Incrementality - Next Advertising Era
What does incrementality mean in marketing? Why does it become even more important in a world where measurement is more uncertain?
Incrementality in marketing is a concept used to measure the actual impact of advertising campaigns. Essentially, it refers to the difference between what would have happened without a specific marketing campaign and the results achieved with it.
The principal method for measuring incrementality is the randomized control trial, borrowed from the pharmaceutical world to verify the actual effectiveness of a drug.
Let's assume we want to test the effectiveness of a new fertilizer on different types of plants: garden plants, houseplants, greenhouse plants, etc. Our target includes all these varieties so that the results are applicable to a broad spectrum of situations. After defining the target, we choose a percentage of plants that will continue to receive the traditional fertilizer (our control or holdout group) and another percentage that will receive the new fertilizer (the test group). It's crucial that the assignment of plants to the two groups is completely random. This ensures that every other variable (type of plant, growing conditions, etc.) is evenly distributed between the control group and the test group, thereby avoiding unwanted influences on the results. We measure various aspects such as growth, health, and productivity of the plants in both groups for a certain period. After the testing period, we compare the results. If the new fertilizer shows a significant improvement in terms of growth or health of the plants, and this difference is statistically significant, we can conclude that the new fertilizer is more effective in an "incremental" way compared to the baseline treatment for those metrics.
In the world of marketing, we must add an additional element of control. We cannot distribute our light bulbs (our advertising message) to just anyone, but we must select a group of users to randomize homogeneously so that the test is not affected by biases, for example, that the two groups have very different purchasing capacities. There are many more variables at play that would make the explanation of incrementality very technical and are beyond the scope of this article, but know that the identification of the two segments and their realization in the real world is much more complicated.
If it's not possible to apply the randomized control trial methodology, other solutions can be used that try to simulate the two groups:
Cohort Matching:
Description: Select similar individuals or cohorts and compare those exposed to the advertising campaign with those who are not.
Advantages: Useful when it's not possible to take a single group and apply causal division.
Limitations: Requires careful analysis to ensure the groups are truly comparable.
Time Series Modeling:
Description: Analysis of historical data to predict what would have happened without the advertising intervention and compare it with the actual results.
Advantages: Effective for large-scale campaigns or over long periods of time.
Limitations: Can be influenced by external factors (such as seasonal or economic events).
Geo-Controlled Market Tests:
Description: Select different geographical areas, some of which receive the campaign while others do not.
Advantages: Useful for testing the effectiveness of specific campaigns in different markets.
Limitations: Local factors can influence the results.
Each of these methodologies can provide valuable insights into the effectiveness of advertising campaigns. The choice of the most suitable methodology depends on the specific situation and the resources available.
What is Brand Lift?
Brand Lift measures the impact of digital marketing campaigns on how consumers perceive and behave towards a brand. This is especially relevant for online advertising and social media. Here are the details:
Measuring Impact: Brand Lift evaluates the influence of specific advertising campaigns on brand perception aspects like awareness, attention, and purchase intention.
Key Metrics: This includes metrics such as increased brand awareness, enhanced brand image, higher purchase intentions, and other psychographic and behavioral factors.
Testing Methods: Brand Lift tests are typically conducted using surveys before and after an ad campaign, or by comparing control groups with those exposed to the campaign, to gauge differences.
Significance in Digital Marketing: In today’s world, where online advertising is everywhere, understanding a campaign's impact on brand perception is key. This insight helps companies fine-tune their advertising strategies and strengthen their brand.
Data Usage: Insights from Brand Lift tests enable companies to better understand their audience, refine their advertising messages, and allocate their advertising budget more effectively. Overall, Brand Lift is a crucial success metric for marketing campaigns, offering valuable insights into how advertising affects public perception and behavior towards a brand.
Brand Lift tools are available on these platforms, incorporating randomized control trials in their analyses:
Google Ads offers Brand Lift for in-stream and bumper ads bought through auctions. It's not available for outstream and in-feed video ads. Brand Lift analysis isn't accessible for all accounts; you need to contact your Google Account representative.
Meta Ads doesn't have the same creative limitations as Google Ads, but you still need to reach out to your Meta Account representative.
TikTok Ads allows Brand Lift for all types of creatives, but it must be set up by a TikTok Account representative.
Amazon Ads makes Brand Lift tests available for all campaign types via Amazon DSP. These are active in self-service and managed modes in the US, UK, and Canada, subject to certain spending thresholds that vary by marketplace."
Conversion Lift
Conversion Lift is a digital marketing term that refers to measuring the direct impact of an ad campaign on conversions or sales. This concept is essential for gauging the true effectiveness of online advertising efforts. Let's dive into the details:
Measuring Conversion Impact: Conversion Lift measures the increase in conversions (like sales, sign-ups, app downloads, etc.) that a specific ad campaign drives, compared to what would have happened without it.
Incrementality Tests: To assess Conversion Lift, incrementality tests are often used. These compare the behavior of a group exposed to the ad against a control group that hasn't seen it.
Importance for ROI: Understanding Conversion Lift is crucial for evaluating the return on investment (ROI) of ad campaigns. It's all about figuring out if the ad spend is really paying off in terms of increased conversions.
Data Analysis: Conversion Lift relies on data from conversion tracking tools, such as tracking pixels or conversion tags. These tools track user actions following ad exposure.
Making Data-Driven Decisions: Insights from Conversion Lift testing help businesses make smarter advertising investments. It's about fine-tuning campaigns to get the most conversions and cut down on wasted spend. In a nutshell, Conversion Lift is a vital metric for assessing the impact of online ad campaigns, focusing on the tangible increase in conversions directly linked to the ads.
In Google Ads, like with Brand Lift, Conversion Lift is available only for video campaigns and requires a Google Account representative to set it up.
In Meta Ads, Conversion Lift tools have been around for a while and are now available in self-service mode again, as long as certain conversion thresholds have been met in the last 30 days of campaign/account activity.
Other Tools for Measuring Incrementality
There are open-source tools available that allow for measuring the incrementality of any type of campaign, provided you have control over the campaign management.
GeoLift is an open-source solution developed by Meta for calculating geographic incrementality. It's designed for conducting robust geographic experiments, considered one of the best methods for measuring the effectiveness of advertising through incrementality, especially in a constantly evolving advertising ecosystem. The tool leverages the latest developments in Synthetic Control Methods, enabling decisions based on incrementality and measuring the true value of marketing campaigns. GeoLift offers a comprehensive solution for geographic experimentation that includes data ingestion, potential analysis, market selection, and inferential analysis, all in an easy-to-use R package. The open-source nature of the tool ensures transparency and reproducibility. GeoLift operates exclusively on aggregated market-level data, without using personal information.
"CausalImpact" is an R package developed by Google for causal analysis using Bayesian time series models. This tool helps estimate the effect of a specific intervention, like an advertising campaign, on a time series, for example, the increase in a website's sales. It uses control data, such as the trend in total sales, to build a model that predicts what would have happened without the advertising campaigns. This package is powerful and relatively easy to use, but it requires strong and accurate assumptions about the control data and its relationship with the observed data.
Why introduce incrementally into our planning?
Incremental analysis emerges as a valuable resource in the post-cookie era, where direct measurement becomes more challenging. This analysis not only facilitates the verification of underlying hypotheses in Marketing Mix Models (MMM) and attribution models but also provides a clearer view of the actual contribution of marketing activities. With these insights, teams can optimize budget allocation, refining strategies in real time. Moreover, incremental analysis supports adapting models to changes in consumer behavior and the dynamic digital ecosystem. This interdisciplinary process, combining first-party data and incremental learnings, underscores the importance of a collaborative and integrated approach. The end of third-party cookies necessitates a more intuitive measurement, inherently focused on the added value of each customer touchpoint, thereby ensuring more informed and data-driven marketing decisions.
Previous Episode of: Next Advertising Era
Reading Suggestion
discusses the evolving concept of Customer Data Platforms (CDPs), contrasting them with Customer Data Infrastructure (CDI) and examining the transition from traditional, bundled CDPs to a more composable, modular approach. It highlights the importance of syncing customer data across various platforms and the challenges in seamlessly integrating multiple tools within a modern data stack. The article also touches on the evolution of different types of CDPs from their original forms, like email tools and event pipelinesYour primary tag needs to be Switzerland by
explores the importance of primary event tracking in digital environments, emphasizing its role in data governance, customer privacy, and performance. It critiques client-side tag management's weaknesses, like governance issues and performance impacts, and argues for a singular, primary tag that observes and governs digital behavior data. The author advocates for a neutral, versatile primary tag, exemplified by Snowplow, to avoid vendor lock-in and ensure data integrity and ownership. This approach is presented as crucial for effective, compliant, and efficient digital data management