It’s been four weeks since you presented your market segmentation to the executive team. Everyone at the meeting seemed really excited about how your team will use the segmentation to deliver a step-change in sales. After fighting for funding to get the project off the ground, you just need just an incremental increase to your marketing budget for implementation. Despite reassuring comments from the executive team, new concerns keep coming up in meetings, and now there is silence. What went wrong?
Even for businesses that operate as mass marketers within a mature market, market segmentation can play a critical role in a business. Market segmentation provides businesses with a way to understand their market, prioritise which markets to target, it provides marketers with the needed information for developing their campaigns, and it helps design teams focus in on what consumers want and why. In a study reported on in the Harvard Business Review, 81% of business felt their market segmentation was critical in delivering business growth. Despite this apparent endorsement of segmentation, only 25% of executives felt they used them effectively. Unless you plan to avoid the common pitfalls of market segmentation, you are planning to disappoint.
“Unless you plan to avoid the common pitfalls of market segmentation, you are planning to disappoint.”
So why do so many market segmentations not meet expectations?
As a strategic piece of market research, market segmentation projects come with high expectations. After the initial excitement of presenting a segmentation solution to senior management, the hard work for marketing begins. If you have sold-in the market segmentation as a solution for solving all your marketing needs, then senior management, and others in the business, will have unrealistic expectations. This was a lesson learnt by a utility client that segmented their market using only personality and lifestyle values. After selling the segmentation into the business as a solution with universal consumer insights, managers in operational areas wanted to use the segmentation for customer experience and channel use. While the segmentation provided great insights for creative development and media planning, it was never designed to explain why customers chose different service delivery channels, or predict service needs. Once some managers felt disappointed, the disappointment rot set in, and the entire segmentation’s value was questioned.
To help marketer with their segmentation, back in 1967 Philip Kotler, one of the founding fathers of modern marketing, put forward an initial list of five criteria for judging your market segmentation (Relevance, Size, Access, Measurable, Reachable). These criteria provide useful guidance once you have a solution, but it misses key reasons why segmentations often fail to meet internal expectations and are not implemented. Based on both my experience with designing and in implementing market segmentation projects across consumer and business markets, I’ve found market segmentations are more likely to fail when they fail in one or more of five areas. These areas incorporate the initial list proposed by Kotler with more operational and design areas.
Failure starts early. The most critical mistakes in market segmentation begin with segmenting the wrong market or letting personal bias and agendas drive the results. Businesses always have multiple markets. Even at the most basic level, from a specific business’s perspective markets are generally made up of consumer/ non-consumer, commercial/ retail, trade channel/ direct, buyer/influencer channel, and customer/ non-customer markets. For segmentations that are based on customer databases, there is often the added complexity of customer selection definitions.
A basic error is segmenting the wrong market. If your business objective is to grow the business, then you will need to focus your market segmentation on your non-customers. If the business objective is to increase conversion and your business uses a third-party channel, then your segmentation will need to either be based on decision influencers or take them into account in your design. The market you need to segment may appear obvious; however, time spent making sure you are segmenting the right market increases the chances of success. A healthcare client that focussed their segmentation on the general market became disappointed when many of the strategic issues focussed on current users, who made up only a small part of the general population. Product usage occurred only a few times a year and not every year.
When choosing what market to segment, you will also need to take into account your businesses resources. For a business that has a customer database, segmenting your customer base may not be optimal, but it may be affordable and offer a platform for later your broader market. For businesses that are unsure about implementing a full market segmentation, customer segmentation may offer a cheaper way to prove the value of a market segmentation strategy.
The second main design issue is personal bias and its more insidious form, prejudice. Personal biases can that shape every aspect of the market segmentation project, but in the initial design phase, it is particularly dangerous. Personal biases and prejudices affect market choice, what measures are selected, and what segments are selected. If you start with the wrong assumptions, your solution will be wrong. A client, and leading consumer grocery brand, had used qualitative research for their market segmentation. When they designed the study interviews among frequent users of their brand was done in wealthier suburbs around where the marketing team lived, and among younger more affluent consumers. In contrast, users of competitors products were undertaken in less well-off areas, and among older less well-off consumers. Despite the bias being pointed out and how it had an impact on their market segmentation, the self-reinforcing influence continued to affect creative development and media planning for several years. What makes personal bias dangerous is that to the person with the bias, they seem perfectly natural and logical. Market segmentation should be built on evidence, not personal worldviews.
“If you start with the wrong assumptions, your solution will be wrong.”
Business Alignment Issues
Unless your segmentation reflects your business or the characteristics of the market you compete within, there is little chance that your market segmentation will be useful. Interesting, yes; useful, no. This is because you can group people into an almost infinite number of ways. Unless you include variables that are linked to how you will use the results or aspects of category decision-making or behaviour, any segmentation isn’t going to explain your market. For example, I once worked on a segmentation project for a utility client that was created by a different consultant that was based on personal values. This values approach was popular in the 1980s and still makes occasional reappearances under different brand names. In the profiling stage of the market segmentation, we could find no consistent link between product choice, usage, or how the business could select any segment for communication development.
An excellent way to check business alignment is to ask yourself, which variables used the segmentation describe how consumers engage with your business or category?
Putting people, or businesses, into groups does not make them segments. Because people and businesses can be grouped many ways, many segmentation solutions are no better than random assignment. How you test that you have an actual segmentation is by understanding how similar members in a segment are to each other, and how different segments are from each other. Statistical techniques like k-means clustering and latent class analysis segmentation are designed to find solutions that maximise within segment similarity while also maximising between segment differences. However, to find the best solution, you will need to profile segments by variables used to create the segments. By adding and removing segmentation variables into your analysis, you increase the chances of finding the optimal solution.
If senior management cannot easily see what defines a segment and how they are different, then they won’t want to use the segmentation. If your segments are clearly defined and distinctive to each other, keep the descriptions short and focus on these two things.
If a market segmentation cannot deliver value to the business, then the business will not value your market segmentation. From a practical perspective, your market segmentation needs to deliver a return on investment (ROI). To return a positive ROI, your segments will need to be large enough to warrant developing targeted campaigns, offers or products. Having large segments comes at the cost of reducing internal similarity and between segment differences. To increase your segment sizes, it is tempting to combine multiple segments. Unless these segments are part of a larger segment, then combining segments will undermine the whole segmentation strategy.
When choosing your final market segment solution, know how big you need a target segment to be for it to justify the investment. Knowing this can help with setting the number of segment solution. For example, if the smallest size that your business could profitably target is 25% of the total category size, then you should look for a segment solution with five to seven segments.
To achieve a positive ROI from your market segmentation, each segment should respond better to different offers. That is, if the segmentation reflects different mindsets, then these differences should also be reflected in what they find attractive. In practice, we find several segments will still find the same offer or communication appealing but to a different degree. Some segments will find an offer very appealing, while others will show marginal interest. From a campaign design perspective, this has the benefit of running campaigns that do not alienate some segments.
Great ideas that cannot be realised are just dreams. Market segmentations that cannot be implemented, won’t be implemented. Some of the best segmentation solutions I’ve worked on have come unstuck when we have tried to match the segmentation back to a customer database or use the segmentation to identify people in that segment. If you need to match your segmentation back to a database, then you should have either used database variables in your segmentation or have variables in your database that are proxy-variables for the ones you are using in our segmentation. You need to build your implementation strategy into your initial design; plan for implementation.
An issue that often occurs with market segmentations that use a large number of variables to create the segments, is that a large number of variables defines their segment solutions. For one project, we needed 60 measures to allocate people to each segment accurately. So that you can use your segmentation, you will need to reduce the number variables down to a smaller core group of questions. Using segmentation techniques like discriminate analysis, multinomial logistic regression or decision trees can help you reduce the number of variables down to a manageable set.
“You need to build your implementation strategy into your initial design; plan for implementation.”
A successful market segmentation strategy requires broad management buy-in. So at every stage, you need to engage all relevant managers from across the business in the project. Understand what they need and be clear about what your project can deliver. Also, create quick-wins. Expecting managers to support a significant change in how they do business entirely is a big risk. Make the risk small through pilot projects, or by using the segmentation in tactical projects that demonstrate its broader business potential. For an international market segmentation project for a multinational healthcare client, we first introduced the segmentation into the brand tracking a communication pre-testing to show how segments differed before using the segmentation as part of the communication and product development briefing process. During the staged implementation, the segments were refined to improve the relevance and ability to be implemented in other business areas.
Wrapping it Up
Market segmentation can be a powerful way to improve your marketing. Management is likely to have high expectations for the results and want to implement them across the business immediately. If you follow a good market segmentation and market research process and avoid the types of issues outlined in this article, then you will greatly increase your chances of delivering a successful market segmentation that delivers your business growth.