Collecting and visualizing non-aggregatable metrics

This guide explains how to collect and visualize non-aggregatable metrics.

Introduction

For some metrics, it does not make sense to add up the values. Examples of non-aggregatable metrics are reach, frequency, and bounce rate. When non-aggregatable metrics are not collected or visualized correctly, discrepancies can occur in your data.

For example, you collect data about daily reach which shows the number of unique people who come in contact with your ads each day. In this case, it does not make sense to add up the values for different days to obtain the weekly reach values. The reason is that some people can come in contact with your ads on multiple days during the week. Each day these people come in contact with your ads, they are counted in the daily reach. Simply adding up the daily reach values means these people are counted multiple times. For this reason, this method does not provide you with weekly reach because weekly reach shows how many unique people your ads reach each week.

This video guide explains what non-aggregatable metrics are, how to fetch these metrics from your chosen data source, and how to visualize them.

Working with non-aggregatable metrics

To avoid discrepancies in your data as a result of non-aggregatable metrics, follow these steps:

  1. Visualize non-aggregatable metrics with the same level of granularity you used when you collected these metrics. The level of granularity can mean, among others:

    • The datetime breakdown (for example, daily, weekly or monthly).

    • The data filters (for example, the selected accounts, campaigns and advertisers).

  2. Create a separate datastream for each level of granularity you want to use for non-aggregatable metrics.

  3. In the separate datastreams, map each non-aggregatable metric to a metric dedicated to the level of granularity used in the datastream. For example, map daily reach to daily_reach and weekly reach to weekly_reach. Do not use reach for both the daily and the weekly breakdown. For more information on Data Mapping, see Applying custom Data Mapping.

  4. When you visualize non-aggregatable metrics, select all available dimensions. If you create a widget in Explore & Present and do not select all available dimensions, Adverity displays the number of available values for the combination of dimensions you selected. For example, if Adverity finds two values for a certain combination of dimensions, it displays found 2 values in the table instead of aggregating the values of a non-aggregatable metric.