A few days ago a colleague asked around if anyone had an approach to segmentation in a world when social media is integrated with mobile and retail transaction data. So, here's the answer I came up with in an attempt to answer the questions: What did they do before buying? what content did they consume before trying?
Customer transaction data is like a silent movie, we see the final action but don't understand the nuances of how they got there. Just before the 'submit button' or landing page is a vast region of content and interactions that most likely influenced the choice a consumer made, we just can't see it and if we can, we can't make sense of the enormous quantities of granular data without some type of framework.
Our objective as marketers is to align solutions with customer's needs to the mutual satisfaction of both. And to do that we need to not only understand the benefit provided by the product/service but also how the choice was ultimately made and the path to purchase. So, the first opportunity is to segment people on what technologies, sources, devices etc. people use to consume information - are they mobile, social, or site users?
It is also generally accepted that various types of information are utilized when making a decision - we decide emotionally based on aspirations and needs, we defend the decision rationally based on facts and figures. This content can be divided into several types that influence choice – emotional or brand, promotional or incentive, informational or reference and communal or recommendation. If we could tag the content consumed according to such a framework then we have a new opportunity to segment based on content type.
Put the two together and we can build a picture of 'content consumption' based on type and channel.
By linking transactional data (customer value and product mix) where we understand the value of a customer based on LTV, visits or other metric we can create a series of product-specific Paths to Purchase that map content consumption to their value to the organization.
So take for an example a segmentation scheme based on standard RFM (recency, frequency, and monetary) metrics that produces for groups: Best, Frequent, Spenders and Uncertain. Over lay that on the type of content consumed and you might find patterns that impact how you market. A hypothetical view might look like the following chart.
For each value-based segment there are two content consumption segments:
- Best customers, those with above average spend and visits, either use branded content, maybe the self-help site you created, or they use a lot of recommendations from social sites.
- Frequent customers who spend less than the average but shop often, are either deal shoppers or search driven.
- Spenders, those who spent a lot but on only one transaction, are either attached to brand as a 'badge' or use information and search to find you.
- Uncertains, those with low transaction value and count, rely on incentives or the recommendations of others.