Generating insight about your customers is a powerful enabling force behind better decision-making. Yet you still hear people say:
“We have spent money on customer insight: research and profiling, and yet we didn’t really get any value from it or understand how to use it.”
Sound familiar? These circumstances are not uncommon. The reason behind this is often due to the generic nature of the insight that is generated. For the majority of SMEs generating insight is not a core competency. So getting insight is something that is usually done two ways. By generating a survey using a free tool like SurveyMonkey, which allows them to gain direct feedback from customers. Or by buying a report from an agency like Experian or Call Credit, which provides geo-demographic segment information about customers based on their address. There is nothing wrong with these methods, but the outcomes can often be frustrating.
Surveys can provide some interesting feedback. But usually in my experience they’re crafted in a way that only really yields a high-level set of feedback that often leads to more questions than answers.
As for geo-demographic information, it too can provide quite a broad set of information which can be hard to translate to your specific business, what you sell and your market. It’s interesting to know the proportion of your customers who are Aspiring Homemakers vs. Transient Renters (two of the segments within Experian’s mosaic segmentation). But being interesting doesn’t necessarily translate to business value. Companies often have to interpret these reports creatively to try and get something out of them.
Going down the survey or geo-demographic profile route is really akin to dipping your toe in the water when it comes to customer profiling. The reality is that we’re complex individuals. While we share some characteristics with our fellow segment members we can be quite different in other ways.
Customer attributes and individual profile building
So rather than thinking about creating a single segmentation, and profiles for each segment within it, companies should think about creating a number of customer attributes, which in combination can create a profile for each individual customer.
Here’s an example for a retailer who sells men’s and women’s shoes, both smart and casual.
First of all you’d want to create a gender attribute for every customer. So every customer is either male, female or unknown.
Then you’d want to create customer lifecycle attribute for each customer based on whether they have bought or not, how many times and how long ago.
Then you’d want to look at high-level product preferences, so each customer is either smart, casual, both or none
Then you might add colour preference; black, brown, blue, multiple, none etc.
Then you might want to add something about a customer’s value or potential future value so you have an idea how important they’re likely to be financially.
This list of attributes could go on to include: favourite products, recommended products or categories, campaign engagement, channel preferences and so on.
Here’s what a snapshot of 4 example customers might look like for this retailer:
Using attributes, we could conclude that:
Customer W is a new female buyer, of casual brown shoes and is quite valuable. She is a great target for a follow up purchase and to build a relationship with, to prevent her spending elsewhere.
Customer X is a high value male buyer who has bought smart and casual shoes. He is someone to target with exclusives as a reward for the amount he has spent so far.
Customer Y is a female buyer who bought a smart pair of black shoes a while ago; maybe for work. So perhaps she would benefit from a win-back offer for another smart pair of shoes.
Customer Z hasn’t bought yet but we know they are male. They have subscribed to email newsletters only. Getting them to buy for the first time is the priority so sending them campaigns about popular male shoes could be prioritised.
Following this ‘bottom-up’ approach of building a granular profile of each customer allows a marketer to test different tactical marketing initiatives to build engagement and encourage a purchase. More importantly it allows a marketer to personalise the way they communicate with the customer, be that through email, customer service interactions, Facebook and indeed through the online store itself.
A ‘top-down’ segmentation approach to putting customers into one of, say, 6 segments might actually result in grouping customers from the table above together. As a result, it dilutes the marketer’s ability to connect with the customer in a more one-to-one, personal way. Imagine you have excess stock to clear in a particular line and colour of shoes. Without the bottom-up granular approach of creating customer attributes the marketer has little choice but to send a mass communication out to all customers or a pre-defined segment of customers. This creates irrelevancy and can put recipients off. Using customer attributes, allows the marketer to focus on only those customers who are most likely to buy.
Why bother doing this?
According to the Connected Commerce survey, “62% of respondents claim they buy more and/or more often when met with personalised retail experiences.” To me this provides hard evidence for something that supports a more socially intelligent way of marketing that simply makes sense. Why wouldn’t you want to make the experience relevant if it is possible to do so?
About the guest author
Will Young is the CEO and Co-founder of rais a Customer Intelligence and CRM software start-up for SME retailers. Since the start of his career at dunnhumby, Will has worked with many retail clients, helping them make sense of their data and improve customer acquisition and retention.
rais (retain. acquire. interpret. sell) is software that integrates customer data, automatically generates customers attributes, reports on customer performance and enables users to act across incumbent communication channels (email, facebook, customer service etc.)