Getting to the Root of Innovation Part 1 of 2

A Wider View of Segmentation

Throughout the fifty-plus year history of market research, we have attempted to understand the marketplace by breaking it down into manageable pieces. Traditionally, this has been via segmentation, with either simplistic cross-tabulation or more complex multivariate forms. However, this goes only part of the way to providing a total understanding, and perhaps stems from the fact that we have too narrow a view and definition of what segmentation is and what it can achieve.

Much of the emphasis we place on segmentation as researchers is in terms of the consumer. There is clearly nothing wrong with this. However, there is another perspective that merits our segmentation skills: how consumers themselves segment the products and brands available to them--which is often quite different from the marketers' perspective--which we might term a brand landscape segmentation. After all, the Latin root of segmentation is segmentum, to cut, and segment is a component part, any one of the parts or sections into which an object or group is divided. And we shouldn't be myopic to the opportunities to break away from traditional segmentation approaches.

This paper is therefore about segmentation, but from a broader perspective than the traditional consumer-oriented taxonomy, which is nevertheless an important component. How consumers see the brand landscape in which they search, shop for, and consume products is the starting point. An understanding of the brandscape--how to determine, analyze, and interpret it--provides not only a total understanding of how attitudes, needs, and beliefs combine to drive behaviors and how these play out on the brand terrain, but also where white space--innovation-driven opportunity--can be identified.

What is White Space?

Technically, of course, white space is a misnomer, in that the space defined is only white if the color of the background is white. Blank space might be more appropriate, but that's perhaps not as attractive a concept.

White space also infers gaps; in years past, this might have been called "gap analysis," although this was performed on a set of numbers rather than a set of spatial inferences. White space is a visual representation of gestalt and the principle of proximity. Quite literally, without white space, objectives would be close together, adjacent, essentially the same, and substitutable. Because objects obviously create occupied space (we might call it gray space), it can be described essentially by what comprises it (in this case, the brands or SKUs). However, where we see white space could indicate we have a lack of knowledge, a lack of opportunity, or simply a lack of something to say about a product or specific brand in order to fill and exploit the void. White space might be evident simply because we know what is around it, but it is unknown in the sense that we are not fully clear as to what it offers, if anything.

But does white space really exist in markets, and if so, where? Some might say it is purely imaginary, a marketer's creation and yet another passing buzz word. It is important enough for Wrigley to make experience with white space analysis a job requirement for a global marketing manager. Finding white space may be as difficult as finding the appropriate candidate, since the brandscape is becoming increasingly crowded. According to ProductScan, there were some 33,000 new product introductions last year, a 50% increase over a decade ago. And Booz Allen Hamilton reports that for every seven concepts that enter the new product process, only one becomes a commercial success, and 46% of resources devoted to new products are spent on products that fail or are cancelled. Add to that a 2001 estimate by the TV documentary series Frontline that we receive about 3,600 brand exposures a day, and it is hard to believe that there is any room to move.

White Space and the Innovation Process

It is fairly widely agreed that there is less and less differentiation among brands, both in terms of functional benefits and perhaps even emotional ones. Therefore, from a manufacturer perspective, while white space may exist within the core brand portfolio (sometimes called Horizon 1) or even the contiguous area (Horizon 2), which are more likely to contain products and brands that are similar, interchangeable, and adjacent and offer low to medium incrementality, it is in Horizon 3--new areas--that the biggest opportunity and most incrementality will probably be found.

However, as Steve Coley notes in The Alchemy of Growth, "What distinguishes the corporations that carry on growing is...they can innovate in their core businesses and build new ones at the same time."

This is not to say that profitable white space is only in new markets and categories. It can certainly emerge and be identified in a mature marketing landscape, as consumer needs change and brands mature and the whole landscape begins to transform or is redefined; look at the dynamic change in the coffee market by the white space that was exploited by Starbucks. In order for a company to exploit "new" white space, it will need to make strategic moves, invariably requiring significant capability development, commitment of time and investment, and possibly new brand creation, or at least a line extension. Therefore, capitalizing on white space is unlikely to be a tactical quick-fix initiative.

White space and the identification process that accompanies it should not be viewed in a vacuum, in the sense that discovery of white space is a call to find a product to fill it. Rather, its identification is just one component at the beginning of the early-stage innovation process. White space identification is a by-product of market structure work, and is often undertaken in conjunction with innovation platform development. Such studies and resulting segmentation are also part of a regular process to monitor customer and market change and to act as a roadmap for broader strategic development and brand planning. On this basis, landscape structure studies should be considered on a three-year cycle for relatively stable markets and perhaps more frequently for markets experiencing dynamic change and an appreciable number of new entrants.

White Space in the Landscape and Market Structure

White space identification is part of a holistic approach to marketplace understanding. Because white space, at the point it is discovered, is essentially nothing, empty, colorless, it can only be determined by what is around it and what it is not. As a short research exercise, we could easily undertake only the core similarity sort task that is the basis of building the brand terrain. White space would be identified, along with current brand clusters, but that would be about it. In order to gain true perspective, a significant amount of additional data is required with respect to attitudes, needs, and beliefs, both general as well as category and brand specific.

This facilitates the overlay of such information on the basic landscape, subsequent drilling down, and further segmentation in the more familiar multivariate analysis sense to make the whole initiative actionable. Put in this broader context, the objective is to unlock the code and maximize product and brand differentiation and to refine consumer targeting.

For decades, marketers and researchers have cycled through the gamut of potential keys to better understand the relationship between the consumer and the markets in which they operate to better identify, define, and capitalize on the dynamics. Psychographics, values, lifestyles, life stages, sociodemographics, geodemographics, attitudes, and behaviors have all played parts in a variety of segmentation schemes. What is perhaps different about first identifying the brand landscape is that this segmentation is conducted through the lens of the consumer.

In data collection terms, this is a fairly straightforward task, requiring consumers only sort through brand images such as pack shots or brand logos in terms of their similarity. There are no preconditioning attributes that can potentially lead or constrain thinking. A well-executed landscaping exercise is merely a brain dump of how each individual consumer sees the market as defined by the brands in the task. It is their "head set": their organizational chart for classifying the brands and their personal consideration set at the "moment of truth,"; when planning, shopping, or making a selection for the next meal, for cleaning the sink, or for applying in the bathroom. It is their segmentation of the available choices.

Issue Development

Large-scale market structure and landscape studies should not be undertaken without significant due diligence in terms of issue development. The worst thing to do is to respond to an initiative prompted by "we need to do a segmentation study." To build any study purely around the tools you might use to generate solutions is like saying "I need to use a saw," when what you really want is to make a table. It is important to appreciate that landscape studies are not merely delivering data, but the very act of applying multivariate analyses induces holistic information delivery that deepens understanding, adds value, and becomes a more prescriptive, forward-looking tool. It has to in order to justify its longevity. Therefore, objectives need to be carefully set out, not be couched in tactical terms too much, and consider both likely strategic company and market direction over several years. Prior to starting such an engagement--which is likely to total three or four months from beginning to end--all current knowledge, hypotheses, and expectations from research, brand, and marketing teams need to be brought together and fully expounded. Although the immediate needs might be for a single brand, it is rare that a structure study does not impinge on the real estate of several other brands in a company's overall portfolio, let alone new product development. Therefore, polling all possible constituents is also advisable.

Critically, it is important to define the playing field. The danger with such strategic initiatives is that they become too myopic, too limited in their scope. You may currently market cola, but your competition extends not merely to other colas and sodas, but also to juices, power drinks, water, coffee, tea, beer, perhaps even ice cream, depending on occasion, function, location, and a host of other factors. Simple attitude and usage surveys are likely to remain narrowly focused, traditional segmentation perhaps less so, but landscape structure studies, especially when exploring white space, need to adopt as wide a perspective as possible. This is because white space is more likely to be meaningfully defined by what is around it. White space on the periphery is harder to qualify because it may be influenced by products or brands that represent further-out competition and were not included in the evaluation.

Although structure studies are ultimately large-scale quantifications, input from exploratory qualitative research is often beneficial in determining whether all consumer issues are being covered and appropriate language in terms of underlying attitudes, functional, and emotional benefits, key drivers, triggers and barriers, and aspects of brand image and equity are identified. The objective is to attempt to get beyond the obvious and look for more subtle and perhaps more predictive underlying motivations.

Building a Landscape Structure Study

Two principal factors control the nature and extent of market landscape structuring exercises: breadth and depth--reasonably large samples and a thorough information drill-down. Typically, a brand landscape will generate perhaps eight to 12 brandscape clusters and a consumer segmentation, four to eight segments to be actionable. So sample sizing has to be sufficient to accommodate this. Although capable of online deployment for small-scale initiatives, we favor in-person data collection, as experience shows that the similarity sorting task is less constrained with greater sorting and resorting in a real (as opposed to a virtual) scenario. However, as CLT facilities add broadband capabilities, a combination of offline and online data collection is becoming feasible.

To profile the market landscape fully, you need all the interpretive help you can get; remembering that white space has no direct data. While there are some basic templates for building these types of studies, it is important to recognize that specific goals are likely to vary by project. The need for a lot of data is often because the first pass of a segmentation may not necessarily yield fruitful solutions due to the heterogeneity of respondents. Increased learning from segmentation analysis and the application of more complex and inclusive segmentation analysis approaches such as Latent Class and Two-Step can overcome this. However, no matter how good the analytic engine, it cannot utilize data that is not present; it is not completely true, because Geographic Information System GIS data and mapping can add data post-hoc, providing that individuals can be geo-coded (i.e., at minimum their zip code can be established). GIS data ranging from hard "categorical" data such as residence in a newspaper circulation area, distance from a named store, level of water hardness, etc., or "soft" probabilistic data such as likely of class of automobile ownership, consumer expenditure level, etc., can be appended. Suddenly the power of segmentation has been kicked up a notch, and white space has possibly become just a shade grayer.

Regardless of the specific issues involved, there are nevertheless some key building blocks that need to be incorporated in order to map the basic brandscape, and understand the core structure and market dynamics. These include awareness, usage and evoked set or repertoire derivation, key category drivers/benefits, determination of barriers, occasion need states, brand/product benefit image ratings (both functional and emotional), user personality, category and lifestyle/lifestage attitudes, and purchase and channel behavior.

There is one other important element that assists in the identification and segmentation of white space, and that is the inclusion of new product ideas into the task. The insertion of 10-15 simple concepts into an essentially strategic study would generally be considered an anathema, and it certainly needs to be recognized that this is not a forum for concept testing. However, it is useful for exploring market opportunity to see where new ideas might extend a company's products or brands, and especially to see if they fall into white space, thereby providing further definition of the space itself.

Read Part 2