Yes. You can configure the look back window (Attribution Offset) from the UI and Daton will repeat the data after the selected number of days.
Look back Until: Repeat data until selected number of days. You can select up to 31 days to replicate changes up to 30 days. Look back After: Repeat data after selected number of days. Multiple days can be selected up to 31 days.
Look back Until: When you consume Amazon reports every day for your business insights. Look back After: When you consume your reports on a weekly or monthly basis. Note that the metrics may not be accurate for previous 7 days as the minimum look back date is 8 days in Daton.
It depends on your use case. If you use 30day attribution for your reporting, then you must choose 31day offset on Daton. Amazon Ad console will use 14day attribution for most of their reports. If you want to match with Amazon reports, then you must use 15day offset on Daton.
Since we replicate the data after a few days (based on selection from UI), we append the new data to the tables by default. You can also configure the tables or integrations to ‘upsert’ instead of ‘append’ to handle the duplicates.
You can use _daton_batch_runtime to identify most recent record for a primary key of the table. Use the deduplication guide at the bottom of this page.
Amazon Ads APIs share previous 60 days of reports. Daton can lookback and replicate previous 60 days of data from the date of integration.
Yes. Daton will drop and create the table again for reloads. Historical data replicated so far would be lost and previous 60 days data would be replicated again.
No. Amazon will not share present day’s reports through APIs, until the day ends. Daton can replicate only previous days data after 2 AM of seller time.
Daton replicates previous day’s data in the first job after 2 AM of seller time. All jobs after that will be skipped and marked success with zero records processed. This is to reduce the volume of data and duplicates on destinations. If the first job fails, then Daton will repeat the job based on frequency selected.
ReportDate is the date for which the data belongs to. For example, reportDate=’2021-04-01’ will have all metrics such as clicks, spend, attributedSales14d generated for 1st April.
There will be <1% variance after 15 days. There can be 5-10% variance for previous day (applicable for previous week if you select look back after option).
There could be various reasons for this – • Deduplication mistakes: You could have made mistakes while removing duplicates from Daton populated tables. Use the guide at the bottom of this page for deduplicating the data. • Are you using right attribution metric while comparing with Amazon reports? • Errors in integration: Check if there are errors in Daton jobs for the subject dates and report to us. • Check if all campaigns present in Amazon reports are also present in Daton tables. Are there any name changes for campaignIds? • Check if you have used campaignName for deduplication. You must always use campaignId for primary key as campaignNames can be changed.
Yes. You can change them on edit integration workflow.
NUMBER() OVER (PARTITION BY campaignId, reportDate order by _daton_batch_runtime desc) row_number from `daton-210514.AmazonDataset.AmazonSP
4xxxxxxxxxxxxxxxxxxx1_SponsoredProducts_CampaignsReports` ) where row_number=1