“We need to be able to use predictive modeling to identify when shifts in shopping behavior are most likely to occur, and estimate the direction, magnitude and duration of these shifts”, manager on Marketing Science Institute 2018-2020 Research Priorities
These tough times make it harder to look beyond the urgent short-term changes and also consider the more enduring changes to customer behavior and attitudes about priorities, consumption and brands. However, the value of actually doing so is high because these are the times consumers abandon old habits and ideas and pick up new ones. We are all looking for support and some good news in this crisis, and organizations that deliver will earn a place in our hearts for years to come. For interesting data on donation and production news in brand campaigns during COVID-19, see https://docs.google.com/spreadsheets/d/15fd5OU2rRtYcVDFcgFrPsxfL9yQiHR83YQeGIKZWhAY/edit#gid=0
This week I got the good news of publication acceptance for my paper with Bernadette van Ewijk on ‘Enduring Attitudes and Contextual Interest: When and Why Attitude Surveys Still Matter in the Online Consumer Decision Journey’. With long-term predictive modeling, we show that survey-based attitudes and online consumer behavior provide complementary information for managers of 32 brands in 14 categories, from low-involvement fast moving consumer goods (such as toilet paper, cheese and beer) to high-involvement services and durables (such as insurance, lodging and cars). While online search, web visits and page views are key for services and to explain same-week sales, survey-based attitudes are especially important to predict brand sales (3 months out) and for fast moving consumer goods. We use these empirical insights to below new framework relating enduring attitudes to the contextual interest expressed by online actions.
Our starting point is the familiar pre-purchase, purchase and post-purchase distinction in the consumer journey, shown centrally in the Figure. On the sides, we distinguish enduring attitudes from manifest contextual interest, which may show up both in online as in offline actions – such as visiting a retail store. Enduring attitudes and contextual interest show dual causality (P1): consumers are more likely to act on brands they have strong feelings about (Brand Right becomes Brand Right Now), but they may also develop feeling for brands they discovered when exploring (Brand Right Now becomes Brand Right). Next, both enduring attitudes and contextual interest metrics drive sales, but enduring attitudes excel in sales prediction and contextual interest excels in sales explanation (P2). After purchase, customers experience with the product/service can induce them to share it through word-of-mouth, to update their attitudes (left bottom of figure) and their repeat purchase behavior (right bottom of figure). Both sharing and attitude updating are more likely in high involvement categories while low involvement categories are characterized by habitual buying, in which scenario both online and offline pre-purchase metrics are less informative in explaining purchase (P3). The Figure explicitly shows the feedback loops of post-purchase to pre-purchase as, e.g. word-of-mouth by current customers feeds back into the actions of prospective consumers.
How can managers use our framework to assess metric importance? The above bar graphs show the % of brand sales that can be explained by the metrics and marketing variables (online and offline marketing budgets) in low-involvement categories (blue bars) and high-involvement categories (red bars). Brand Awareness continues to be an important driver of brand sales, while consumers’ answer to the question ‘which brand do you prefer?’ is less predictive of their buying when preferences may quickly change due to online comparison, search or reading Amazon reviews. Interestingly, offline marketing is still an important sales driver – a deep dive reveals it has great power in driving online behavior such as search and website visits, as demonstrated by other papers (e.g. Pauwels, Aksehirli and Lackmann 2016) covered in previous blog posts.
For the full paper, here is the pre-review version we submitted to the journal: enduringattitudesandcontextualinterestwhenandwhyattitudesurveysstillmatterinconsumerjourneypauwelsJIMsubmission. Please see the Journal of Interactive Marketing for the accepted version.
How useful is this to you? What are your experiences? Please let us know in the comments!