Adthena performs extensive overnight processing, validation and modeling before publishing the latest day’s data.
As part of this process, Adthena uses a complex machine learning model to estimate, nightly, both CTR and CPC for every search term, for every competitor, in every position they have been seen appearing that day.
Once this processing is complete, a series of automated quality checks are performed before data is made available in the platform. Our goal is to have data available by 9:00am local time each day.
On occasion, these checks may identify anomalies or incomplete processing that require additional validation or reprocessing before the data can be published. In some cases this can require manual intervention from our engineering team.
When this happens, we may delay publication while the issue is investigated and resolved. We take this approach because we believe it is better to delay data briefly than publish data that may be incomplete, as it would mean we would have to update again later that day.