As an industry standard, directly comparing delivery metrics between Sharethrough for Publishers (SFP) and 3rd party ad servers should be avoided. There are many things that may cause discrepancies between SFP and 3rd party ad servers, but there is no need for escalation if the difference is less than 15%.
Discrepancies of 15% or less are within the industry standards, and are commonly explained by many things, including:
- Tracking Differences - Ad servers define each event slightly differently. Detailed definitions of each Sharethrough Metric can be found here: SFP Key Terms
- Impression Requests - Impression requests from anything but the following browsers and devices are ineligible for Sharethrough delivery and should be disregarded for discrepancy analysis:
- Browsers: Chrome, Firefox, Safari, Internet Explorer (version 11)
- Mobile Operating Systems: iOS, Android
- Additional Demand Sources – By rotating Sharethrough tags through additional mediation, you increase the latency between page load and ad load. This can cause an increase in discrepancies.
If you are seeing a discrepancy larger than the industry standard of 15%, please email email@example.com. Including the following items will allow for an expedited resolution:
- Ad Server Delivery Report – A report from the 3rd party ad server, broken out by day. Please adhere to the below criteria for the report:
- Date Range: Ensure that SFP and the 3rd party ad server are using the same date range.
- Time Zone: If possible, pull your data in UTC.
- Geography: For discrepancies with Sharethrough Exchange (STX) demand, please limit the ad server report to the United States, Canada, Australia, United Kingdom, France, Spain, Italy, Germany, Belgium, and Finland.
- Ad Server Information – Please provide either: a) access to your ad server for sysops@sharethrough, or b) screenshots of the ad server setup.
- Test Page – A test page with the placement, campaign, or creative in question receiving 100% fill. For best practices on creating test pages, see the following article: Best Practices for Creating Test Pages