Sometimes, the data fetched from an API (through an Adverity datastream) does not contain the data you expect. For example:
- the aggregated metrics are not as you expect
- specific dimension records might missing
The reasons for this could be:
- Incorrect datastream configuration
- Time zone discrepancies - particularly the case with e-commerce platforms.
Incorrect Datastream configuration is the most common reason for data discrepancies. This article aims to provide a how-to guide to check the validity of Datastream configuration. This guide also provides instructions on how to report any data discrepancies if you suspect the Adverity platform is not working properly.
Data discrepancy in Destination
If a data discrepancy occurs in the Destination or Dashboard, first, check the data source to see if the same data appears there. Data discrepancy in a Destination can also occur if the following are not properly configured:
- Overwrite options
- Other Destination settings.
If the data discrepancy occurs in the raw data extract, move to the next step.
How to investigate Data discrepancy in a raw data extract
Most common reason of data discrepancy issues is the fact that report from the platform user interface (UI) is not set up in the same way as the Adverity datastream. The differences are usually in dimension or time breakdowns, dimension groupings, date ranges, account selections or some additional filtering.
The easiest way to check if the data provided via Adverity datastream is correct is to generate a report in the data source UI and compare it with the data extract in Adverity. Your original data in the datasource and the Adverity data extract must have the following:
- The same date ranges.
- The same date time breakdown (e.g. daily, weekly, yearly, lifetime).
- The same report type.
- The exact dimensions and breakdowns. Good practice is to have the same metrics too.
- Identical accounts/campaigns/advertisers selected in both Adverity and the data source.
- Identical time zones.
- The same attribution windows (if applicable).
Some of these parameters mentioned above may not be configurable, in which case, ignore that parameter.
What to do if the data discrepancy continues
If the data discrepancy persists, forward the issue to Adverity Support Team. When doing so, please attach all details and steps from your data discrepancy investigation. Such details can include:
- A link for datastream or extract where you expect the correct data.
- A generated report from the data source user interface that is configured in same way as your Datastream. If it is possible, provide such report with column names described in English language, otherwise, may require a translated version.
- If the report can not be provided, a screenshot from data source user interface will suffice. Ensure this screenshot contains:
- The configuration of report type.
- A list of columns.
- The report date range.
- Any other important properties are listed in the Investigation section above.