Choosing the Right Time Granularity for Your Data

Some data connectors provide multiple time granularity levels for their tables: _daily, _weekly, _monthly, and _lifetime. Choosing the appropriate granularity depends on the type of metrics you want to analyze.

Aggregatable vs Non-Aggregatable Metrics

The fundamental distinction lies in the nature of the metrics:

Aggregatable metrics: These metrics can be summed without loss of accuracy. For example, impressions: if you have 100 impressions on Monday and 150 on Tuesday, you can confidently state that there were 250 impressions over those two days.

Non-aggregatable metrics: These metrics represent unique interactions and cannot be simply summed. A classic example is reach: if 100 people saw your content on Monday and 150 on Tuesday, this doesn't mean 250 unique people saw it over both days, as some may have seen it on both days.

Usage Guidelines

Use _daily tables for:

• Aggregatable metrics like impressions, clicks, or spend
• Detailed day-by-day analysis
• Calculations where you need to sum daily values

Use _weekly, _monthly, or _lifetime tables for:

• Uniqueness metrics like reach, unique users, or unique interactions
• Analysis where you need to capture behaviors over a specific period

Practical Example

To calculate the total reach of a campaign over a month, you cannot sum the reach values from the _daily table. You must query the _monthly table directly, which correctly calculates the number of unique users who saw your content over the entire period.