Marketing professionals are accountable for the marketing spend and the return on the marketing investment that they bring to the business. However, when it comes to measuring the impact or ROAS (Return on Ad Spend), marketers usually find the measurement methods slow, inadequate or outright misleading. Marketers are increasingly applying data science and advanced analytics to measurement to gain a competitive edge over their competitors. They use methods like Marketing Mix Modelling (MMA) or Multi-touch Attribution (MTA) – sometimes both – to help them understand the impact of their marketing initiatives. How do marketers select a particular approach? Which is a better strategy to adopt?
To begin, we need to understand a bit about each method.
Marketing Mix Modelling (MMM)
Marketing mix modelling looks at the historical relationships between marketing spending and business performance to help you establish your business drivers and marketing decisions—along with the best provision across products, markets, and marketing programs.
MMM incorporates historical aggregate data from a variety of sources in different formats and estimates the aggregated impact of marketing activities on desired outcomes. The data is at a campaign or market level instead of at an individual customer level. It requires the data to be regularly updated, cleaned, deduplicated and normalized while aligning time series.
Multi-Touch Attribution Modelling
MTA involves collecting information about individual prospects and customers through the various digital media that they interact with. The data usually includes every interaction that an individual has had with the brands’ messaging across paid search, display and video ads. Given the high proliferation of smart devices, a typical customer journey involves various touchpoints. MTA is a way of assigning credit to marketing touchpoints allowing you to see how each interaction is contributing in the customer journey.
There are three main types of attribution models that are adopted by marketers:
First Touch: In this model, the first touchpoint with a potential receives the entire credit.
Last Touch: This is the most commonly used model where the last interaction gets the full credit of the conversion
Multi-Touch: This model gives credit to every touch point. Depending on whether the model is linear, time-decay, U-Shape or W-Shape, the proportion of weightage may vary.
Time-period: MMM uses historical data usually spanning 2-3 years whereas MTA utilizes a much shorter window from 1 week up to 1 month based on cookie preferences.
Holistic Evaluation: Since MMM takes a variety of internal and external factors into account, MMM tends to be more holistic in its approach. MTA only considers digital marketing often ignoring offline interactions or other factors.
Incremental Impact: MMM is capable of reproducing the incremental impact of marketing since it shows the direct correlation of marketing on sales or other KPI metrics over a period of time. MTA model is able to reflect the short-term absolute impact of marketing.
Which method is better?
The choice of attribution model will depend on the brand and other factors such as:
- Media spend – whether the major proportion of the spend is on online or offline media
- Data availability – whether marketer has access to all the required data points at all times
- Data format – whether the available data is in suitable formats that can be utilized in the attribution model
While each method has its pros and cons, there are benefits of combining both MMM and MTA approaches. Combining the detailed view available from MTA with the more comprehensive analysis provided by MMM, the marketer stands to gain a more accurate view. While this approach could be ideal, there are a few significant challenges like data uniformity, duplicity and granularity that marketers may face.
The marketer should choose the most suitable approach based on data availability and brand presence.