Incrementality: An Explainer

HomeInsightsBlogs | Last Updated July 5, 2023 - by bradley kossmann under data science & analytics

Published onMay 27, 2022

Incrementality is an increasingly hot topic in the marketing and digital analytics space. At Softcrylic, we wholeheartedly embrace this renewed interest in the relatively old concept of incrementality – we have been recommending incrementality measurement to our clients for years. It is our position that marketing should be evaluated by its incremental impact whenever possible. Incremental KPIs (e.g. incremental revenue, incremental ROAS, incremental conversion) reflect business value more directly than their non-incremental counterparts.

We are currently producing a content series that addresses all things incrementality: what incrementality is, why/when/where to use incrementality and how to approach the measurement of incremental KPIs. This explainer serves as a quick primer for how we define incrementality to set the stage for deeper dives.

Definition

An incremental unit cannot occur without an intervention (organically). Incrementality describes the extent to which a given entity drives incremental units.

Example

At Softcrylic, one of our favorite services is what we call Audience Modeling (learn more here). Assume that we want to run a paid search campaign on the keyword “softcrylic audience modeling.” What would appear if one were to Google “softcrylic audience modeling”? Our advertisement would appear, immediately followed by a series of organic links to the Softcrylic website. If the searcher were to then click through to the Softcrylic website through the ad and convert, most attribution schemes would give credit to the advertisement. But was that conversion incremental?

One way to conceptualize incrementality is to ask what would have happened without the intervention? In this case, what would have happened had Softcrylic’s advertisement not appeared at the top of the page? The searcher clearly had the intent to find Softcrylic’s audience modeling services and would have easily been able to click through an organic link. Looking at our definition above, this conversion was clearly not incremental because it would have happened organically without the intervention. If our audience modeling campaign in general behaved in this way, we would describe it as having low incrementality.

Approaches

There are two categories of incrementality analysis:

  • Calculation – there are many ways to calculate incrementality during or after a marketing campaign. While not as accurate as live media testing, incrementality calculations tend to require less effort and can be performed on historical data without needing a test design. Some algorithmic attribution approaches also attempt to simulate incrementality through what are known as removal effects, although this type of multi-touch attribution is somewhat limited in its practicality.
  • Testing – the most accurate approach to measuring incrementality is to run a live media test. Many types of testing that are already familiar to marketers, including A/B and holdout testing, are very similar to incrementality testing in execution. There are three drawbacks to incrementality tests relative to calculating incrementality from historical data:
    • There is potential to lose revenue during the test.
    • Tests can take a very long time to produce insights.
    • Test design and execution can be time-consuming and difficult.

    Regardless of the drawbacks, we recommend a rigorous incrementality testing regime. There is no replacement for knowing the value of your campaigns.

Benefit

Put simply, the benefit of adopting incremental KPIs is that incrementality provides transparency in marketing measurement. If a campaign is driving very little incremental value, it is worth considering whether some optimizations or budget re-allocations are in order.

If you are interested in learning more about incrementality and how to evolve your organization in the direction of incremental measurement, please don’t hesitate to reach out through the form below. Again, we are very pro-incrementality at Softcrylic and would love to have a discussion with you on the topic. If you want to learn more, but aren’t ready to reach out just yet, that’s ok too! You can follow this series through our newsletter (sign-up here) or just keep an eye on our insights blog, where you are guaranteed to find some other interesting topics to dig into. See you there!

Schedule 30 Minutes with our Consultants

    Bradley Kossmann

    Brad leads Softcrylic's Data Science & Analytics practice, a team whose capabilities span every stage of the analytical project lifecycle. He loves helping clients build data-driven cultures with engaging, powerful insights and analysis. Brad spends a little too much time dwelling on the problem of incrementality in digital marketing and the problem of AI alignment more generally.

    Contact Us

    We're not around right now. But you can send us an email and we'll get back to you, asap.

    Not readable? Change text. captcha txt

    Start typing and press Enter to search