Education

Understanding the Metrics in Lotame Insights

Lotame Insights provides visibility into the behavioral attributes that make up campaign interactions, and how those attributes contribute to the performance of each step along a consumer’s path to conversion. Whether you have one simple path to purchase or you have dozens of varying length and complexity, no two conversion funnels are the same. Lotame Insights tracks online events and interactions and provides intelligence about the audience and behavioral composition at each step along the path. Track dozens of events occurring within display campaigns, email campaigns, checkout flows, for instance.

Armed with knowledge of how different behavioral attributes are aiding (or prohibiting) the action you seek, you can fine-tune your audience(s) to ensure you’re running the highest-performing and most efficient audience targeting possible.

When a preceding campaign event, such as an impression, is tracked in addition to a secondary campaign event, such as a click, Lotame Insights calculates the interaction rate at the campaign level, using all profiles that qualify for both campaign events, and it calculates the expected interaction rate at the audience- and behavior-level for every audience and behavior identified in profiles qualifying for your campaign event.

Estimated Click Rate vs. Campaign

Estimated Click Rate metrics reflect the performance you can expect from that audience or behavior (depending on which report view you are in) should it also qualify for the preceding event. The Estimated Click Rate vs Camapgin reports the expected increase (or decrease) in your interaction rate for that audience and behavior compared to the interaction rate for that campaign event.

 

Events such as impressions and clicks, and their associated interaction rates, are tracked and reflected in Lotame Insights as Estimated Impressions, Clicks, and Click Rate. These events, and others you may wish to track as part of your Insights Campaign, are estimated as Lotame Insights relies on sampling to return audience composition and performance reporting as quickly as possible.

These numbers are derived based on the total impressions (or clicks) we recorded over the last 30 days x the affinity (defined below) for the audience or behavior listed.  For example, if the campaign delivered 1000 impressions in the last 30 days and we have an audience with a 50% affinity, we display 500 impressions. Please note that the affinity numbers driving the derivation are based on sampling.

Unique profiles, and the data attributes appended to those profiles, are extrapolated to provide the composition and performance metrics. The events tracked as part of your Insights campaign are based on the number of unique profiles qualifying for each behavior or audience, and not the absolute number of campaign events.

Affinity

Affinity

Affinity is a measure of similarity, or overlap, between two pools of unique profiles. In Lotame Insights, the Affinity metric can be used to measure the similarity between unique users collected from campaign events and other preexisting first-party behaviors or audiences. You can even compare the affinity of campaign events to data points from third-party data providers.

To visualize, consider the diagram below. The pool of unique users that clicked on your campaign have a higher affinity to the “Enjoys Horror Movies” audience and a lower affinity to the “Attended 2+ Music Festivals” audience. In other words, more unique clickers qualify for the “Enjoys Horror Movies” audience than the “Attended 2+ Music Festivals” audience.

Music Festivals and Horror Films Venn Diagram

Affinity is represented as a percentage and is calculated by dividing the number of unique users qualifying for a single audience or behavior and profiles which have also performed a campaign event by the total number of unique profiles tracked campaign event and multiplying by 100.

Using the above example, the yellow circle represents 125 unique users who clicked on a campaign. Of those 125 clickers, 45 qualify for the “Enjoys Horror Films” audience (orange semi-circle, seen between the yellow and red circles) while only 10 of the 125 qualify for the “Attended 2+ Music Festivals” audience. Therefore, the “Attended 2+ Music Festivals” audience’s Affinity is 8% and the “Enjoys Horror Films” audience’s Affinity is 36%.

Index

The probability that you will reach a user with a given behavior if you targeted a specified audience relative to if you targeted the all members audience (run of network / RON). If the index for health vs. sports is 200, then the probability of finding a user with health behavior in a sports audience is 2x higher than if you targeted the all members audience.

This number tells us what audience to target efficiently and it reduces impression wastage.

Index is a metric used throughout the DMP. In Lotame Insights, Index is simply a comparison of two Affinities: the Affinity of an audience (or behavior) to a campaign event tracked in Lotame Insights, such as clickers or converters, and the Affinity of that same audience (or behavior) to all your first-party profiles. Understanding the prevalence of an audience in a pool of unique profiles relative to the prevalence of that audience across all first-party profiles can be helpful when determining the likelihood of reaching that audience, should you decide to target it as part of your campaign.

To visualize the Index metric, consider the diagrams below. There are 125 unique converters tracked as part of your Insights Campaign (red circle). The “Household Income $225K+” audience (yellow circle) contains 80 thousand unique profiles. Lotame has identified 1 million unique first-party profiles across all properties and campaigns.

Unique Converters and First-Party Profiles Venn Diagram

Of the 125 unique converters, 15 also qualify for the “Household Income $225K+” audience (orange semi-circle, seen between the yellow and red circles). The Affinity of unique converters to the “Household Income $225+” audience is 12%.

Of the 1 million unique first-party profiles, 50 thousand also qualify for the “Household Income $225+” audience (green semi-circle, seen between the yellow and blue circles). The Affinity of all first-party profiles to the “Household Income $225+” audience is 5%.

Index is calculated by dividing the Affinity of the campaign event to an individual behavior or audience, by affinity of the campaign event to all first-party profiles and multiplying by 100. In this example, the Index is 240%.

An Index over 100% indicates the audience or behavior is more prevalent in the campaign event than in your pool of first-party profiles. An index under 100% indicates the audience or behavior is more prevalent in your pool of first-party profiles than in the campaign event.

By leveraging metrics like Index, Affinity, and audience- and behavioral-level interaction rates and understanding how they relate to the overall campaign performance, you can begin to understand how various audiences or data points are affecting your campaign performance, and how they could contribute to performance improvement if appended to campaign targeting.