One of the essential topics taught in introductory marketing courses is the concept of market segmentation, which is the division of a market into groups of consumers that share one or more characteristics. Although most marketers understand the value of segmentation as a means to better target their efforts, many are daunted by the task of applying segmentation in an age of Big Data, where a deluge of information—both internal and external—is available. In this post, I’ll quickly review traditional B2C and B2B approaches to segmentation and touch on the concepts of anti-segments before relaying story of how a retailer used micro-segmentation to significantly impact the probability of sale.
B2C Segmentation in the Age of ‘Mad Men’
If you took an introduction to Marketing class more than a decade ago, you were probably taught that there are four primary bases for segmentation: geographic (e.g. region, population growth or density), demographic (e.g. age, gender, education, income), psychographic (e.g. values, attitudes, lifestyles), and behavioral (e.g. usage patterns, price sensitivity).
Although the recent popularity of the hit TV show ‘Mad Men’ may make some nostalgic for simplistic demographic segments like ‘housewives’ or ‘professionals,’ segmentation of B2C markets today is very granular and typically accounts for specific online and offline behaviors.
B2B Segmentation is a Different Animal
For business (B2B) markets, however, these variables are often less helpful. Traditionally, B2B marketers relied on SIC or NAISC codes to perform industry segmentation. However, these code systems are antiquated and new industries and business models—such as cloud services, SaaS and social media—are not represented.
B2B segments are typically based on such factors as company size, buyer type, purchase criteria, etc. A recent survey by the Content Marketing Institute (CMI) and Outbrain of 1,416 business-to-business (B2B) marketers from North America ranked the frequency that B2B small business marketers targeted their content based on such characteristics (see chart).
The advantage of segmentation schemes based on these traditional approaches is that they produce a relatively small number of easily understandable segments. The disadvantage is that they produce relatively broad segments that may have limited effectiveness.
Anti-Segments (or Explicitly Defining Who You DON’T Want to Sell To)
First introduced in The Lean Entrepreneur – How Visionaries Create Products, Innovate with New Ventures, and Disrupt Markets, an anti-segment is simply a market segment that you want to avoid selling to. As outlined by Brian Gladstein in a recent blog post titled “Turn Away a Customer? Yes… If They Are In Your Anti-Segment,” there are a number of reasons why this might be the case, including:
- You can’t reach or service these customers profitably
- They have product requirements that would consume your resources, making it impossible (or extremely difficult) to meet the needs of the rest of your customers
- These customers don’t provide the upside that other segments do, in the form of longevity, up-sell opportunities, cross-sell opportunities, or other sources of revenue
Why take the time to define anti-segments? If you don’t define your anti-segments, incorporate them in your internal sales and marketing training, and embed them in your targeting approach, you will find yourself squandering marketing dollars to move them through your funnel and wasting sales bandwidth to close them. It’s a classic problem where opportunistic sales may increase short-term revenue, but these customers and the accompanying lack of focus may ultimately sink the business.
Segmentation 2.0 or Segmentation in the Age of Big Data
In today’s age of Big Data, a multitude of new types of data are readily available. These include:
- Activity-based data, e.g. web site tracking information, purchase histories, call center data, mobile data, response to incentives
- Social network profiles, e.g. work history, group membership
- Social influence and sentiment data, e.g. product and company associations (e.g. likes or follows), online comments and reviews, customer service records
This data explosion enables the definition of increasingly finer segments. These micro-segments enable ever finer targeting of content, offers, products and services, which can deliver real and substantial returns.
Case Study: A Retailer Increases the Probability of Purchase Using Segmentation
In a recent project, a house ware retailer with more than 400K customer records in its Salesforce® CRM system wanted to increase sales by better leveraging various streams of customer data they had been collecting. These included:
- Basic demographic information, such as age, income range
- Geographic data, such as location of purchase (store and/or web)
- Behavioral data, such as purchase history, website viewing history, shopping basket contents
While identifying meaningful segments from such data is a fairly straightforward project for an experienced analyst or data scientist with knowledge of the appropriate tools, the retailer had no such staff on hand. Instead, they turned to a SaaS solution that leveraged advanced analytics and machine learning to analyze the data and present the results visually. Even though this retailer had multiple products purchased at multiple prices, the analysis was setup and completed in only a day, rather than weeks.
The analysis results increased their understanding of customer loyalty and customer life-time value (CLTV), which helped them develop a better sales engagement and product offering strategy. The 16 micro-clusters produced by the analysis were immediately deployed for better targeting of offers delivered by email and via the website.
Six weeks after deploying the results, the retailer was able to increment revenue by more than $1 million by significantly increasing the probability of a sale to these 16 micro-clusters, increasing the probability of purchase of the next product suggested and decreasing overall churn.
Now It’s Your Turn
The use of micro-segments based on past customer behavior—such as observing product / service purchases over a period of time, price changes, response to incentives, location and many other factors—allows for diminished turnaround times and easy to invoke programs. And with new solutions that employ sophisticated analytics, machine learning and visualization, market segmentation in the age of Big Data is easy to implement, while producing truly actionable results.
For more information on how marketers can quickly implement micro-segmentation without having to hire a data scientist take a look at the Quant5 Customer Segmentation solution.
A version of this post originally appeared on the MarketBuildr Blog.