The Best Identifier of Needs is Behavior
When marketing to customers, dividing them into subgroups with similar needs provides for optimal results: this is frequently referred to as need-based segmentation of the customer population. However, since it is hard to know everyone’s needs, marketers use proxy information like (a combination of) customer attributes. As social information became available over the last decade, marketers started using it to enrich the demographic attributes. But what people say is not what people do. Or to go Apple-far, sometimes customers do not know their own needs.
Cohorting started as an advanced method to segment users based on start of product use (e.g. users who started using product in March of 2014). Observing that cohort (group) of users provided insights like when does average usage drop off, or what percent of users remains active. You can get insights into your product over that timeframe, or compare one time period with another: which cohort found the product stickier — March’14 cohort vs June’14 cohort.
In addition to user start date cohorting, or simple user attribute cohorting, one can cohort users by behavior. For example, all users who read reviews prior to buying a product can be a behavioral cohort. Once you cohort users by behavior, you can ask interesting questions like are the users who read reviews having a higher conversion rate (buying a product more frequently) than those users who do not read reviews. Or are they more engaged: longer sessions, more time in app, venturing deeper into the funnel with fewer drop-offs?
Behavior Data + Social Data + Demographic Data
Demographically predisposed users who read reviews and mention those products socially, might have them promoted, yielding an uptick in conversions. Behavioral data enlivens social and demographic data. We have seen clear trends emerge once customers are segmented by behavior. We invite you to discover the insights in your own data after a behavioral segmentation. You will see demographics for what they are — customer attributes rather than proxies for customer needs.