Blog

Get the latest news and views from the Datalytics Team

Segmentation

Posted by admin

Orange SegmentsConsumers are sharing more and more data and expect a greater level of personalisation as a result.  So how do we keep up?  Segmentation is one route to split the customer base into meaningful groups to target with more relevant marketing messages.

There are many ways to approach segmentation.  The customer base can be segmented based on demographics, purchase and engagement behaviour, digital and social interactions.  We recommend that you start simple and build it up over time.

 

An Approach

My personal preference is to start with purchasing behaviour.  By understanding the patterns of previous purchases, we can begin to estimate likely next purchase, opportunities for cross and up-sales and when retention might be required.

The key measurements needed on purchasing behaviour are (RFM):

  • Recency
  • Frequency
  • Monetary value

Consumers who have purchased from you recently, are more likely to purchase again.  The brand choice is fresh in their mind and hopefully product and service satisfaction is high.  Consumers who buy from you often, have already endorsed the brand choice, stated their loyalty and many are brand advocates.  Finally, consumers who spend more of their share of wallet with you are more likely to spend again.

So the data requirement here is to categorise consumers along these three measure and score and weight them accordingly.  Sum the scores across all three measures and rank the customer base highest to lowest.  Although there will be some exceptions and some outliers, generally the data should categorise into top, middle and bottom level customers.

Five Principles of Segmentation

Having gone to a significant amount of work to segment the customer base, you’ll need to be sure your approach has been robust.  To validate the segmentation, run it past these five key principles:

  1. Highly differentiated – are the customers in one segment different enough from their neighbours in the next segment to warrant being in their own segment? If not, re run the model changing the weights and scores on RFM.
  2. Homogeneous – within each segment, do the customers have enough in common to belong to the same sub group? Again, revisit the model to identify how it can be refined.
  3. Identifiable – is the segmentation or classification recognisable to stakeholders outside the data team? Does it make business sense to have the customer base categorised this way?
  4. Clearly Actionable – we are all unique, aren’t we? So we could have segments on one, but we’d never be able to develop a marketing plan to drive actions at that level.  Start out with somewhere between three and eight segments and you can get more sophisticated as you go.
  5. Stable – while customers will change segments overtime, ideally in response to a targeted marketing campaign, the model should be stable enough to enable a strategic plan for growth to be developed and implemented.

What are the benefits of segmenting the customer base?

The benefits are wide ranging and should be tracked and measured to demonstrate the value in segmenting the customer base.  Depending on the rational for segmentation, you may be looking to address a retention issue and so an increase in retaining the most valuable customers should be tracked.  Perhaps your business is going through a growth spurt.  Segmentation can be used to profile your optimum customers and support the acquisition of more of the same type of customers.  Ultimately segmentation will open up opportunities for data guided decisions that will enhance the performance of your marketing campaigns and have an impact on the bottom line.

If we’ve hooked you in and you’d like to talk more about segmentation, give us a call, we love to talk about data.

05 April, 2016 by Tara Grehan

Got a project you’d like to talk through?

Call us on Tel: +353 1 644 9828

or Email: info@datalytics.ie