Good Morning. I am Kent Stephan and I am the President of Princeton Brand Econometrics. I was asked to come here today to discuss The Process Theory of Brand Choice. It is a general theory that explains how individual brands behave through fundamental laws of consumer purchase behavior. The theory results from modeling work PRINCETON BRAND ECONOMETRICS has been doing.
I am going to have to limit my discussion to established brands in categories that are frequently purchased. Time does not permit me to do more.
Let’s start with the basic concept of brand switching. The brand switching that goes on between established brands is not like divorce and marriage. It is really more like polygamy. Individuals often have reasons for using 2, 3 or even more brands in a category. This happens when people perceive that different brands provide different important benefits. Brand switching then results not from people’s changed perceptions of brands but in response to their fluctuating desires. As a matter of fact, once someone has used a brand, their perceptions of it become fixed and quite unresponsive to the brand’s own advertising.
Advertising persuades by making the brand more desirable but rarely by changing how it is perceived. Perception research proves this to be true.
If one were to hold prices constant (in other words, neither the brand nor its competitors are allowed to promote), you would see that some users of a given brand purchase it to the exclusion of all others. No matter what they desire from the category, this one brand always provides the best solution. Other individuals would use a fixed repertoire of certain brands in the category. For these individuals, no single brand satisfies all their requirements of the category.
Now, if in addition to assuming that there is no promotion, we also assume that there is no advertising, you would find that people who use a repertoire of brands allocate their purchases across the same set of brands – and in the same proportions–from one year to the next. For example, a group of people who used just two given brands and each allocated their purchases between them 60% / 40% one year would do the same the next year, under the assumptions we’ve made so far.
If you were also to measure these consumers’ preferences, again holding brands’ prices constant but not equal, you would arrive at the same 60% / 40% split. In the absence of market stimuli, purchase behavior would conform to brand preference.
This tells us something important. It means that even though we don’t know precisely which brand one of these people will purchase the next time the category is shopped, we do know it will be one of these two brands and that the probability for one brand being purchased is .6 and the probability for the other is .4.
Advertising and promotion intervene in this process of brand choice by temporarily changing the probability of a brand being selected the next time the category is shopped. Advertising and promotion can also have an influence on the amount someone purchases and how frequently purchases are made. Promotion can also temporarily pull in consumers who are not willing to pay full price for a brand. However, this discussion must focus mainly on purchase probabilities, not as much on volumetric considerations.
How these purchase probabilities are distributed across the brand’s user base (in other words, the proportion of consumers who purchase a brand 100% of the time, 90%, 80% and so forth), strongly predetermines how the brand will respond to its own advertising and promotion. For example, individuals who have a low probability of purchasing the brand will have that probability increased proportionately more by a given price discount than will individuals who have a high probability of purchasing it. This means that the more a brand’s user base skews toward low purchase probabilities, the more it will benefit from promotion.
The distribution of purchase probabilities also predetermines how the brand will respond to competitors’ advertising and promotion. Purchase probabilities represent a zero sum game; no one can have more than 100% probability of purchasing your product. So, if a brand that is in the set of products your consumers use does something to increase the probability that it will be purchased, this decreases the probability that your brand will be purchased.
Finally, the distribution of purchase probabilities predetermines how profitable a brand can be. If a brand has high purchase probabilities, this means that its consumers are automatically inclined to purchase it. If you look at scanner panel data, you will see that brands with large shares or true niche brands have user bases that skew toward high purchase probabilities. This is consistent with, and also explains, the PIMS finding that large share brands are more profitable than their smaller competitors.
By the way, it is possible to accurately predict how purchase probabilities will be distributed across a new brand’s user base, once it has become established. This means you can see in advance how profitable it will be to sustain the brand after it has reached equilibrium.
As a brand’s users move through their purchase cycles and make their purchases in response to their underlying preferences, as well as intervening price promotion, six distinct consumer segments emerge. These segments can be identified and quantified through preference research and sophisticated scanner panel analysis.
First, we have Sole Users. These are the people who prize your brand so much that they do not even respond when a competitor is offered at a discount. The more of these individuals in a brand’s user base, the more profitable the brand is. Sole Users require only enough advertising to sustain their desire for the category. Often, it is a stupid idea to deliberately promote to them, because they were going to purchase your product anyway. However, if the brand is in a category where usage is elastic (in other words, when people have a lot of product on hand, they use it much more quickly), then promotion can make sense. For example, cola marketers have found that if they reduce price and sell people a lot more product, the use-up rate increases dramatically.
For elastic categories, product availability in the home increases usage. In such categories, total category purchases are also responsive to increases or decreases in total category GRPs.
Next, we have Semi-Sole Users. These individuals are very much like Sole Users, except that they have at least one other brand positioned close to yours in their minds. So, when this other brand is being promoted, they are willing to purchase it. However, when the competition is at full price, they will always pick yours. If there are a lot of Semi-Sole Users in your user base, the competition is promoting, this can be a rationale for defensive promotion. However, our research has shown that each time someone purchases a brand at full price, their odds of purchasing it the next time go up a little. One the other hand, each time someone purchases at a discount, their odds of purchasing it again at full price decrease slightly. Over the long run, ongoing promotion converts purchases that were once based on preference into purchases that are now based on price. In general, large share brands and niche brands, because they have already won the preference war, will often be better off if they do not allow their smaller competitors to get them into a price war.
Discount Users. These are the other guys’ Semi-Sole Users. They are not prepared to purchase your brand at full price, but they will purchase it at a discount. If there are a lot of these individuals, this is a rationale for offensive promotion, even if the category is not terribly elastic.
Aware Non-Triers are not users of your brand, but merit brief discussion anyway because brand managers frequently think they can pick up a large number of Aware Non-Triers by spending a little more money. Reality is that if Aware Non-Triers have gone through several category purchase cycles without purchasing a brand, they likely never will. They simply have not bought into the brand’s positioning and more GRPs will not change this. To have a reasonable shot with them, you will have to start saying something new. Because they have not used the brand, their perceptions of it are still malleable.
Trier-Rejecters bought into the brand’s positioning, but not the product itself. Unless they can be given a dramatic new reason for using the brand, you will have to change the product itself in order to recycle them back through the brand.
Finally, there are Repertoire Users. These are the individuals who have a preference for your brand but also prefer at least one other brand in the category as well. They are the ones who generally determine the success or failure of annual marketing plans. If your advertising is persuasive, these are the people it persuades. And, in all likelihood, they will have been persuaded not because the advertising improved their perceptions of the brand but instead because it made them desire the brand more. The key to creating persuasive advertising is to first understand why it is that individuals select your brand over others and then make that reason more important to them relative to the reasons they have for using competing brands.
I said a few minutes ago that ongoing promotion converts purchases that were once based on preference into purchases that are now based on price. There are two very practical implications to this. The first is that promotion should only be used to accomplish what advertising cannot. And resources should be allocated on that basis. Also, promotion makes no sense at all if it is not currently profitable. Ongoing promotion erodes the brand’s baseline business over the long term, so if a promotion is not profitable right now, it should not be continued.
Advertising on the other hand, does not have to be currently profitable in order to be justifiable. In fact, most advertising is not currently profitable in terms of the incremental sales it produces within a purchase cycle.
Advertising helps sustain the brand’s ongoing baseline business–its equity. This is the most profitable thing advertising does for an established brand. Therefore, on-going advertising must be evaluated both in terms of the incremental sales it produces purchase-cycle-to-purchase-cycle as well as on its ability to sustain the brand’s baseline business. In fact, it is now possible to accurately predict the incremental sales an advertising execution will produce with a given amount of GRPs by using the ARS persuasion score. It is also possible to reliably predict, for example, the five-year trend for a brand if its advertising were pulled beginning today. Advertising can be evaluated on a sound financial basis.
This conference is about brand equity, and brand equity is really only a metaphor. What I am about to say may not be quite fair, but it makes a point. Each year advertising agencies walk into their clients and ask them to spend millions of dollars to buy a metaphor while, at the same time others come to those very same clients and ask them to spend money in order to receive some rate of return. Whether these other people who are competing for corporate dollars are honest in their projections or even have a clue what they’re talking about is irrelevant. They at least seem to offer more than a metaphor.
For brand equity to play the role it deserves in budget decisions, it must be expressed in terms of dollars and cents. This is the metric of business. It is the metric decision makers use to evaluate alternative choices. From what we have seen, advertising deserves a larger share of the pie than it now gets in many cases but it gets less than it deserves because it is perceived to be a high-risk crap shoot when compared to alternative tactics.
Advertising, the primary source of sustained equity, does not have to be a roll of the dice. Its returns can be reliably predicted and subjected to the same ROI evaluations as other business decisions. Admittedly, this may be a scary prospect to some people. However, brand equity is important. But it will continue to become an irrelevancy discussed mainly in conferences and corporate bull sessions as long as it remains a metaphor instead of a dollars and cents measure of value.