Attributing Conversions in a Multichannel Online Marketing Environment: An Empirical Model and a Field Experiment
Journal of Marketing Research, 51 (February), 40-56, 2014
47 Pages Posted: 23 Jun 2015 Last revised: 12 Mar 2019
Date Written: 2014
Abstract
Current technology allows firms to produce a granular record of every touch point a consumer makes in their online purchase journey before they convert at a firm’s website. However, firms still depend on aggregate measures to guide their marketing investments in multiple online marketing channels, such as display, paid search, referral, e-mail, and affiliates, which upon click-through become conduits, or “channels,” to firms’ websites. For example, the widely used “last-click” attribution metric assigns purchase credit to the last touched channel and entirely ignores all the other channels a customer might have touched prior to the purchase. Such aggregate and incomplete measurements bias the investment decisions for subsequent marketing campaigns. This paper provides a methodology to attribute the incremental value of each individual marketing channel in an online environment using individual-level data of customers’ touches in their purchase funnel. We propose a three-level measurement model to analyze (1) customers’ consideration of online channels, (2) their visits through these channels over time, and (3) their subsequent purchases at the website, and estimate the carryover and spillover effects of prior touches at both the visit and purchase stages. Based on customers’ path data from a hospitality firm, we find significant carryover and spillover effects – for example, e-mails and display ads trigger visits through search and referral channels, while e-mails lead to significant purchases through search channels. Attributing the conversion credit to different channels based on the estimated carryover and spillover effects, we find that the relative contributions of these channels are significantly different from the contributions based on other metrics currently used in practice. A field study conducted at the firm’s website by pausing paid search for a week validates the ability of our proposed model in estimating the incremental impact of a channel on conversions. In targeting customers with different patterns of touches in their purchase funnel, our estimates help in identifying cases where e-mail re-targeting may actually decrease the conversion probabilities.
Keywords: Multi Touch Attribution, MTA, Online multichannel attribution, attribution modeling, touchpoint management, online advertising, display ads, search, carryover, spillover, field experiment, path analysis, purchase funnel
Suggested Citation: Suggested Citation