STI: LandScape - Methodology

What do political contributions, daily vegetable servings, and urbanicity have to do with consumers' propensity to purchase specific items or shop in particular areas? A lot, according to the companies using STI: Landscape to more precisely target not only who their customers are and where they live, but also their attitudinal attributes - including everything from their physical activity levels to their propensity to use the latest technologies.

The innovative methodology used to create this unique new market targeting approach significantly elevates consumer segmentation to a whole new level, far beyond today's traditional segmentation systems - giving companies distinct views into consumers' purchasing propensities that have never been available before.

Traditional geographic-based consumer lifestyle segmentation systems are tools that address the "who" and "where" of market targeting. Simply stated, these systems identify consumer groups that share a set of common demographic factors, and they identify where these consumer groups live. Their goal is to classify each neighborhood as a specific "market segment" in order to give decision makers demographic insight that improves decision-making in site selection and target marketing.

While there have been developments in the science of market segmentation since it was introduced in the 1970s, it has remained largely based on standardized socio-economic databases. However, as every market-focused company knows, business decisions are becoming increasingly difficult to make. There are a number of reasons for the growing difficulties, but generally the problems revolve around the expanding complexity and maturity of today's consumers. As a result, it's no longer possible to dependably and precisely pinpoint new markets using segmentation systems that have not evolved to meet today's complexity and challenges.

Now, what's needed is a sophisticated system that hones market segmentation to a much more precise level and consistently delivers better numbers than "traditional" systems. In other words, businesses need answers, not just to "who" and "where" their customers are, but also to "what" additional factors determine their purchasing decisions. Delivering this additional insight is exactly the goal of STI: Landscape.

To create STI: Landscape - and achieve this new level of market segmentation exactitude - we took a completely new path to market segmentation by augmenting traditional economic and demographic segmentation processes with several unique non-demographic consumer lifestyle indices. To create this new approach, we leveraged the wealth of non-traditional consumer databases that are available, such as health surveys conducted by the Centers for Disease Control and Prevention (CDC), political contribution as reported by the Federal Elections Committee (FEC), and much more. We also developed a unique methodology to create a truly innovative and powerful new "lifestyle with attitude" market segmentation system - one that hones consumer group targeting to the most insightful science ever available.

STI: Landscape's Development Philosophy

Although market segmentation has a 30-plus history, Synergos Technologies approached the industry from a non-traditional perspective. We are not demographers; we're data experts. As a result, we didn't realize that there's a traditional way to create market segments and "everyone does it that way." Without this historical perspective limiting our judgment, we took what we now know is a truly innovative approach to consumer segmentation by adding attitudinal attributes to the equation. To us, it seamed like the perfect way to bring a stronger personalized "human" element to traditional consumer data. The hypothesis paid off, because now our clients are enjoying much more accurate consumer targeting than they had ever achieved with traditional market segmentation tools.

In evaluating today's market segmentation systems, the most important shortcoming we discovered is the lack of empirical data on "consumer attitude" measurements. Aside from the basic socio-economic attributes that every household possesses, each household also has a set of attitudes that have a significant influence on its residents' buying habits.

A simple example would be to compare two families: both wealthy, with 2.5 children, living in large homes in highly urban areas, and having heads-of-the-households in their mid-40s. Demographically speaking, the two households appear to be the same. However, when we introduce new lifestyle attitudinal indices, we find out that their attitudes, and therefore their purchasing habits, are very different. For example, one family lives in an area that has consistently voted for conservative candidates and issues, while the other has consistently voted along liberal lines. As a result, one household buys Ralph Lauren Purple Label™ and the other Versace™ - both very different from an attitudinal point-of-view.

With this in mind, we proceeded to merge traditional demographic databases with non-demographic consumer data. However, we found that the number of segments created (about 500) would be too unwieldy to manage. So, we switched our strategy and started with a traditional set of segments at the block group level using a classic segmentation methodology. Then we appended each block group with attitudinal indices that further characterize the block group. This process allowed us to create a consumer segmentation system that captures the best of two world: A set of market segments that is manageable, coupled with a set of consumer characteristics that marketers can use to understand the attitudes of consumers and more accurately target their specific customers.

By creating several attitude indices, we are able to give companies the ability to select the consumer attitudinal insight that best defines their particular target consumers, markets, and products. This flexibility also allows companies to add or subtract new attitudinal indexes as their customers, markets, and products fluctuate and change.

Segmentation Methodology

The segmentation methodology for STI: Landscape includes two parts: (1) The process used to develop the primary segments, and (2) an overview of our current selection of 10 secondary (attitudinal) indices.

Part 1 - The Segments

STI: Landscape segments are created using a combination of two mathematical techniques: Factor Analysis, which is the process used to identify the primary factors that characterize a neighborhood, and Recursive Partitioning, which is the process used to refine those factors into smaller and smaller groups.

Factor Analysis. In most cases, only a handful of factors describe the majority of discrepancies between groups or events. Therefore, as we progress through our analysis we constantly evaluate which factors are the keys to describing and, more importantly, differentiating market segments. For example, we found the following demographic categories have the greatest impact on distinguishing the market segments used in STI: Landscape:

• Age
• Income
• Ethnicity
• Education
• Marital Status
• Dwelling Type
• Presence of Children

Once again these are the primary categories. In addition, our Factor Analysis allows for many other demographic characteristics to enter into the analysis.

Recursive Partitioning. In data analysis, Recursive Partitioning means to split a dataset into two or more subgroups to improve the homogeneity of each subgroup. The partitioning process recurs until a desired outcome is achieved: which, in our case, was when a reasonable size and number of market segments were created. Therefore, the STI: Landscape model was constructed by first identifying the factors that best subdivided the data into a set of groups. Then each subgroup was evaluated again with Factor Analysis to determine the best way to subdivide it, and so forth and so on. To insure that certain highly specialized sub-groups did not influence the factoring process, they were first removed from the equation (e.g., group quarters).

Part 2 - 10 Lifestyle (Attitude) Indices

STI: Landscape contains 10 Lifestyle Indices, which can be appended to our market segmentation system. Each index is mutually exclusive from the market segment. Therefore, when using an index it can be used alone or as a gradient to the market segment. For example, while two contiguous neighborhoods may be classified as Empire Builders, one neighborhood may be significantly more conservative than the other. While both neighborhoods may be classified as conservative, one may have a higher percentage of smokers. These significant differences impact their attitudinal influences and, therefore, spending habits.

The 10 Lifestyle indices are broken down into two categories: Social and Health. Social Indices relate to the human condition and the attitudes that form that condition. Health Indices relate to lifestyle decisions that ultimately affect health. (Note: Health indices can be highly correlated with one another.)

Lifestyle Social Indices

1. Conservatism. This is the degree to which individuals in a neighborhood support conservative issues in general elections. The data source is precinct-level election results. (Note: Some states do not use precincts to tally results, so this index may not be available for some states.)

2. Power Broker. This is a measure of financial contributions to political activities or social issues. Although, in most cases, Power Brokers tend to be affluent, the reverse is not necessarily true: that affluent people are always Power Brokers. The data source is political contributions reported to the Federal Elections Committee (FEC).

3. Innovators. These individuals are either at the forefront of innovation or are early-adopters of new technologies. This index was created based on a highly correlated relationship between innovation and certain occupations, educational backgrounds, and industry affiliations.

4. Gay. This index is a measure of the likelihood that people are associated with a homosexual lifestyle. Our index is based on several studies that correlate certain demographic characteristics with the probability of homosexuality. Since the extent of homosexuality in the U.S. is a much-debated figure, ranging from 2% to 15%, we opted to base our model on a relatively conservative estimate of 7%.

5. Urbanicity. Urbanicity is defined as a combination of the density of people and businesses in a specific location. It should not be confused with population density only: for example, a person living in a part of a city with no other people living nearby would live in low density, yet highly urban market. The measure is created by combining data from STI: PopStats and STI: Workplace.

Lifestyle Health Indices

6. General Health. This is an amalgamated score of multiple health indices that measure the general health of a population. The data source is an annual survey conducted by the Centers for Disease Control and Prevention (CDC).

7. Body Mass Index (BMI). The BMI measures the relationship of height-to-weight and is considered by most doctors to be the primary statistic for understanding health as it relates to obesity, diabetes, and many heart-related diseases. The data source is an annual survey conducted by the CDC.

8. Veggie. This index measures the number of servings of fruits and vegetables a person eats each day compared to what is considered adequate by most nutritionists and the United States Department of Agriculture (USDA). The data source is an annual survey conducted by the CDC.

9. Smoking. This index measures the incidences of smokers in a neighborhood. The data source is an annual survey conducted by the CDC.

10. Activity Level. This index is the level activity (or inactivity) in which the average individual participates daily relative to an ideal activity level. This measure accounts for both casual activities, like walking, and strenuous activities, like heavy lifting. The data source is an annual survey conducted by the CDC.

Neighborhood Personality Formation

In every market, there are distinct neighborhoods with specific "personalities," which are based on the combined characteristics of the people who live in them. Neighborhoods typically form for one of three reasons: tradition, environment, and development. Understanding how neighborhoods form can lead to a better understanding of lifestyle market segmentation.

• Tradition. Specific neighborhoods have traditionally attracted certain personality types, so the people who live in them tend to share similar lifestyles.
• Environment. Similar types of consumers are attracted to similar environmental factors, including cultural mores, school quality, work proximity, and natural features, such as water access and views.
• Development. When developers de-velop specific tracts of land, they have certain demographic characteristics in mind, such as "single professionals" or "married with children." Each group prefers a specific type of domicile, and the developers build homes according to the consumers' preferences and needs.