“If markets are to be segmented and cultivated, they must meet certain requirements. Segments must be measurable, substantial, accessible, differentiable and actionable.”
– Philip Kotler, Marketing Guru
I can’t emphasize enough how important segmentation is to strategic leadership. While I’ll focus this tool on segmentation analysis, segmentation as a concept drives many strategic solutions to problems. As an example, let’s focus on the elegant segmentation solution to always cooking the perfect steak at Outback Steakhouse. While most of us struggle to cook the perfect medium rare and medium steak, these guys always get it right, and they even cook every steak the same amount of time. How do they do it? They segment each steak by their thickness. If you order a medium-rare steak, they give you a fat steak, while if you order a well-done steak, they give you a thin steak. With each steak cooked the same amount of time, the thin steaks get more heat into the core and are cooked to well done since they are thin, versus the fat steaks which get cooked to medium-rare, since the heat doesn’t penetrate the thickness as much. Pure genius!
Segmentation will help you in so many situations, whether it is segmenting the complexity that plagues most companies, segmenting priorities into importance, segmenting customer issues into different workflows, or segmenting customers to enable you to better market and shape your customer value proposition. Let’s go over the different types of segmentation and the best practices of segmentation analysis.
What is segmentation?
Segmentation creates a lens from which to look at things in a new way. Segmentation is organizing items, products, markets or customers into meaningful subsets enabling improved insights, understanding, and ability to tailor and target solutions.
As a super simplified example, imagine over the past year a store had 20 customers. They tracked their spend per visit, how many times they visited, their gender, and age. The data on these 20 customers is below.
Just by looking at it, you would be hard-pressed to get any insight out of the data. It looks like 12 of the customers are male, the average spend is $28, customers on average visit 3.1 times, and half of the customers are 20-35 and the other half 36-50. Interesting, but not very insightful or actionable. But, if you organize the data by age segments, the data begins to paint a more insightful story:
And, then if you organize the data by male and female segments, the data tells a slightly different story:
But, the most interesting segmentation is when you combine the age and gender to come up with four distinct segments:
All of the sudden, by finding the right segmentation, the data tells an incredibly insightful story, that the core customers are 36-50 year-old females, who spend more than twice the average per visit, visit almost twice as many times, and account for more than 70% of the total sales, even though they only represent 30% of the customers.
Once the store understands that 36-50-year-old females drive sales, the store will be able to spend their marketing dollars better, tailor the store and merchandise more towards their core customers, and ultimately grow their business.
This type of segmentation analysis is typically called a demographic segmentation since the segmentation schema is demographic data such as age and gender. There are many more types of segmentation, with the major ones being:
Psychographic segmentation is segmenting customers or people by their preferences, tastes, values, and views. Psychographic segmentation is often used in marketing to target advertising and messaging, and product development and design.
Behavioral segmentation is segmenting customers or people by how their behavior, choices, purchasing patterns, and actions. Behavioral segmentation is often used for customization of online and offline experiences, targeting interactions and messaging, and generating insights into product and experience design.
Value segmentation is segmenting customers or things by value. There are many permutations to value segmentation including lifetime value segmentation, potential value segmentation, and RFM segmentation. RFM involves segmenting customers by the recency of their last purchase, the frequency of purchases over a time period and the monetary value of the purchases. Consumer-oriented businesses such as retail, restaurants, services, and airlines often depend on value segmentation in tailoring promotions and marketing. Business-to-business oriented companies often use potential value segmentation to prioritize targeting of prospecting efforts.
Product segmentation is segmenting products into families, use cases, or other categorizations. Internal analysis related to the performance, quality, and financials of different product segments relies on product segmentation.
Time segmentation is segmenting things by their age, or like periods of evaluation (e.g., same year or month). Time segmentation is often used to segment products, issues, customers, and other things, to understand the overall trends of the different segments.
While these are the most often used types of segmentation, you can segment just about any dataset. Segmentation is simply the organizing of items, products or customers into meaningful subsets. Any time you start analyzing a set of data to discover insights, there are often ways to organize the data to drive more insights. Any time you start problem solving, segmenting the problem or solutions correctly, can lead to tremendous insight and often the right answer.
Why is segmentation important?
To generate insights from data, you typically have to frame things in the right way. Segmentation is the analytic tool to frame data and information appropriately. Over the past 40 or so years, since segmentation became widespread, orienting the activities, decisions, and resources of companies on the right segments have created billions in value. Segmenting things the right way can lead to a high ROI paradigm shift in the allocation of effort and resources to be more focused on essential segments.
What are the best practices of segmentation?
Start with simple segmentation
At a retailer, we conducted a substantial survey to get psychographic, demographic, and behavioral data on thousands of our customers. We tied all this survey data to their actual transaction history and applied some serious amounts of advanced statistical analysis to the massive dataset. Yet, in the end, the most insightful segmentation was a simple demographic segmentation, where the core segment that drove a disproportionate amount of value for the retailer, was mid-to-high income, 35-55-year-old moms and dads.
Leverage pivot tables
Whenever you have a dataset, start segmenting the data using simple pivot tables. It will often open your eyes to some fascinating insights, which may lead you to even better questions and hypotheses.
Use the right segmentation for the situation
You typically segment a data set to drive insight or to better prioritize segments to more efficiently allocate resources and activities. Whatever your situation, a particular segmentation methodology can typically help you. If you’re having issues with the ROI of an expensive direct mail campaign, then use RFM segmentation. If you want to see if a new quality program is working, use a product and time segmentation. If you want to improve the closing rate of sales interactions, use behavioral segmentation to drive insight into what works and doesn’t work.
Outsource the statistics
Segmentation can get fairly complex as you are trying to figure out the right combination of explanatory variables that create a meaningful variance in output values, which is the domain of cluster analysis and statistics. You can outsource this type of advanced segmentation to fairly inexpensive experts.