Market segmentation is a technique used by marketers to separate a market into more manageable and insightful sub-markets. Used in product development and communication development, media strategy, and in both user and customer experience design and management, the value of market segmentation to a business is its ability to improve decision making about consumers that improve investment returns to a business.
In an article in the Harvard Business Review, a study by Bain & Company, found that businesses who properly segment their market achieved success. In the study, 81% of the interviewed executives felt their segmentation was a critical tool for growing profits, but fewer than 25% believed their companies used it effectively. This does beg the question on what is a ‘proper’ segmentation? In this practical article, we step you through how to run a market segmentation project so that you get the best results. In this article, we’ll cover the following areas.
- What is market segmentation?
- Why segment your market?
- What are the ways you can segment your market?
- What are the market segmentation steps?
- How do you know that you have a good market segmentation?
What is market segmentation?
Market segmentation is a technique used to separate a market into smaller groups whose members are more similar on a set of criteria than members allocated to another segment. At its simplest level, market segmentation can include splitting your market into customers and non-customers, or retail and commercial buyers. Even splitting your customers by sex, age or location is a form of segmentation. At a more complicated level, it involves using a large number of observations, transaction and interview collected variables. Market segmentation can also encompass an entire market, or segment only part of a market. For example, businesses often segment their customer database to identify different types of usage, revenue, cost and profitability segments.
Market segmentation does not need to be complicated; it just needs to be useful. To be useful, your market segmentation needs to relate to how you will use the results and how well the segments reflect your market. Many market segmentations, and personas, often fail because they reflect personal opinions on how we want the market to look, rather than reflect reality. Unless they are grounded in data, and a robust understanding of human behaviour as it relates to a business’s activity, segmentation and persona’s are likely to reflect personal biases, prejudices and stereotypes.
“Market segmentation does not need to be complicated; it just needs to be useful.”
Unlike some research which treats consumers as independent of a business, market segmentation needs to take into account how a business is structured and how you will use the market segmentation in your decision making. It is because of the need to reflect your business, that market segmentation is often as a precursor to developing a marketing strategy, and is used as input into other market research. The need to reflect how a business will use its segmentation, also means that businesses in the same industry may have very different market segmentations, and that a business may even have multiple segmentations to reflect different needs. For example, for a consumer goods client, we had different segmentations that were used for communication, products development, and trade channels. Although there was some cross-over between the segments, the different market segmentations used different criteria that were relevant to their goals.
Market segmentation is often used in business-to-business marketing for the same reasons as consumer-based businesses; to understand how a market is structured. For a financial services organisation that served small banks, credit unions, and similar types of banking organisations, we segmented a national market using their annual report results. By using this data, we were able to understand the different strategies being used by these businesses and how those strategies were then reflected in their financial performance. Market segmentation allowed them to improve their product and service offering, and their industry forecasts.
Why segment your market?
Market segmentation is a powerful way to improve your businesses understanding of the market. By segmenting your market, you have a better understanding of how similar behaviour, liking buying your product, are driven by different factors and how those factors vary across your market. For a segmentation project to be successful, it needs to be relevant to your business. The more specifically you understand what your need is, the more likely your segmentation will achieve that objective. Universal segmentations often provide generic segments that offer little real business-relevant insight.
Even if you use a mass marketing approach, segmenting your market can improve strategy development. By undertaking a market segmentation, you are becoming more customer focussed in trying to understand why consumers buy or not buy your products. The process of trying to understand how consumer differences drive similar behaviour, can help you develop communication that appeals to the broadest part of your market or develop products with the greatest appeal.
“Even if you use a mass marketing approach, segmenting your market provides can improve strategy development.”
The table below shows some of the main reasons business undertake a market segmentation project. A business may want to use segmentation for multiple areas, and a segmentation solution may assist in more than one area. However, the more disparate your goals are, the less likely your segmentation is will provide a useful tool for anyone area. While having a universal market segmentation that predicts behaviour across multiple areas would be ideal; this is rare because what drives one type of behaviour rarely drives all other behaviours.
What are the ways you can segment your market?
The number of variables that you can use to segment your market is almost infinite. The number of ways that you can segment a market falls into six main types: Geographic, demographic, behavioural, psychographic, needs-based, and business relationship. When doing your segmentation, you can use more than one type of approach. Sometimes when combining variables from each of the approaches, you may find there is a close relationship between variables that provide new insights.
For example, in an energy market study where we segmented the entire Australian population, our initial focus was on creating a psychographic segmentation. However, we also included demographics, behaviour, and needs. When analysing the results, we found that location and household characteristics, provided the main segmentation factors, and psychographic details provided an additional way to help with communication development. However, without the location and behavioural information, the segments provided no benefit in understanding product or service use.
Below are the main types of ways that a business can segment its market. The first two ways – geographic and demographic – describe a consumer’s context and while they affect consumer behaviour, they are independent of your category. The next two types relate to how consumers engage with your category; what they do (behaviour) and how they think about it (psychographics). Some segmentation projects include broad values, personality and other psychological characteristics that, like demographics, are independent of your category. The last two types relate specifically to how consumers interact with the category. In some studies, the business relationship is part of the project scope. For example, segmenting consumers at the pre-purchase stage, buying or account opening stage, the usage or post-usage stage.
For business segmentation, the equivalent of demographics is called firmographics. Firmographics are the characteristics of an organisation that can cover different measures of business size or operations. This area can also include financial information.
“How is this characteristic likely to affect behaviour?”
When choosing a framework and specific areas to measure, you need to keep asking yourself “how is this characteristic likely to affect behaviour?”. Although you may not know exactly how a characteristic affects a behaviour, you should have some understanding of how it could potentially be relevant. For a market segmentation study of smokers to quit smoking product, physical and mental health related questions were included as demographic variables in the project. Although we did not know how specific conditions were relevant to behaviour, prior research indicated mental and physical health could play a role in what types of quit smoking products were used and their approach to quitting. Both factors ended up playing a major role in understanding the different behaviour and values in each segment.
When trying to work out which types of questions to include in your market segmentation, also keep thinking back to how you are going to use the segmentation. There are many ways you can describe different types of consumers, but only some will be relevant to your needs.
What are the market segmentation steps?
To get the best from any market segmentation, you should think of it in terms of a business project rather than only as a market research study that delivers your segments. The strategic nature of a segmentation project means that the best results are tailored to a businesses needs, and that while the segmentation project itself will provide your business with market insight, the real benefit comes from how useful it is with your strategy development. For a business to realise these benefits, creating the segments is an iterative process.
In this section, I’ll step through the 11 steps for undertaking a market segmentation project process. These steps follow similar steps to undertake a market research project, with the addition of undertaking preliminary research and having an implementation phase. When setting up your segmentation project, the research briefing template will also help you save time and increases the chances of success.
Step 1 is the most critical stage. Like any research project, the better you define your needs, the more likely your project will be a success, and the easier your project will be to undertake. Because segmentation results are used as input into other projects, it is particularly important to get this stage right. Next, decide which market you want to segment. The first stage may partially address which market you will segment; however, you will still need to define the market to ensure you to research the right people. For customer segmentation, this stage will include defining the market based on database selection criteria.
Undertaking preliminary research is always advisable. By doing some initial research that involves your target market, you will be able to test some of the criteria to see how you should measure them, and if they likely to be predictive. Other research you have conducted for other projects, may help provide the initial areas that you need to include. From this preliminary research, you will have a clear idea of what types of variables to include in the project.
Now collect your data! But wait! For large, survey-based, segmentation studies it is a good idea to stop and analyse your data after completing the first 25% or have n= 300-400 completed interviews. The early analysis will show whether your variables are showing adequate variation and the likelihood that your approach will provide a workable segmentation solution. Once you have completed all the interviews or collected all your data, it is time to run your segmentation.
“In your qualitative phase, you may have created some preliminary segments. Be prepared to abandon these once you have your final segments.”
There are several types of techniques used to create segments. Some of the more common approaches are Cluster Analysis, Latent Class Analysis (LCA), and Decision or Regression Trees. For small studies, and ones using a qualitative approach, segmentation can take a more manual approach. If a qualitative study is done before a larger scale survey, sometimes a qualitative segmentation is completed to provide draft segments. If you do create draft segments in your qualitative research, be prepared to kill your darlings once you have your final segments. Your team may become wedded to the qualitative solution and try and force the larger study to come up with your initial solution, which may not exist.
“If you do create draft segments in your qualitative research, be prepared to kill your darlings once you have your final segments.”
Once you have your segments; refine them. See if they can be grouped to form larger segments, split into more distinctive segments, or if you can create them from fewer variables. As part of the refining process, you will need to profile each segment against both the variables that were used to create the segments and other business-relevant information. Segment profiling tells you who your segment is. By profiling the segments, you will know distinctive each segment is from other segments and how similar members within each segment are with each other. Part of the profiling stage is to uncover which segments provide your business with the best strategic direction.
The profiling stage will also indicate which are core variables that define each segment and discriminate between segments. If you are going to use your market segmentation to classify consumers, or to use in recruiting consumers for future studies, having a small set of predictive variables is critical for implementing your segmentation into the business. Techniques like discriminate analysis and logistic regression that use profiling variable to predict membership, are helpful at this stage.
“A well-designed segmentation solution will bring your consumers to life.”
Finally, you need to apply your market segmentation solution to the business and sell the solution into the business. A well-designed segmentation solution will bring your consumers to life in a way that helps others in your business to become more customer-centric in their decision making.
How do you know that you have a good market segmentation?
Because you can segment a market many different ways, during the development and the refining stages of your segmentation project, you will need to check whether you have good your market segmentation solution. An ineffective solution is no better than randomly assigning people to different groups. When creating a segmentation, you should also create competing solutions. The more solutions you have, the more likely that the one you choose will be the best for your business. To help determine which solution to use and whether it is likely to deliver benefits to your business are seven evaluation criteria.
The first two criteria are measures of whether you have segments. Unless the consumers, or businesses, within a segment are more similar to each other than those in other segments, then they are not a segment. Not only should consumers in a segment be similar to each other, but they should also be different from members of other segments. Although this may seem like something that is automatically true, if you are using multiple variables to create your segments, you can end up with segments that are subsegments of each other.
“If all your segments respond the same way to the same offer, then your segmentation is not better than no segmentation.”
For a market segmentation to have practical value, it must also be relevant to your business objectives. Depending on your segmentation approach, this relevance may be found in one of the statistics used to generate the segments, or from profiling the segments against key measures. Relating to their relevance is how actionable a segmentation solution is in regards to different offers, communication or experiences elicit different responses. If all your segments respond the same way to the same offer, then your segmentation is not better than no segmentation.
When selecting a target segment, be careful that it is not mutually incompatible with large segments. If marketing activity that turns on one segments turns off another, then you will need to ensure your strategy takes this risk into account. When developing cold sore communication, a decision was made to target a specific segment that was primarily young females. Because fewer males and older consumers were not part of the segment, and the segment had a message that focussed on triggers and benefits of less relevance to the other segments, most of the market was left feeling the brand was irrelevant and started to buy the competitors offer. Depending on your study design, the profiling and testing of your segments may be part of the original study or as a follow-up validation phase.
The next two criteria for judging the success of your segmentation relate to how you will use the segments and their practical value to the business. Assuming your segments respond differently to different strategies or offers, the target market segment needs to be large enough to generate a profitable ROI. Unless the target segments are large enough by themselves, it is tempting to combine segments to achieve a profitable segment. Unless those segments are similar in some important business-relevant way, combing the segments will render your segmentation useless.
Linked to market sizing is your ability to identify the segments either in your database or in the broader market. If the criteria you used to uncover your segments are not in your database, then you will need to either collect that information, use proxy variables, or only use your segmentation as a conceptual tool. When doing market segmentation of an entire market, or if you do not match customers to segments in your database, you will still need to create a refined decision rule that you can use for media planning, creative briefing, and recruiting consumers to research. A common mistake is having market segmentation that use a large number of variables and complex calculations to identify segment membership. This mistake can greatly increase research costs.
Finally, your market segmentation should provide insights into what drives the differences in your market. If the segmentation variables themselves do not clearly explain the differences, then you may need a follow-up stage of research to understand why these variables predict market differences.
Unlike many other forms of market research, market segmentation is a dynamic process or marching the needs of a business to how best to segment the market to reflect that need. For this reason, market segmentation is not an exact science. While there is a lot of personal judgement in their development, segmentation is not an art form. A segmentation that moves beyond personal biases and market stereotypes is based on evidence. This evidence, combined with a deep understanding of your market and business, will produce a tool for creating a successful strategy.
If you want to know if market segmentation would deliver your business a competitive advantage, contact us to discuss at firstname.lastname@example.org.