ROI in the New Age
The explosion in digital marketing has increased the need for media measurement tools designed with data-driven insights in mind. Before digital, many of the factors behind ROI were a secret because there simply wasn’t any way to access the information. New technologies have transformed preconceived notions of how sales happen into entirely new ways of thinking. The ability to measure ROI across multiple channels, devices and a growing number of touchpoints is a data-capturing dream come true. But when running and managing multiple campaigns across this massive tangle of historical trends, consumer behavior and other data, the dream can quickly turn into an attribution nightmare.
Truth in Attribution
A consumer’s final decision to purchase doesn’t simply span a few clicks from A to B. It’s a path cluttered with obstacles, distractions and invitations to stray into more exciting territory. It’s hard to know whether it was that snappy one-off welcome email or the result of untold exposures to an ad on Facebook that sealed the deal and prompted a purchase. Tracking the customer across all touchpoints and using that information to understand what they’re thinking, where they’re going next and why they decided to buy requires complex analytics.
The days of juggling and measuring just a handful of media channels are gone. Digital platforms continue to emerge and have vastly increased the number of variables to consider when decoding the customer journey. This is where attribution comes in. Following the winding conversion path across all touchpoints and media channels, in a perfect world attribution accurately assigns credit for revenue.
End of attribution story? Hardly. Attribution models like everything else, have multiplied, so efficient optimization of digital campaigns will mean knowing more about the various types of attribution models and how they work.
Single Click Attribution: Simple, but Broad
First touch and last touch attribution models are called single source and assign all credit to a single touchpoint. There are obvious blind spots with these models since they lessen the impact of other parts of the customer journey that are certainly factors that should be included. First touch attribution modeling is most useful in determining where new leads come from. Last touch attribution modeling focuses on the most recent interaction with the customer and doesn’t consider what steps contributed to building awareness. A study released last year reported that 44% of advertisers are last touch focused, still assigning all the credit for conversion to the last interaction. Overall, first and last touch models are a good fit for businesses new to attribution as they do provide a general view of the purchase behavior.
Multi-Click Attribution: Specific, but Complex
For businesses dealing with multi-source data, there is greater need to see the customer interaction in the broadest way possible. In this case, a multi-touch attribution (MTA) model is the best option. The ability to see channels that are underperforming in the media mix is one advantage of MTA and indicative of how important a role it can play in budgeting strategies.
A common multi-touch attribution model is called the even-weight or linear model. It credits each touchpoint equally. It’s great for focusing in on total customer journey engagement across a single campaign but isn’t an ideal optimization tool. Time decay, U-shaped, W-Shaped and other multi-touch models apply credit proportionally depending on a range of variables and have different advantages worth investigating.
Regardless of which multi-touch attribution models make the most sense, they should be preferable to single-source models. Multi-touch attribution better accomplishes goals related to optimization and provides the deeper data insights necessary for a more customized customer purchase path. Ultimately multi-touch attribution modeling is all about spending where the dollars will be the most effective. This is especially critical in digital marketing where there are both multiple media channels and touchpoints feeding into the total data pool.
Media Mix Modeling: Larger Picture
Expanding the tools used to evaluate overall impact on ROI across all channels means using both data-driven models like multi-touch attribution and media mix modeling (MMM). While data-driven modeling tracks engagement as consumers move down the sales funnel, MMM takes a longer view and seeks to gain high-level insights allowing greater understanding of the trends surrounding the overall campaign. These are events and trends that influence over years versus the shorter cycles in digital. Factors like seasonality, weather, holidays and brand authority can be evaluated through MMM modeling.
This in turn enriches the data coming from MTA that is feeding into the MMM but is largely focused on digital channels that can’t see offline conversions. External trends are left out of multi-touch attribution because its purpose is to be more granular and stick to discovery at the user level. Media mix modeling measures both online and offline marketing activities and provides a more complete story about how these are impacting sales and ROI. Advertisers need both MTA and MMM for a comprehensive, unified measurement approach that considers all data sources.
Effective Planning
Before embarking on any attribution modeling project it’s important to define success and get team-wide consensus on priority milestones. Remember that the model is only as good as the data that goes into it, so closing the gaps in the customer journey and other efforts to ensure high quality information is critical. Don’t forget about the people. Invest in expertise first and then the technology. Once those two things click, it’s safe to test, iterate and use what you learn to keep your attribution game strong.
Interested in tracking your touchpoints? Message us to learn more about setting up an intelligent attribution model and optimizing all the channels that feed it.
