This article discusses discrepancies between impressions reports from Triton Digital versus similar reports from third-party vendors, and what to do if there is a discrepancy between them.
Definition of ad discrepancy
Ad discrepancy is when there is a difference between delivery numbers from your ad server and those of a third-party.
According to the IAB, any discrepancy below 10% is acceptable. Our recommendations are:
Discrepancies between 1-5% should cause no concern. However, you should still double-check your third-party tag or tracker set up. (See below for details.)
For discrepancies over 5%, it’s worth digging deeper. Review the section on potential reasons for discrepancies below before submitting a ticket to support.
To quickly catch any potential discrepancy, we recommend you ask for a daily delivery report from the third-party (or your advertiser/agency) two days after your flight starts to deliver.
Potential reasons for discrepancies
Different methodologies used to measure an impression
Often the ad server and the third-party use different counting methodologies to measure ad impressions. Ask the third-party how they count an impression and compare it with how Triton does it.
Review: Triton impression counting methodology for in-stream and podcast:
In-stream (radio stations):
Pre-roll: short sessions are discarded, any ads which might have started in a discarded short session are not counted as an impression. Impression is considered delivered two seconds after the ad starts playing.
Mid-roll: Impression is considered delivered two seconds after the ad starts playing.
Podcast:
Pre-roll, mid-roll, post-roll: Ad is inserted at download. Impression is triggered by the download.
Triton methodology does not come into play for on-demand calls, as the publisher player or server defines when to count an impression.
IVT and IABv2 compliance
There could be a difference between the third-party and Triton Digital on how IVT (invalid traffic) is filtered. For podcast impression counting, Triton conforms to the IABv2 compliance.
Check with the third-party:
Are they IAB certified for podcast impression counting?
Do they filter IVT? If so, how?
Time difference
At Triton Digital, impressions delivery is based on station/podcast time zone, while reporting is in the UTC time zone.
Ask the 3rd-party what time zone they are reporting in.
Different geo definition
All ad servers use a partner to map IP address locations. If the third-party uses a different geo lookup partner than the one used by Triton, it can lead to significant differences.
Triton uses MaxMind for our geo look-up. Double check with the third-party to see which one they use.
Network connection and/or server reliability
A third-party ad server may fail briefly or encounter an issue that prevents it from logging an impression.
Difference in ad campaign settings
The third party might have set targeting rules which are different and maybe conflicting with your flight targeting rules, whether in time, listener, content or all three. For example: you did not set a geo targeting rule, but the third-party limited their targeting to US only. Triton will deliver impression everywhere, following the flight targeting rules, while the third-party will only record the impressions coming from the US.
The third-party might have applied frequency capping that is different from yours.
Ask your third-party if they have set targeting or frequency caps on their end and if so, what targeting/caps.
Multi-use of tags
Your advertiser might have sent the same tags to multiple different publishers by mistake. Hence the third party is recording impressions from all publishers without distinction, resulting in a much higher number of impressions on their end than what you are seeing.
Ask your buyer to double-check that the trackers/VAST tag they provided you is unique to you and was not sent to others as well.
Incorrect setup of third-party tag
To make sure you’ve implemented the trackers correctly, read this article: Working with Third-party Creative Trackers
Missing or unsupported macros
Although unsupported macros might not lead to non-delivery, it could affect what the third-party is tracking on their end.
Some macros, such as cachebuster/timestamp, must be recorded properly; if missing, they will significantly affect what the third-party is tracking versus what Triton is tracking. More information: Using Macros in TAP.
Conclusion
If, after checking all of the above, you still cannot explain the discrepancy, submit a ticket to customer support. Make sure to include your flight URL, and the third-party report (request this from your third-party prior to submitting your ticket).