performance), while the other ad is delivered using SeenThis streaming technology.

The ad delivered using traditional technology is placed in Group A, and the one delivered with SeenThis streaming technology is placed in group B. Data transfer tracking and measuring* is implemented for both groups, which records the amount of data transferred for each impression (excluding all cached versions that can/will occur for an end user targeted more than once by an impression).

The test is then set up for trafficking as an A/B test (programmatic or IO), rotating Group A and Group B. This means that the ads in each group are delivered under the exact same circumstances (placement, targeting, inventory, etc.). The only difference is the delivery technology used.

Once the test is completed, all figures are summed up and a direct comparison of the amount of data sent between both groups is made. We are then able to draw conclusions on the amount of data transfer saved through SeenThis technology.

"There are several ways to measure the amount of data being delivered from CDN data centers to end users. SeenThis uses the most accurate and reliable method to track and monitor this information, per the CDN's recommendation to track egress bandwidth using the resp.bytes_written variable."

Sales Engineering Manager, Fastly EMEA

Automatic evaluation

SeenThis has developed a methodology for automatically evaluating data reduction for both video and static ads based on campaign-specific data. For video ads, a regression model has been developed for predicting the amount of data reduced using our streaming technology. The model is based on a large sample set of historical campaign data and can accurately predict data waste with a mean absolute percentage error of 1.22% and a standard deviation of 0.88% for any given ad specification.

For static ads, SeenThis evaluates data reduction with a similar methodology based primarily on how static ads are delivered with traditional download technology. This high level of accuracy means that SeenThis is able to measure data savings without wasting impressions on traditional technology.


SeenThis provides different alternatives for reporting, depending on which method is used.
The CO2 reporting, the emissions dashboard, A/B testing and the Emissions calculator are all ways of helping our clients understand their saved emissions on digital ad campaigns.

Read more