Column: Customer Futures: not yet passé
Commentary by Sasha Fainberg
Can any of us really predict the future? Not exactly. But, consider those who do not study the past – they are doomed to repeat it! As marketers, we can really be blinded by the future: Beacon technology, push notifications, the Internet of Things, wearables, content… the list goes on.
I suggest discovering the truth about your customers from past data. Transaction history, loyalty programs, surveys, ZIP codes, or promotional responses can be used, even if the state of that information resembles a kitchen junk drawer more than a slick business report [thus far!]. Past data has the advantage of pushing beyond speculation and communicating actual facts. A “data audit” can take the disparate data, cull the wheat from the chaff, and identify gaps. The best defense is to know your own unknowns.
Market research seeks to predict customers’ futures by reviewing past patterns: what they have purchased, which messages were most successful, and which advertising had the highest impact. One of the most successful ways we’ve seen this prediction happen is segmentation analysis. As quantitative data (typically from surveys) is sorted, significant “breaks” can be detected by demographics, habits, or agreement with value statements. Taking it forward one more step…
Modeling may just be the most exciting step about data analysis: a true predictor of behavior. Data can be plugged into software (guided by humans, at least for now) to project how this same segment will make future choices. For advanced reading, may we suggest a CHAID analysis that breaks down decision-tree style?
In this way, a company can understand, for example, how customers ages 40-55 who purchase >$4,500 of home services in urban areas are twice as likely to have been informed by Google-suggested ads than by the word-of-mouth that dominates the suburbs. Now, your company is REALLY equipped to enact a media buying plan without wasting a cent.
Data always tells a story. Are you ready to understand how your existing data can spell out your business future? Only after examining the past can we get back…to the future.