# Incrementality Testing

## What is incrementality testing?

It is a way to measure advertising by comparing exposed and unexposed audiences to determine the actual impact of a media channel and tactic. The idea is to understand who converted because of advertising vs. those who would have converted anyway, regardless of advertising.

## Why is it important?

Measuring incrementality allows you to more accurately identify the impact of any given marketing initiative, providing a better understanding of what advertising tactics are the most effective, how to reach desired audiences more efficiently, and what is worth investing in.

## Is incrementality a form of attribution?

Yes! Incrementality is a method of understanding a partner’s unique contribution to conversions and revenue, similar to how multi-touch attribution measures touchpoints on a customer journey.

In other words, we can tie media to a conversion and use incrementality to enhance that approach. Don’t rely on one methodology; instead, consider using multiple approaches, as each type serves a different purpose.

## Types of incrementality

### Total conversion contribution rate

• The proportion of all conversions (not just the ones touched by the partner) that are generated by working with a partner.
• If the incremental lift on total conversions is 10%, that means 10% of all conversions are as a result of working with the partner and would not have happened without the partner.

### Messaging-based conversion contribution rate (holdout group measurement only)

• The proportion of messaging-based conversions that were incremental to the business overall.
• If the incremental lift on messaging-based conversions is 10% , that means 10% of the conversions touched by the partner’s messaging would not have happened without the partner.

### Incremental return on ad spend (iROAS)

• The measure of how much additional revenue is earned per dollar spent on marketing efforts. To calculate this metric, we use the following formula:
• Incremental Return on Ad Spend (iROAS) = Incremental revenue / ad spend*
• Ideally, this calculation results in a number above 1.0. If we see a 2.0 iROAS, this means for every \$1 spent with a partner, they earned \$2 in return.

## Success factors for audiences

Audience splits are typically created through randomization or match market (geography).

The important part in creating these audiences is making sure that they are balanced and that the difference can be attributed to solely the media.

## What are the types of tests Quantcast can run?

Once you’ve determined your audience methodology, the testing can happen through one of the following ways:

• Public Service Announcements (PSA) Incrementality: Based on balanced groups, branded ads are served to a portion of the audience, and PSA ads are served to another portion of the audience. The difference in conversions can then be attributed to the media.
• Holdout Incrementality: Based on balanced groups, a portion of the audience receives branded messaging, while the other receives no messaging. The difference in conversation can then be attributed to the media.