burger menu

Customer segmentation and clustering algorithms

16th February 2018 by Emily Davies | Data

A company’s customer base may consist of thousands, if not millions, of unique individuals. Marketing to these individuals presents a large challenge because if you try to market to everyone, the message could be too vague, but creating a marketing plan that appeals to every single individual is unrealistic.

Why do we need customer segments?

Customer segments allow you to understand the patterns that differentiate your customers. Here are some valuable ideas of what you can do with segmentation analysis:

  • Increased understanding of customer needs and wants, which in turn can lead to increased sales and customer satisfaction.
  • Identify the most and least profitable customers.
  • More focused marketing efforts – products can be emphasised through more targeted advertising using more appropriate media to enhance message delivery.  
  • Develop products which best appeal to different segments of customers.
  • Companies cannot satisfy all potential customers, all the time. Using segmentation procedures companies can focus on satisfying those segments that they assess to be the most attractive for their products. Once an attractive segment has been identified, an appropriate marketing mix can be developed.
  • Build loyal relationships.

As you analyse your own customer base, it will soon become clear that there are some distinct groups with specific requirements. This enables you to develop a deeper understanding of your customers and discover what makes them tick.

When you’re communicating a message, it will be more effective if the recipient of the message finds it relevant.

What is customer segmentation?

It’s no secret that some customers are more profitable than others. But to be profitable in the long run, businesses must have a clear understanding of how profitability correlates with customer segmentation. Recognising the differences between customers will allow you to tailor your approach to the needs of varying customer segments.

Customer segmentation is the practice of dividing a customer base into groups of individuals that are similar in specific ways. Customer segments are usually determined on similarities, such as personal characteristics, preferences or behaviours. Understanding your customers–their similarities, their differences–is one of the most fundamental and important steps in quantifying the customer’s relationship with your product and company.

How to segment?

Segmentation doesn’t have to be complex. For a small company, it could be about recognising that you have two or three distinct customer types with different needs.

A few common methods to segment customers include:

  • Geographic
  • Demographic
  • Psychographic
  • Behavioural

A simple example of segmentation would be: If you run a hairdressing salon, the type of offers you might make to customer groups would certainly differ on gender and age. If you own an online clothing store, you might be better off analysing buying patterns and splitting customers into groups per how much they spend, how often they buy or what products they are most interested in.

By grouping together all the customers who regularly buy certain products, you can target them with relevant offers encouraging them spend more on products related to this segment’s spending habits.

It’s also good customer service to recognise the interests of your customers. Communication that acknowledges what they’ve bought and when, is much more impactful than a one-size-fits-all email, offering a product to a customer who is totally the wrong fit. This personalisation will build relationships with customers and grow loyalty.

Now What? – Cluster Analysis

There are multiple ways to segment a market, but one of the more precise and statistically valid approaches is to use a technique called cluster analysis. Clustering is the process of using machine learning and algorithms to identify how different types of data are related, and use these differences for grouping objects of similar kind into respective categories.

These homogeneous groups are known as “clusters”. Some common clustering algorithms include k-means clustering, hierarchical clustering, spectral clustering and mean shift clustering.

Unlike the rule-based segmentation process (segmenting by Age, Frequency of Sale, Income, etc.), clustering algorithms are not based on any fixed rules. Rather, the data itself reveals the customer prototypes that naturally exist within the population of customers.

The Disadvantages of Rule-Based Segmentation

In rule-based segmentation, the marketer selects fixed rules, typically in two dimensions, and divides the customers accordingly. For example, segmenting by users who purchased more than 5 products, who are over 40.

  • As rules are predetermined the results usually meet initial assumptions, whereas in cluster analysis the data is given the freedom to find natural clusters, hence discovering more meaningful insight.
  • There will still be large variance within each segment. This is due to rules fitting many customers.
  • It is very difficult to perform the segmentation in more than two dimensions.

The Advantages of using Cluster Analysis for Customer Segmentation

  • Find underlying natural clusters using machine learning by discovering the differences between customers.
  • Able to segment customers over many dimensions.
  • Variances within each resulting cluster, whereas rule-based segmentation typically groups customers who are very different from one another.
  • Clusters could slightly change each time the algorithm runs, ensuring that the clusters always accurately reflect the current state of the data.

Get in Touch

Do you, or your business, need help or advice when it comes to implementing your segmentation strategy? We can help – get in touch today!

Share this post

Related Content

Yard runs the 10k

Well, it’s all over! All those weeks of training (or not) and a great effort fundraising have paid off and all 6 members of the Yard team completed the...

Fantastic February: A Month of Momentum

Firefox 22; One Tough Cookie

Mozilla‘s recent announcement to release a new version of their browser, Firefox 22, has caused quite a stir in the advertising industry. The new version is said to block...