Why Marketers Should Look Beyond “Last Click” Metrics

Marketers frequently use “last-click” or uniform- or weighted-average metrics based on customer touches to measure return on their marketing efforts. But a recent study suggests that such metrics offer a misleading picture of the payoff of dollars spent on search keywords, emails, and display ads.

In “Attribution Modeling: Understanding the Influence of Channels in the Online Purchase Funnel,” University of Maryland researchers Hongshuang (Alice) Li and P. K. Kannan develop a model that measures the incremental impact of channels and interventions on conversions at a firm’s website in a multichannel context.

While last-click and other metrics may provide “passable” results for products with a very short purchase funnel, a wider lens is needed for high involvement categories—like consumer durables and travel services—where customers may visit a firm’s website several times through multiple channels before a conversion occurs, they explain.

Email, display, and referral channels contribute much more significantly to conversions than the last-click model suggests.

Li and Kannan’s model encompasses three stages in the purchase funnel: (1) the customer’s consideration of online channels/sources, (2) their visits through these channels over time, and (3) their subsequent purchase at the website. They also consider the “carryover” and “spillover” effects of channels and interventions at both the visit and purchase stages of the funnel.

To test their model, Li and Kannan used data from 1,997 visitors to a travel and hospitality company’s website from late June through late August 2011. During this time, the visitors in their sample clicked through to the company website 21,046 times and encountered display impressions but did not click through 1,323 times. A total of 815 customers made 1,128 purchases.

Path data for each customer were captured through tools such as Double-Click, Omniture Site Catalyst, affiliate websites, and the company’s email campaign management system.

Their analysis offered a much different scenario of the relative contribution of channels as compared to the last-click attribution model. Specifically, where the last-click method attributed organic search with 25% contribution to purchase conversions, the authors’ model attributes it only 16%. Likewise, paid search dropped from 10% to 6%. Third-party referrals, on the other hand, jumped from 18% in the last-click metric to 24% in the authors’ model, email’s influence jumped from 12% to 19%, and display ads went from 4% to 7%.

In the focal firm, attribution estimates were used to charge franchisees for various marketing programs such as paid search, referrals, and other campaigns. “Even if the attribution ranks were only marginally different it would still make a sizeable difference for such appropriations,” Li and Kannan noted.

With U.S. spending on Internet advertising exceeding $40 billion per year, the budget implications of their analysis are clear. Their proposed model offers a flexible and nuanced framework that may help marketers allocate those dollars more wisely.

From “Attribution Modeling: Understanding the Influence of Channels in the Online Purchase Funnel” (MSI Report No. 12-115)

Related links

Online Ad Design and Purchase Intentions (2012) [Article]


You must be logged in to leave comments. Please login.



Employees of MSI Member Companies enjoy the benefits of complete online access to content, member conferences and networking with the MSI community.



Qualified academics benefit from a relationship with MSI through access to, conferences and research opportunities.



The public is invited to enjoy partial access to content, a free e-newsletter, selected reports and more.



Become a Subscriber

MSI's Online Library of 400+ reports, authored by marketing academics, offers new research and evidence-based insights

Read More

Priority Topics

Stay Informed

The MSI Mailing List

Subscribe to our email list to stay informed about upcoming events, news, etc.

Social Media