But it’s not just something that the marketing team should worry about. It’s something that affects the entire company, from sales to product development to support.
A customer or market segment is the name for the grouping of customers that share certain characteristics. Understanding your customers–their similarities, their differences–is one of the most fundamental and important steps in quantifying the customers’ relationship with your product and company.
Not only does it tell you how to better serve current customer demographics, it also allows you to discover unmet needs and deliver better products and services to new groups of people.
User Experience professionals are usually more familiar with Personas, which differ from customer segments. Personas are profiles of people based on data that comes from ethnographic research, surveys and interviews. They are archetypical users whose goals and characteristics represent the needs of a larger group of users. They function as stand-ins for real users to guide decisions about design and functionality. In fact, many personas come from a customer segmentation analysis, so even if you won’t conduct one yourself, it’s good to know how they’re developed.
Why Segment Customers?
Customer segments allow you to understand the patterns that differentiate your customers. But collecting and analyzing data just to understand patterns doesn’t make sense unless you’re going to do something with it. Here are some valuable things you can do with segmentation analysis:
- Identify the most and least profitable customers
- Better focus marketing efforts
- Improve customer service
- Build loyal relationships
- Price products differently
- Develop better products
- Create personas
- Customize features
While the type of product or service will determine the customer attributes that are worth segmenting, there are some fundamental attributes that most organizations should be familiar with and collect data on.
Unfortunately most customer information isn’t collected automatically. Website tools like Google Analytics don’t track even basic demographics. Instead you’ll need to identify other ways of obtaining the data.
Typically this is done using a combination of existing customer lists, surveys, 3rd party reports and services (e.g. Dunn & Bradstreet) and some Internet searching. You should plan on cross-referencing data where possible.
Segmenting customers can seem daunting: where do you start and how do you segment? While that answer will depend on your goals and products, it can be helpful to think in terms of the 5 W’s–and an H (who, what, where, when, why and how) when segmenting.
At the most basic level you should know the demographics of your customers. The core “who” demographic questions include:
- Education level
- Marital Status
- Number and age of children
Knowing these basic demographics is interesting in themselves. For example, if 80% of the customers are aged 25-34 then this helps with identifying effective sources of advertising or products that this age group is more likely to purchase.
Knowing where customers live isn’t just an exercise to place pins on a map. Instead, it’s about understanding the geographic diversity or concentration of your customers. Geographic attributes include:
- Rural vs. urban
- Domestic vs. international
- City names and market size (e.g. San Francisco vs. St. Louis)
- Regions and states
In thinking about the “what” you should think about past, present and future. What have customers done, what they are doing and what they are likely to do.
What they’ve done
The easiest place to start is the past. What have customers done that distinguishes them? If helpful, think in terms of recency, frequency and monetary value (revenue and profits) for segments.
- Most recent purchase
- Total number of transactions
- Product experience (new vs. repeat customer)
- Years of experience
- Total revenue
- Total profit
- Time spent with support
- Number of Customers they’ve referred
What they do
Understanding the context, goals and motivations of customers help identify gaps in product features and opportunities for improvement.
- Motivations: business vs. pleasure
- Experience level: power vs. novice user
- Their goals: (e.g. looking for a lawyer vs. getting answers to a legal question)
- Their tasks
What they think
Look to identify attitudes and psychographics that differentiate.
- Lifestyles: traveler vs. homebody
- Values: frugal vs. spendthrift
- Technology: early adopter vs. tech laggard
- Personalities: risk seeking vs. risk averse
- Overall product satisfaction: low vs. high
- Active vs. occasional investor
What they are likely to do
You also want to think in terms of long-term relationships and the lifetime value of a customer. Using a combination of surveys and past behavior you can estimate:
- Likelihood to recommend to friends
- Likelihood to repurchase
There are often significant differences in the type of customer you have based on when you measure.
- Seasons: Christmas shopping vs. ordinary time
- Weekends vs. weekdays
- Life Events: after a baby, marriage, or move
- Daytime vs. evening
- Periodic activities: haircuts and toothpaste every five weeks
How do customers interact with the product or service?
- Online vs. in-store
- Phone vs. in-person
- Through a reseller vs. direct
It’s not essential to collect every one of these attributes for your customers, many won’t apply and some won’t make much difference. After you’ve collected the data, here are the next things you’ll want to do next.
- Estimate the size of each segment: Compute the percentage of customers in each segment. Use confidence intervals to understand how much percentages could fluctuate based on your sample. For example, if 5 out of 100 of respondents to a survey were over age 65, a good estimate about this segment is 5% and we can be 90% confident it’s between about 2% and 10%. In other words, it’s highly unlikely more than 10% or fewer than 2% of customers are over age 65.
- Estimate the value of each segment: Compute the average revenue and profitability by each segment. Also use attitudinal variables like the Net Promoter Score to differentiate between “good and bad” profits. Customers that generate a high proportion of revenue but who have a bad experience are more likely to say negative things and lead to ephemeral revenue streams (bad profits). Size alone is often less important than size with profitability. If only 10% of the customers are over age 65 but these customer represent 40% of the revenue, then that’s good to know for product development but also sales targeting. Look for the Pareto principle (the 80-20 rule) when segmenting to understand who the critical few customers are that are driving the most value for your organization.
- Cross-Tabbing: In the earlier example we used the single variable of age to estimate the percent of customers above age 65. After estimating the size and value of single variables you’ll want to “cross” more than one variable. For example, cross gender by age or company size by industry to look at the number of men in urban areas or large manufacturing businesses. Look to mix many of the different dimensions to find patterns or valuable segments.
- Cluster-Analysis: A more advanced technique to identify segments is based on clustering algorithms such as cluster analysis, factor analysis and multiple regression analysis. These techniques identify statistical patterns that are hard to detect intuitively. You’ll want to call in the professionals to help with this as it involves both sophisticated software and statistical know-how. It’s a service we offer when helping our clients determine segments and it often uncovers undetected patterns that shape better product design or more profitable selling.
If this is your first segmentation analysis, don’t try and do everything at once. Collect enough data to get a good handle on your customers but don’t hesitate to go in stages. Running smaller surveys with open-ended questions and conducting a few interviews can provide a lot of insight into what questions to ask.
Finally, try and focus your efforts by having some research questions and hypotheses to answer before collecting data, but don’t be afraid of following the patterns seen in the analysis. Oh yeah, and get as much buy-in as you can. You don’t want people arguing too much over who “owns” the customer relationship.