Closing the Gap: Linking Marketing Decisions with Consumer Decision Making

Alastair Gordon
Managing Director (Asia-Pacific)
R&D and Brand Health Management
ACNielsen, New Zealand

It is clear from ACNielsen brand equity research that many brands suffer from relatively low levels of brand differentiation and emotional commitment. At the same time, marketers are under increasing pressure to fund tactical marketing initiatives and find it harder to justify increases in “above the line” spending. This need to better differentiate brands and simultaneously concentrate more marketing dollars at retail is leading marketers to seek more information on exactly how to link information on buying behavior with data on perceptions and motivations.

Yet it is becoming increasingly apparent that there is a gulf between what we are able to discover about consumer's personal feelings, brand associations and psychological drivers and how a consumer actually behaves when confronted with purchase decisions. This clearly impacts our ability to offer our client's actionable and specific information about consumer behavior. It has also been noted that consumers are dealing with increasingly more information about more and more brands; predicting their behavior is becoming correspondingly more difficult.

While the attributes of awareness, trial and regular use are highly associated with brand equity, it is still the case that even brands with lower scores can attract reasonable rates of regular use. While high brand equity is surely correlated with better brand performance, it is also true that despite most brands having fairly low equity, they also have regular usage rates of around 20%.

Many studies have shown consumers with repertoires that go well beyond merely using their “favorite” brand. The rise of private-label brands in many markets should be warning enough to marketers that brand choice is not simply driven by overall preference or brand associations.

Neither is a discrepancy between preferred and chosen brand necessarily driven by price alone: ACNielsen data has shown local brands eroding the market share of leading multinationals, despite the latter's huge advantage in preference, improving levels of disposable income and a declining price gap between locals and foreign brands. Put simply: there are many people using brands they really are not that committed to.

Of course, overall, brand equity does distinguish between “stronger” and “weaker” brands. Yet brand consideration and use are only partially described by brand commitment measures. By modeling the drivers of brand equity, we may improve our insights into this phenomenon.

Brand Equity Modeling
A more formal method of evaluating impacts on brand equity involves analysis of “brand knowledge” variables against brand equity indices. This modeling allows us to develop an estimate of the relative contribution of brand awareness, consideration and brand associations to explaining variations in brand equity scores. In our analysis, approximately half of the consumer brand equity differentiation is related to basic familiarity issues rather than specifics of brand knowledge [See chart 1].

It is also interesting to note that brand associations tend to group into a rather small set of factors (e.g., brand imagery, product benefits, etc.). Typically, the first two identified factors account for two-thirds of the brand equity variance accounted for by associations. As a generalization, it is also fair to say that most of these “major” factors identified are fairly holistic assessments of “product quality” and/or general affirmations of brand reliability (trust, reputation, etc.). In other words, consumers tend to distinguish, at the level of brand equity, between brands based on a few rather general criteria rather than a mass of quite specific attributes.

What Does This Imply for Investigating Consumer Behavior?
Research in this area suggests that people base much of their decision-making on a fairly small set of rules called heuristics. The concept is that people do not waste huge amounts of brainpower to make most everyday decisions (especially in areas like consumer products). Instead, people tend to have a few fairly simple choice criteria that they use as their standard means of deciding.

For example, a consumer's rules in buying laundry
powder might be:

• It should work in a front-loading automatic machine.
• I prefer the brand I bought last time.
• If my last brand is not available, it should be well-known/popular.
• Don't buy the “budget pack” brand as my partner hates the smell.

The core features of such heuristics are that they: a) tend to be simple and easy to mentally compute; b) are relatively few in number; and c) are often good “surrogates” for other attributes/benefits (e.g., if it is popular, it is a good guess that it is also safe, reasonably effective, won't have an offensive smell and is “socially acceptable”).

There are good theoretical reasons why “simple familiarity” should play such a large part in consumer choice, and studies of consumer psychology point out that as consumers face increasingly complex choices, the role of such basic decision criteria is likely to increase. Put simply—people don't necessarily want too many choices, and when they employ an “elimination strategy” in the context of brand selection, it is likely that “familiarity” is a very efficient decision rule.

How Do People Actually Behave?
For years, market researchers have been exasperated by the consumer who insists they cannot answer the big “why”—the respondent who says, “I don't know why I do it, I just buy the same one every time.” Much effort has been put into trying to “get beneath” this “superficial” response. Yet heuristic-based analysis would suggest this response is neither superficial nor a simple rationalization of “hidden” drivers. Rather, “just buying the same” may be simply a mechanism for navigating through a very complex but relatively unimportant (in the context of a consumer's life) set of brand alternatives.

If this hypothesis is correct, then we would expect people
to choose brands based on:

• Broad rules to establish core acceptability (often “veto” or elimination rules)
• Very specific rules to provide easy guidelines (to quickly choose between various “acceptable” alternatives)

ACNielsen has recently developed a series of qualitative research tools to try and improve access to memories of how people actually arrive at choice “rules” and uncover the way these operate. Initial studies in Asia-Pacific countries are beginning to show a hierarchy of mental rules that seem to be applied in much choice [See chart 2].

The milk category example illustrates what seems to be a common set of phenomenon: people operate on two very broad types of heuristics: a “veto set” based on broad basic prejudices about “what works” in this category and a second, macro heuristic that tends to cause them to favor familiar brands unless given a very good reason for not doing so. The resultant choice set is then refined with fairly specific and often idiosyncratic rules based on personal experience and desires.

This is as would be predicted by heuristic theory. The notable addition to the model introduced by our findings is the degree to which, in many categories, social drivers are used heuristically to refine choice. Very often the choice is based on a simple trade-off; for example, a consumer may say, “I don't care much about this brand of coffee; however, my partner does, so I just go with their preference...”

One study of coffee brands revealed consumers defining choice on the basis of the exact size (not too large, not too small) of coffee granules. We would not necessarily expect “size of granules” to come out as vital in any quantitative study. They are indicators of other things (solubility)— gatekeepers to choice, not the underlying motivations
themselves.

Implications for Marketing
The following four themes are generally consistent across our research:

1. Autopilot purchase behavior—Many people choose brands automatically, as a matter of habit and with little thought.
2. Familiarity in initial differentiation—Top-level differentiation between brands happens on the basis of simple familiarity and broad, fairly basic assessments.
3. Social Influence—This plays a key role in mitigating individual choice patterns.
4. Trivial criteria before decision—The final choice comes down to “minor” and seemingly trivial decision criteria between similar alternatives.

All four of these patterns seem to offer explanations of why positive general brand imagery does not always seem to translate into purchase:

• Auto-pilot behavior may well be cognitively quite divorced from the current needs and motivations.
• Brand knowledge (and hence, by implication, strongly differentiated “feelings”) seems not to be essential in differentiating between brands.
• People seem to trade off their own needs against the needs of others fairly regularly.
• People regularly utilize quite minor but easy-to-implement decision criteria.

These trivial but “gatekeeper” criteria occur all the time—in studying paper towels for instance, we encountered several people who had a basic rule: “avoid any paper towel with animal patterns on it.”

All this makes sense only if we move our perspective on what underlies consumer decision-making. In the past, consumers have tended to be seen either as actively seeking a “best solution” in terms of choice, or as being almost “victims” of deep-seated needs, manipulated by underlying psychological drivers into making certain choices. In fact, it seems that consumers are not entirely the victims of needs they can't control, nor do they expend needless energy deciding between myriad alternatives. Instead, they are better seen as navigators, primarily concerned with making reasonably safe decisions as fast as possible. The process (efficiency of arriving at a decision) is thus as important as the end result (what I buy).

And if consumer decision-making is characterized by very broad general “appropriateness” and “veto” rules, which are refined by very specific “indicator rules” or tests, then market researchers may need to consider if the kinds of attributes and motivations we commonly describe are at an appropriate level on the decision-tree. In some cases they seem to be too specific to be broadly repertoire-defining and yet not discriminating enough to be helpful in understanding how people choose between brands which are “very close.”

Clearly, there is a need for more detailed studies in the area of how consumers use mental rules to refine brand choice decisions. Overall though, this paper represents a first attempt to show that the process of focusing on the mental rules that underpin habits can result, for marketers, in a more specific and actionable set of guidelines for marketing. If, as a profession, we are to be a catalyst for decision-
making, we need to better understand the process by which consumers make decisions. This implies market researchers need to build models which allow smaller, more precise set of consumer priorities to be identified, and to utilize category-specific analyses wherein consumers will be segmented primarily on the basis of easily identifiable and understandable cognitive decision processes rather than broad attitudes or motivations.

This article is an excerpted version of the paper presented at the 57th ESOMAR Congress and Exhibition, Lisbon, Portugal, September 19–22, 2004.

Decision-Making Segmentation:

Identifying Key Choice Mechanisms
Can we see evidence of habitual decision making and heuristic choice in quantitative studies, and if so, does it make sense to utilize this to categorize consumers?

Preliminary evidence suggests that it may well pay dividends. In a study of female chocolate bar purchasers in China, we attempted various segmentations based on personality and general brand association criteria, but they did not tell us too much about the key question in this category: why the main brand (with over 50% “most often” brand usage) was so dominant, and what was leading to experimentation when it did happen.

We focused on a series of “purchase trigger” statements asked about consumer's last purchase occasion. This revealed that, on their last occasion, many people were reporting significant amounts of “searching” behavior, many were actively basing their search on the needs of others, and that simple “look” criteria (attractive packaging, “interesting looking new brand,” etc.) were commonly mentioned.

Our analysis identified, very clearly, four purchase behavior clusters:

1. A very habitual, “auto-pilot” cluster, characterized by respondents, 97% of whom said that they “just chose the same brand they always do,” and who showed very low incidences on all other key reported behaviors.
2. An “active social shopper” group characterized by both a very high propensity to search for chocolates to share and who reported a large number of other searching and checking behaviors especially “brand related” ones (interesting new brands, “famous” brands, etc.).
3. A “cautious sharer” shopper group, that was above average in the degree to which they were looking for a brand to share, but who were well below average in terms of “superficial” search criteria (e.g., guessing what they'd taste like, famous brand, interesting new brand, etc.), and instead were above average on buying on promotion or carefully checking ingredients.
4. A “visual impulsive” orientated group, that were well below average on both buying to share with others and on “just buying the same brand,” but who were well above average on such behaviors as buying famous brands, searching several for one that looked like it would taste best, buying based on attractive packaging, etc.

It is clear that the more “habitual” a shopper's last purchase habits are, and the less they report “ephemeral” choice criteria (like attractive packaging), the more likely the brand leader is to enjoy an advantage over rivals. Patterns identified in terms of last purchase behaviors seem to relate to longer-term differences in brand choice, supporting the idea that what we are seeing are indeed signs of on-going strategiesÑnot just idiosyncratic “one-occasion” behaviors. Clearly this analysis needs to be repeated on more, and larger data sets, but initial indications are that segmenting people by very specific purchase behaviors is likely to enhance understanding of brand choice.