March 19, 2015
When it comes to customer segmentation, customers tend to be pushed into pre-fabricated segments driven by criteria such as demographic, age, and income. And as we become more sophisticated, transaction details, online usage and CRMs have allowed for more specific segments. Focus groups and surveys provide even more data. However, when it comes to defining a segment around this more qualitative data, it is still up to marketers to make assumptions or educated guesses in order to make customers fit into a certain bracket, (all nice and tidy).
All this guesswork, and rigid structures forcing customers into static segments, has led to ‘multiple segment’ targeting to ensure we are capturing as many of the right audience as possible. However, this may mean that the messaging used may not be correct at all.
In short, customers can’t be boxed into rigid segments. There are too many influential factors to take into account – for example, peer reviews, trends and memes. Even offline factors like the weather can change customers’ decisions and purchasing intentions. With multiple stimuli from brands and peers it is hard to keep up with the ever evolving customer.
But what if we allowed customers to segment themselves naturally?
A new trend that is reaching the digital world is natural segmentation and building a ‘segment of one’. This allows customers to be in charge of their own segmentation.
So how is this achieved?
With the high computing power now available to us, it is possible to combine existing customer data, such as CRM or transaction information, with social data taken from logins. Social logins have given marketers access to a previously hidden world, i.e. customer likes, dislikes, music listened to, things they find funny, charities they support, etc.
All this data is then gathered into a topic graph; a flexible dataset that can reveal the relationships between millions of topics. Algorithms can then be applied to identify patterns and trends and groups of interests will begin to appear.
By using models that detect convergence, natural occurring segments can then be identified.
Putting it simply, natural segmentation is all about using interests gathered directly from consumers, which, when aggregated, can reveal a bigger picture. As this data is ever changing, it allows customers to dynamically self-segment themselves, based on their affinities and interests. This stops marketers from making assumptions about what else their customer may or may not like.
And why is this important? Knowing this information will allow brands to add value to their content when targeting their “segment of one” and will allow brands to build more one-on-one relationships and loyalty. Ultimately it will increase the lifetime value of your customers.
Source: The Marketer