Advanced methods like fractional attribution, third-party SDK services, and consumer ID technology help piece together more accurate digital attribution models.
Faced with an increasingly mobile consumer base, marketers are devising newer methods for more accurate and detailed digital attribution. These practices and technologies measure digital advertising effectiveness while tracking complex customer journeys across channels.
And desktop is just half the battle. Mobile devices — including smartphones, tablets, and other connected devices — have quickly risen to the dominant mode of internet browsing and communication. Overall mobile browsing traffic surpassed desktop last year, and 77% of Americans now own a smartphone. Experts predict there will be 11.6 billion mobile-connected devices spread across 5.5 billion users around the world by 2020.
Given that most people now own multiple devices, tracking their complete customer journey and giving key touchpoints their due attribution has become more difficult. Someone who researched a product for days on mobile but eventually bought it on their laptop could appear as two disconnected people, even though they are one and the same. Without proper attribution, marketers don’t know which touches had the greatest, or any effect on conversions, and thus deserve more of their attention.
“Applying an attribution model to assess the contribution of different digital advertising elements is vital for advertisers if they want to understand the effectiveness of their digital marketing mix,” according to the UK’s Internet Advertising Bureau. Proper attribution enables marketers to allocate marketing spend sensibly and optimize campaigns effectively.
Thankfully, new developments in attribution are helping to give each digital touch its due credit.
Evolving the Model: Multi-Touch Attribution
Multi-touch attribution (MTA) models, also known as fractional attribution, trade out the traditional “last click” or “post click” model, where 100% of the credit goes to the last thing a converted lead sees or does. Instead, each touch is addressed along the customer journey and weighed to algorithmically assign a percentage of attribution based on its influence.
“The approach appears to be gaining steam,” observes Mobile Marketer. Based on a Mobile Marketing Association survey, “more than 150 of the top 500 marketers are using MTA, and an additional 250 plan to implement it in the next 18 months.”
Under a fractional model, key touches are recognized for their contribution to an endpoint, usually a completed sale. If for instance, someone sees an ad on a favorite blog for a new laptop, bookmarks a few reviews, but ultimately buys the laptop after seeing an ad on social media, then post-click models give 100% credit to the social ad. Fractional attribution breaks the 100% into smaller pieces to acknowledge each touch.
The challenge with splitting up attribution is that ad publishers are wary of sharing credit with others. Partial attribution becomes a zero-sum game because any increase in attribution to one takes away from another. Additionally, assigning weight to touchpoints — or which ones influenced a purchase the most — are hard to interpret.
Ultimately, rigorous testing and common standards across industries are needed to settle debates about who deserves credit beyond the final click.
Bringing Mobile Into the Fold
Mobile comes with its own set of challenges. Apps with built-in or grafted-on ad platforms used to have difficulty tracking attribution for clicks and purchases from display ads. “In this new, siloed world that mobile has created, referrer tracking doesn’t work the same way as the web,” explains one attribution service provider.
Recently, third-party services have created ways to help app owners attribute these actions more accurately, which allows them to earn a share of the ad commission. To accomplish in-app attribution, most third-party services rely on a software development kit, or SDK, which adds code to send data from outbound ad clicks. This can allow app owners to monitor activities more effectively, and advertising companies can connect identities between in-app actions to those across the larger web.
As channel movement increases from app to mobile web, from push notifications to app, and from other cross-channel activities, tracking consumer activities through every channel becomes vital, but still very difficult. Companies willing to invest in attribution — be it third-party or in-house solutions — will benefit from a more complete customer journey that provides better attribution across channels.
Cross-Device Tracking Using Consumer ID Management
In a recent post, we covered the fundamentals of consumer identity and its two main methods — deterministic and probabilistic. These methods for connecting identities are crucial for fusing cross-device actions into a single, cohesive journey. The importance of this capability is particularly evident when trying to obtain accurate attribution.
Forty percent of all e-commerce purchases involve multiple devices along the consumer’s journey. Without a way to connect actions taken on mobile with actions taken later on a different device, the customer’s journey appears fragmented and incomplete. By using an identity management solution, such as verifying ID probabilistically through an identity graph and algorithms, these journeys become complete and each touch can be attributed more properly.
Evolving the Attribution Framework
Many of the latest developments in digital attribution have less to do with new technology and more to do with new ways of using old technology. Namely, marketers are shifting the way they measure, strategize, and approach attribution models.
For instance, some companies define “cohorts” — users with similar patterns of behavior — and then analyze the different reactions of the cohort to different variables. For example, a cohort that uses an app daily may be more likely to purchase a product after reading a review. Based on this information, a company can prioritize conversions to product review pages as opposed to conversions to a product page or landing page offer.
Similarly, someone who spends a long time on an e-commerce site without making a purchase can present a mystery to marketers. But using tools to track screen touches or mouse movement creates granular data about browsing behavior and reveals possible triggers needed to convert. Observations like these form the basis of hypotheses used for A/B testing. By moving away from metrics taken for granted toward ones that explain behavior, better attribution models and more effective customer journeys can be developed.
Through internal changes, advancing technologies, and new approaches like these, it is easier for marketers to accomplish what was once difficult: accurate digital attribution. Further advances will require changes in the way the industry chooses to operate as a whole, especially when it comes to fractional attribution. Ultimately, these changes will pave the way toward a more nuanced and accurate view of the customer journey.