Custom Fit: The "Right" Segments for Your Business

The goal of segmentation is to produce differentiated groups of customers that look, think, and behave similarly. Having distinct segments in place allows for the business - marketing in particular - to focus on desirable relationships, customize the content of communications, and manage and track the overall health and growth of the customer portfolio.

Unfortunately, segmentation projects can sometimes leave marketers asking: "How do I know these are the right segments?"

In the simplest terms, the right segments are those that are not only statistically accurate, but also are the most appropriate for your business. Not simply a statistical grouping of customers, the segments must be reflective of your vision and organizational goals.

To build a successful segmentation solution, it is important to ensure that your team not only focuses on the data and statistics, but is also attentive to the pertinent customer behaviors for your business. The goal of your segmentation project should be to produce customer groups that are:

• Statistically Valid: There should be statistically significant differences between the customer segments. The solution should identify the relevant data and apply the appropriate statistical techniques.
• Distinctively Homogeneous: The within segment behaviors and needs must be similar enough to allow each group to be clearly defined.
• Relevantly Differentiating: There should be meaningful differences between the groups, not only in terms of their behaviors, but also in terms of how you will communicate with them.
• Differentiating on Future Expectations: While segmentations are not designed as predictive models, the resulting segments should be informative of future behavior. The differences captured between the segments should be "true" and repeatable over time - not simply a random variation in the sample.

Unfortunately, segmentation projects do not always meet these goals, which may leave marketers hesitant to integrate the insights and the segments into communication strategy. Generally, shortcomings of a segmentation project are the result of one of the following obstacles, each of which can have a dramatic impact on the outcome of a project.

1. Data Synthesis: When it comes to identifying the right segments, too much data can actually obscure important customer differences. To avoid this, there is an important step between gathering data and producing relevant insights…data synthesis. A segmentation project should allot the appropriate time to maximizing the available data. This is not necessarily done by incorporating every byte of data, but rather through the identification of the pertinent information and the manipulation of the data elements to best summarize and capture the relevant behaviors and insights.
   
   A segmentation solution can be as simple as the development of RFM tiers, or as complex as multi-dimensional summaries of customer behavior. The nature of your data and your business should determine the behaviors that are analyzed and how the data is synthesized to give you the maximum benefits. It is important to avoid throwing everything but the kitchen sink into an analysis simply because the data is available.
   
   
2. Statistical Methodology: Statistical procedures such as cluster analysis, regression models, and even business rules are routinely used to distinguish customer groups. All are acceptable alternatives for creating customer groups, but they each must be employed properly. In addition to being suitable for the data being analyzed, the chosen methodology should be appropriate for the goals of the project. The proper methodology may be as simple as a frequency distribution, or may require the latest in analytic techniques - the goals of the analysis should drive methodology selection.
   
   Additionally, it is important to realistically apply analytic procedures. Asking a methodology to do too much (i.e., create complex segmentations solutions in a single step) often produces inaccurate results or segments that are insufficient from a business standpoint.
   
   
3. Business Insight: The blending of business and science must be at the foundation of an insightful solution. The right customer groups cannot be created simply by identifying the most differentiating behaviors or statistically significant variables (please see Beyond the Numbers - Finding Insight in Statistics ). Rather, all differences must be considered in the context of your overall organization, including the different actions or measurement they may elicit.
   
   To create a successful segmentation solution, it is important to analyze the relevant customer behaviors: those that are insightful and actionable for your business. Customizing your segmentation plan to accommodate business and marketing goals will allow results to go beyond a summary of the data, into pro-active tools to manage your customer relationships. Pre-segmentation efforts to explore the relevant aspects of your relationship with your customers will have a dramatic impact on the overall direction and outcome of a segmentation project.
   
   

Clearly statistics and data expertise are integral for creating a customer segmentation, but equally important is the ability to extract and inject relevant business insights into the solution. While it is necessary to hire a good statistician to build your segmentation solution, that alone is not adequate to ensure you get the right segments. By identifying and planning for the three obstacles listed above, a company can undertake a segmentation project with confidence that the outcome will be meaningful and useful for their business.

The Art of Science
The process and methodologies developed and employed through SegmentologyTM ensure your solution meets the requirements of a good customer segmentation, and that you get the right segments. SegmentologyTM results in a solution that instills complete confidence throughout the organization, because it integrates your organization and marketing needs every step of the way. The final solution is right for your business, because it is built for your business. There are two ways that SegmentologyTM goes beyond science to ensure your segments will be the right snapshot of your customer portfolio:

• Customization of Dimensions
• Selection of the Segments

Customization of Dimensions
The SegmentologyTM process includes a set of pre-identified behaviors, or dimensions, found relevant across most vertical industries. However, your business is not forced into a one-size fits all solution. Rather, we work with you to determine the critical components of your organization's relationship with customers (for more information please see: Explore Your Customer Relationships Through SegmentologyTM ). Together, we will select the appropriate list of customer behaviors, and even add new dimensions when applicable. This will be the groundwork for getting the right segments.

Additionally, once the dimensions are selected, they are not created in isolation of the nuances of your business. The components and methodologies of each dimension are carefully considered and tailored to your unique challenges. For example, your definition of value may be based on margin, or it may simply be a measure of customer gross revenue. Maybe your customers demonstrate regular and consistent purchase cycles, or maybe your business is sporadic/impulsive in nature. Either way, the process of SegmentologyTM relies on the integration of your industry's unique attributes and challenges to ensure the dimensions are the right dimensions for your business.

Selection of the Right Segments
Once the right foundation has been laid for the creation of the segments selecting the final solution requires a balance between statistical significance, business relevance, and actionable penetration.

We do the hard part - identifying statistically valid solutions that are differentiating on meaningful characteristics. This process involves utilizing various combinations of the customer dimensions in a statistical procedure called Latent Class Modeling. By mixing and matching the dimensions, we can create multiple solutions that emphasize different characteristics.

For example, one solution may be heavily focused on customer value, allowing you to isolate and communicate with the most valuable of your customers. An alternative solution may be found to differentiate your customers' attitudes. While the second solution would still have high and low value groups, the predominant characteristic of each segment would be the customers' motivations. This would allow the identification of different types of valuable customers and the customization of messaging to multiple high-value groups.

When modeling the various combinations of dimensions, we evaluate dozens of solutions. This provides us with a "feel" for how your customers naturally group together - manifesting in recurring themes or common customer segments found in most of the solutions. We compare the pros and cons of multiple solutions side-by-side and determine the top 2-4 statistically valid and insightful solutions, keeping in mind your organization and future marketing goals.

We share and discuss these solutions with you, reviewing the similarities and differences, the strengths and weaknesses, and the relative importance of specific dimensions. Then together, we craft the final customer segments that are the most meaningful for your business. Since all of the solutions statistically valid, it enables you to choose the final segments that "feel" right for the organization.

Conclusion
Grouping customers into meaningful and actionable segments is a complex task that spans statistics, marketing, and organizational goals. Marketers should be aware of the obstacles inherent in a segmentation project and take steps to ensure the output will be right for the organization. Following some simple guidelines will allow marketers to embark upon a segmentation project with confidence.

SegmentologyTM provides a reliable option for marketers interested in customer segmentation. The processes and methodologies employed in SegmentologyTM ensure the requirements are met, and that you receive the right segments. This process, as we've experienced time and again, results in the absolute best of all possible scenarios, since you have an opportunity to guide a segmentation solution that will have the highest impact on your organization.