TABLE OF CONTENTS
Incrementality, in marketing, is the measure of desired outcomes in the form of direct conversions, website visits, revenue, clicks, etc provided by the marketing activities performed by a team. It is essentially a measurement that serves as a north star for all your advertising management activities.
Incrementality gives an overarching understanding of the activities that incrementally add value to your customer, so teams can accurately estimate and analyze the return on efforts made on a marketing channel.
The terms attribution and incrementality are often used together, but they mean very different things. Attribution is the process of attributing credit to a particular channel or touchpoint in your marketing campaign. In most cases, teams only account for the last touchpoint which gives a skewed notion of what marketing activities actually contributed to the conversion.
As the attribution model suggests, most contemporary marketing campaigns make use of different mediums, channels, and touchpoints to enable conversions. In this case, attributing the right value to these mediums serves an important role in future advertising decisions. Incrementality, helps with just that, providing marketers with a reliable way to measure the true effectiveness of any advertising activities.
It’s also important to understand if the favorable outcomes are a result of advertising efforts or if they are organically generated. Based on the results of incrementality testing, teams can accurately understand if the conversions are due to the additional advertising campaign. This will make sure that teams only spend on advertising as necessary without blindly pumping cash into advertising campaigns.
Incrementality testing and measurement essentially answer the following questions -
Next, let’s dive into the different types of marketing incrementality and how they are used by marketing teams.
Channel-silo incrementality, primarily asks the following question - How many desired outcomes received from paid marketing can be achieved through an organic search? Incrementality measurement can help teams figure out this metric through controlled experiments on audience segments, where one is exposed to paid advertisements and the other is not. With the channel-silo, incrementality teams can focus on optimizing their marketing efforts in an individual channel.
Some marketing campaigns span across a number of channels that can be clubbed into a single stack. For example, teams generally run advertisements on various Google-owned platforms. Testing for incrementality across these cross-stack platforms is considered as cross-stack incrementality.
Most cross-stack channels in different contexts are known to produce similar results. Hence, performing analysis on cross-stack behavior of customers in controlled regions so teams can optimize spending for an entire stack can be quite a valuable incrementality tactic.
Incrementality hopes to measure the impact of marketing efforts across all channels and mediums, which can turn out to be a mammoth task. But the end goal remains the same nevertheless. Marketing teams have to answer the question, “What is the incrementality across all marketing activities that require resources?" Performing this analysis across all marketing channels is called marketing-portfolio incrementality.
This can turn out to be a difficult task, but if performed with precision can result in invaluable insights about your ad spending. Since most marketing channels are also very dynamic, having a sense of incrementality in that channel can have far-reaching effects.
Now that we know what incrementality means, let’s understand how marketing teams can test incrementality. Incrementality testing is performed similarly to how A/B testing is performed. The basic idea is to segment the audience into 2 main subgroups - group A and group B.
The goal of segmenting the audience is to understand the incremental difference in metrics caused by ad spending. So naturally, each group is exposed to a different campaign. For example, group A will see no ads, so their conversion metrics will all be organic. In the case of group B, they will be shown all the ads currently active. Here it is made sure that the audience types in both the groups are nearly identical, making sure to avoid confounding variables that might affect the results.
Now, based on the incremental difference in group B with respect to group A, teams estimate the value of ad spending. The increase in conversions or installs in group B compared to group A is known as lift. So if group B witnesses 50 more installs compared to group A then the lift is calculated to be 50.
Similarly, incrementality is the percentage of converts in group B that was enabled due to ads. With this information, teams can estimate the cost of each incremental conversion
As discussed earlier, there are two terms that we are dealing with in incrementality testing. The term “lift” refers to the increase in desired metrics that the incremental conversions bring in. Incrementality, on the other hand, represents the portion of conversions enabled by paid media.
To calculate incrementality let’s first consider an experimental test case that actively withholds paid advertisements from a section of the audience/subgroup. Now that we have two segments of the audience, we check the conversion differences between these two segments.
Group A be exposed to paid ads
Group B is not exposed to paid ads
Therefore, the incrementality formula is given as;
Incrementality = (Conversion Rate of Group A - Conversion Rate of Group B)/ Conversion Rate of Group A
So, for example, if Group A witnesses a conversion rate of 13% whereas, Group B witnesses a conversion rate of 10%, the lift is 3% while incrementality is as follows -
Incrementality = [(13 - 10) / 13]*100 = 23%
An alternate method to incrementality testing is the last-click attribution method. In this method, the success is attributed to only the last click or the last touchpoint of a marketing campaign. Suppose, for example, your customer navigates through a bunch of different touchpoints in their customer journey to perform the desired outcome in form of an installation, subscription, clicks, etc. In the last-click attribution method, only the final action is taken into account ignoring all the different touchpoints or ads the customer initially came across that eventually facilitated the conversion. For any such multi-faceted campaign, the last-click attribution method doesn’t provide a clear understanding of what enabled the conversion.
There are also other methods, like media mix modeling, an analysis method that uses historical data from marketing and non-marketing sources to facilitate marketing decisions. Another such method is the multi-touch attribution method. However, both these methods only help measure clicks and not impressions and views.
Incrementality testing accounts for all the variables providing the true incremental contribution of your paid advertising.