- Enterprise40
- Oct 14
- 7 min read
Updated: Oct 21
Table of Contents:
Accurately measuring and attributing sales or other key performance indicators to individual marketing channels is crucial for brands that would like to make informed decisions on marketing budget allocation and spend to pursue growth. There are several methodologies for measuring the impact of individual marketing tactics, including A/B testing, geographic testing, time series analyses, and marketing mix modeling (MMM). Although each of these methodologies for incremental measurement have their advantages, most of them also have limitations. Unlike these other methodologies for incremental measurement, marketing mix modeling provides brands with a holistic view of their marketing efforts in terms of how each channel drives sales directly. By focusing on sales, revenue, and other core business drivers rather than secondary metrics like impressions or website visits, marketing mix modeling gives you accurate insights that you can trust to make informed decisions on marketing campaign spend.
What is Incremental Measurement in Marketing?
Incremental measurement refers to the process of quantifying the impact that specific marketing activities have on key performance indicators such as sales, revenue, or business growth. Such measurement allows you to make decisions on which channels justify additional focus or spend, and which channels are less impactful than presumed.
Incremental measurement is not simply tracking the number of sales that occur within a retailer or website, but instead quantifying the specific driver that led to the sale. It is definitely not last-click attribution, either.
One of the most effective ways to measure the incrementality of your brand’s marketing activities is through a process called marketing mix modeling, which uses years of historical data and intricate mathematical models to accurately attribute individual marketing tactics to sales or other KPIs. Other methods used for incremental measurement in marketing, such as A/B testing or brand lift studies, are more simplistic and offer a narrow view that may miss the bigger picture and lead to inaccurate, misinformed decisions. Marketing mix modeling allows brands to measure many factors at the same time, resulting in a more complete view into the performance of their campaigns.
Why is Incremental Measurement Valuable for Brands?
Incremental measurement is incredibly valuable for brands because it allows them to see the true impact of each of their marketing tactics. By accurately attributing sales or other key performance indicators to specific marketing activities, brands can make data-driven decisions on budget optimizations that will improve their incremental return on investment (IROI).
A/B testing, an incremental measurement method, uses two versions of a marketing asset in order to observe which performs better in terms of a predetermined metric like sales or views. While A/B testing has a relatively quick turnaround time, it does not necessarily provide brands with a complete picture into why or how their marketing activities are resulting in core business drivers.
Marketing mix modeling, on the other hand, uses years of historical data and statistical modeling to measure which sales can be attributed to which marketing activities with a high degree of accuracy, resulting in granular incremental measurement that can be used to make budgeting decisions with confidence.
Incremental Measurement Methodologies: Benefits and Limitations
A/B Testing
A/B testing is essentially an experiment that uses a control group and a test group to measure the impact, sometimes referred to as the “lift”, that a tweak to a marketing tactic makes on sales, conversions, or other target performance indicators. This experiment lets brands see the number of sales or other KPIs that the change to the test group made, therefore providing them with insights into whether or not the change to their marketing strategy increased revenue, engagement, or other business drivers.
Circana’s Sales Lift, a form of A/B testing, is highly accurate because we use sales data to measure the impact of A/B testing for incremental measurement, whereas other services use secondary KPIs like views or engagement, resulting in less accurate insights. A/B testing and Circana’s Sales Lift are shorter-term solutions than other incremental measurement methodologies like marketing mix modeling. Sales Lift is great for boosting sales in the short-term, while marketing mix modeling is used for constant growth over a longer period of time. With that being said, these two incremental measurement methodologies work well when used in unison, as marketing mix modeling provides a holistic view into which of your marketing tactics have or have not been working well, and A/B testing can be used to iterate activity and prove those findings. Circana’s granular sales data results in incredibly accurate insights from both A/B testing and marketing mix modeling due to our industry-leading granular sales data.
Marketing Mix Modeling
Marketing mix modeling is an analytical tool that provides brands with a holistic view of how each of their marketing tactics is performing in terms of the amount of sales or revenue that each tactic is driving. Marketing mix modeling uses a time series regression based mathematical model to analyze the impact that different marketing channels, along with other external factors, have on sales. Marketing mix models are built to be as accurate as possible using pooled data (rather than data requiring customer consent), identifying the correct incrementality for each marketing channel. While consumers are undoubtedly influenced by more than one marketing channel, marketing mix modeling tests and isolates campaigns to find out which tactics result in more sales and which channels should be run together.
One of the best features of marketing mix modeling, which sets it apart from other incremental measurement techniques, is the ability to then simulate and optimize. The outputs of a marketing mix model include metrics that describe how marketing is driving sales based on the time on air and varying levels of weekly spend and impressions. These outputs, which are not available through other incrementality measurement, take marketing mix modelling to another level and allow brand marketers to directly move into action by simulating and optimizing their marketing budgets.
Marketing mix modeling is designed to be run less frequently than other incremental measurement methodologies and is used for long-term growth rather than short-term boosts in sales. While marketing mix modeling should be run every six to twelve months to inform strategic planning, brands typically see the largest improvement after running their very first marketing mix model. The accuracy of the insights gained from marketing mix modeling is also highly dependent on the data the brand provides.
Geographic Testing
Geographic testing for incremental measurement — sometimes referred to as geo-lift testing — is an approach that uses two similar groups of consumers in comparable geographic regions. One group (the test group) is shown an ad campaign or marketing strategy and the other (the control group) is not. The sales numbers or other KPIs from these two geographic regions can then be compared in order to provide the brand with insights into the effectiveness of their marketing efforts. Geographic testing is particularly beneficial for brands that want to measure their marketing campaigns' incremental return on investment when scaling up campaigns, giving them the insights necessary to confidently increase marketing spend in certain geographic regions.
However, geographic testing has several limitations. One of the biggest downsides to geographic testing is the lack of granularity in the data it provides; geo-lift testing only delivers data on the overall change in sales or other KPIs by test region, and not by other factors that may play important roles in the effectiveness of a marketing campaign such as medium, time of day, consumer demographics, etc. Geographic testing can also result in data that does not accurately represent national trends. For instance, geographic testing may show you that a campaign is highly effective in one area of the United States, but these findings may not apply to other regions, resulting in less promising results once the campaign is scaled up. Other limitations of geographic testing include contamination between test groups, inaccurate data or data quality problems, platform constraints, lack of long-term insights, and more.
How AI Can Be Beneficial for Marketing Mix Modeling
Marketing mix models that leverage artificial intelligence, like Circana’s Liquid Mix, are typically faster and easier to gain actionable insights from than traditional marketing mix models due to shorter measurement cycles and data interpretation assistance. Our Liquid Mix solution uses AI throughout the modeling process, from data cleaning, to data checking, to the model itself having machine learning capabilities, resulting in superior insights and a more intelligent tool across the board. Liquid Mix also provides brands with AI-driven insights that help explain what the data is saying in clear marketing language, reducing guesswork or interpretation. Liquid Mix empowers brands to run studies as frequently as needed without sacrificing speed or granularity.
How Many Individual Marketing Channels Can a Marketing Mix Model Evaluate?
Marketing mix models are highly adaptable and in the last two decades as they number of media channels and tactics used by brands has grown, so have marketing mix techniques and the data best practices. Circana models based on store-level data are able to provide much more granularity for marketing tactics than aggregate models that are based on national or brand-only models. As such, Circana’s marketing mix models are typically constrained by the number of marketing channels a brand uses, not the other way around. In order for marketing mix modeling to be most effective, it is typically recommended for brands that use four or more separate marketing channels, such as social media, TV, online video, print, retail media, etc. The effectiveness of marketing mix modeling can also be impacted by the amount of historical data your brand has collected for each of these channels. If your brand has only used a particular marketing channel for a very short period of time, the insights gained through marketing mix modeling will likely be less statistically reliable.
Get Accurate Growth Insights with Circana’s Marketing Solutions
Circana’s incremental measurement solutions like Marketing Mix, Liquid Mix, and Sales Lift provide brands with a third-party, independent view of their marketing channels using industry-leading point-of-sale consumer and market data, resulting in highly accurate insights that lead to confident business decisions that will improve incremental return on investment. Circana syncs all models into one platform and provides brands with access to actionable insights rather than raw data. Circana’s marketing solutions enable you to be confident in your budget allocation decisions.
