[Tableau] Missing Data with Incremental Refresh

In Tableau, there are two ways to extract data source updates: full refresh and incremental refresh. By default, full refresh is performed.

With full refresh, all rows from the data source are extracted, so Tableau reflects an exact copy of the data. However, for large datasets such as transaction data, full refresh places a heavy load on the source database. When using BigQuery, large amounts of data processed by queries can result in significant costs. In such cases, consider incremental refresh.

With incremental refresh, only new rows are extracted, excluding data that was already extracted last time. For example, using a date field as the key, only rows where the date is after the previous update are extracted. However, with this setting, data added retroactively to past periods will not be reflected in Tableau.

When using incremental refresh, you must carefully consider in advance whether any data might be missed in Tableau, and whether missing data would be acceptable.