TL;DR
Market segmentation is the process of dividing a market into distinct groups and tailoring strategies to each. It’s widely used across industries—from marketing and advertising to pricing and customer service. Key terms include:
- Customer segmentation: Focuses only on existing customers.
- Segmentation study: Uses surveys and algorithms to identify segments.
- Segmentation algorithm: Groups people using techniques like k-means or latent class analysis.
- Segmentation strategy: Involves choosing the best segmentation and deciding how to target each group.
What is market segmentation?
Market segmentation involves splitting a market into segments and developing different tactics and strategies for the segments. The term market segmentation is often used interchangeably with the terms segmentation, segmentation study, customer segmentation, segmentation strategy, and segmentation algorithms, although there are meaningful distinctions between each of these terms.
Download your free DIY Market Segmentation ebook
The 9 Phases of Market Segmentation
Segmenting a market is a structured process with nine key phases. Here’s an overview of the full process:
- Strategic decisions: Define the business goals and determine whether segmentation is the right approach.
- Define the market and unit: Identify which market you’re segmenting (e.g., consumers, users, buyers) and the unit of analysis (e.g., individuals, households, businesses).
- Collect data: Design and conduct surveys or gather internal data relevant to the segmentation.
- Build segmentation database: Clean, prepare, and structure your data to make it ready for analysis.
- Statistical analysis: Apply segmentation algorithms (e.g., cluster analysis, latent class analysis) to identify patterns in the data.
- Create segments: Define and label the segments based on the statistical results, ensuring they’re actionable and distinct.
- Communicate: Develop visualizations, presentations, and narratives to help stakeholders understand and adopt the segments.
- Implement: Use the segments across functions—e.g., marketing, product, sales—to inform strategy and tactics.
- Refresh: Periodically review and update the segmentation to ensure it remains relevant and accurate.
Applications of market segmentation
All firms segment their markets. Many people in business spend their days developing and working with segments. For example:
- Marketers use market segmentation in their planning (e.g., Coca-Cola develops some products for the diet segment).
- Advertising agencies use segmentation to assist in developing advertising strategies. For example, working out which message will resonate best with different segments.
- Media companies use segmentation as a tool to assist in efficient media buying. This could be purchasing Facebook advertisements to target moms, and ads in games to target young men.
- Direct marketers use segmentation to better target their campaigns. For example, targeting more campaigns at the segment of households that have historically been found to respond more regularly and/or spend larger amounts.
- Call centers use segmentation to work out the best way of adapting their scripts to differences between customers.
- Retailers use segmentation to work out which items to put next to which others on shelves.
- Pricing departments use segmentation to identify how to maximize the amount that every customer pays (e.g., offering discount airfares at inconvenient times to price-conscious flyers).
- Sales departments use segmentation to set sales territories.
- Data miners use segmentation as a way of summarizing vast quantities of data.
- Management consultants use segmentation as a way of better aligning organizations with customers.
Customer segmentation vs market segmentation
Customer segmentation is the same basic idea as market segmentation, except that the scope is limited to a company's actual customers. That is, people who are not customers are excluded from segmentation. Typically, customer segmentations are based on data held in a company’s internal databases (e.g., CRM and transaction databases).
A segmentation study typically refers to a survey that is used to collect data, where the data is analyzed via a segmentation algorithm to identify segments in the market.
Market segmentation strategy
Segmentation strategy refers both to the decision regarding which of multiple possible segmentations to use, and decisions regarding how to allocate resources to segments. This includes determining which segmentation to adopt, how to allocate resources, and how to tailor offerings, messaging, and operations to maximize value from each segment.
A strong segmentation strategy involves several key components:
- Choosing the right segmentation means selecting a model that aligns with your business goals and balances detail with actionability.
- Prioritizing segments involves focusing on the most valuable, accessible, and strategically important groups.
- Activating the strategy requires tailoring messaging, products, and channels to each segment, and continuously monitoring performance.
Segmentation algorithm
A segmentation algorithm is an algorithm that is designed for aggregating observations (e.g., people) into groups, where the groups are typically referred to as segments, clusters, classes, or taxa. The most widely used segmentation algorithms are k-means cluster analysis and latent class analysis.
These algorithms analyze patterns in the data to ensure individuals within a segment are more similar to each other than to those in other segments. The choice of algorithm depends on the type of data, the number of variables, and the desired interpretability of the resulting segments.
Download your free DIY Market Segmentation ebook
Ensuring successful segmentation
A technically sound segmentation isn’t necessarily a successful one. Just because the math checks out doesn’t mean the results will be useful to decision-makers. Ultimately, a good segmentation is one that’s managerially useful—something that guides real-world strategy, inspires action, and makes intuitive sense to stakeholders.
So how do you get there?
The secret is to generate and compare lots of different segmentation solutions, rather than relying on a single output from a favored algorithm. This might include:
- Trying different numbers of segments
- Changing the input variables or standardizing them differently
- Using dimension reduction techniques
- Increasing the number of random starts
- Testing alternative algorithms (e.g., k-means vs. latent class)
- Adjusting assumptions, especially if using latent class analysis with numeric data
Once you’ve created multiple candidate solutions, evaluate them systematically. Key questions to consider include:
- Do the segments relate meaningfully to other business data?
- Are fewer segments just as effective?
- Are the segments easy to understand and name?
- Do they inspire action or feel like marketing personas?
- Could the differences be due to response biases?
- Are the results replicable?
This evaluation process should be structured and take time—usually a few days. A helpful approach is to create a comparison table summarizing the pros and cons of each solution. The goal is to find the segmentation that best balances technical quality with practical value.
Learn more about Market Segmentation
To learn more, download our free guide to Market Segmentation!
