Synovate White Paper - Brand Value Creation - Communications and Equity

Brand Value Creation, Communications and Equity
Prepared By Jannie Hofmeyr and Ged Parton, Director of Innovation and CEO respectively, Synovate Brand and Communications Practice
Copyright Synovate 2006
Brand Value Creation, Communications, and Equity
This is a great time to be in marketing research. The opportunities for both improvement and for changing the way things are done, are huge. Weve put this document together to summarise our current thinking on a range of issues. What follows, therefore, is the framework which informs our practice and our offer. The document is a joint effort though mostly penned by Jannie. It stems from a week spent together in Cape Town in May 2006, hammering out a joint point of view about our industry. Numerous telephone calls and emails subsequently refined what we wanted to say.
To start: Some words and definitions

There is no dominant theory of brand equity or accepted definition of the word brand.1 Our science is in what Thomas Kuhn (1970) called a pre-paradigm phase. In other words: it is in a state in which many key words have no universal definition because there is no dominant theory. In what follows, we will be adding to the clutter with some definitions of our own. They are not especially original. i.) What is a brand? It would be tedious to review the many different ways in which the word brand is used. Id rather start where I always start when theres confusion and disagreement, and thats with a bit of history and a dictionary. So here goes, lifted from the Oxford English Dictionary. Brand: the word goes back to the medieval period in England. To brand was to burn with a hot iron, whether for marking or for cauterising. By 1587 it was already being used in a modern sense i.e., to mark indelibly, as proof of ownership, a sign of quality and by 1602 it was being used in a way which implies a bit of cognitive psychology: to impress indelibly on ones memory. Branding took off as an activity when manufacturing got into full swing in the 19th century. It was a simple way to indicate origins and promise quality. Pinker (1999) suggests that language has a universal basic structure. There are naming words and doing words. People organise their sense of reality by tagging bits of experience with names; and by describing what the names might or do get up to. Its clear from the Oxford English Dictionary that branding was intended, whether consciously or not, to exploit this natural tendency. Manufacturers who were confident that proof of origin would help to create sales, knew that branding would help both as a guarantee and as an aid to memory. Those who were less confident knew that one way to generate sales, was to counterfeit well-known brands. With this as the background, lets jump to the present and elaborate on the cognitive aspects of what a brand is: a brand is a set of hooks the mind uses to organise its experience of a commercial offering. Why a set of hooks and not just a hook? Because contemporary cognitive psychologists suggest that brand memories do not have specific locations (Ward, 2002); and that the simple network concept of how brand experience gets stored i.e., as a node with tentacles is not just simple, its simplistic. Memory and experience are more dynamic than that. The feelings and images that attach to a brand are widely dispersed across the brain. This means that there are multiple entry points for brand evocation. Call out the name Coca-Cola, present someone with a picture of a Coke bottle, or walk past a McDonalds every time that happens, tracks light up in the brain which evoke the brand Coke. Each time, the tracks vary as a function of previous encounters and the entry point, whether name, bottle, or smell of McDonalds, etc. Some tracks get reinforced and others get dampened. The point: brands can be evoked in many ways and each time what lights up and what gets laid down will be modified, even if only slightly. Understanding that brand memory and experience are dynamic does not affect our core definition of brand as a set of hooks. But it does have implications for how we model the formation of brand
experience through communication and use because it means that the decision about which hook to evoke when designing brand communication, is not trivial. ii.) The structure of brand experiences? If the cognitive character of brand is dynamically stored experience, then its important to have a view about the sources of brand experiences. There are mainly three: Experiences that come from use of the brand. This ranges from standing in a queue at a bank (say, HSBC), to driving in a car (e.g., Toyota), to eating a chocolate bar or pouring laundry detergent into a washing machine These interactions involve what I call transactional touchpoints (Type I). Experiences that come from exposure to brand communications created by marketers; e.g., TV advertising, mobile phone communications, public relations, and so on. Marketers do this to influence our brand memories and experiences. In saying this Im not assuming that exposures have to be explicitly noticed for brand activation to take place. All it requires is that tracks are laid down in the brain, whether they pass through focal consciousness or not. These interactions involve what I call marketing touch-points (Type II). Experiences that come from exposures to the brand that are not under the marketers control. Unmanufactured word-of-mouth is the most obvious of these. Another would be un-manufactured news. These kinds of interaction are incidental touch-points (Type III).
We dont have to be too philosophical about how to classify every conceivable brand interaction. Some interactions are fuzzy. For example, back in the 1980s, Levis jeans had a problem with the way discount retailers were treating the brand in store. People would walk into stores and turn the Levis stack upside down. By the end of the day, the Levis were an unsightly mess. I categorise this as a Type III interaction i.e., it is incidental because it does not involve a conscious effort on the part of the brand marketer. Nor does it involve use or consumption on the part of the shopper, even though it is a very direct confrontation with the brand. Levis responded by no longer supplying to those stores, and by opening stores of its own. By taking control of the visual merchandising they replaced damaging Type III interactions with brand-building Type II interactions. Lets take another example: suppose a friend gives you a lift in her car is this incidental or transactional? I would argue that its transactional because its an example of use or consumption on your part, even if passive. And finally: What about a news story that has been substantially influenced by a company spokesperson I think we have to consider this a mixed interaction. It is part marketing and part incidental. iii) Minds are active (even when apparently passive) William James, the great American psychologist of the 19th century, coined the phrase blooming, buzzing, confusion to describe the mind of an infant. Even though he was writing more than a hundred years ago, many of his ideas remain current. But on this score, he got things wrong. The human mind, and particularly the infant mind, is phenomenally well-organised. The speed with which infants learn language is nothing short of amazing. It would not be possible were their minds buzzing with confusion. The point is this: our minds interact with reality in ways which are powerfully directed by our goals. We are not passive receptacles of stimuli. Some people use the phrase the minds filters to draw attention to the fact that we sense only some of the stimuli that we confront every day. But this suggests a process in which we think of the mind as a kind of sorting machine; i.e., as attending to a range of stimulations, sorting through them to decide whats relevant and what isnt, and then letting some through and rejecting others. Its not like that. Our minds point towards the world in certain ways as a function of current tasks and personal goals. We are set-up to focus on certain stimuli and ignore others. We do not merely receive and sort. At the same time, evolution has designed us to pay attention to peripheral stimulation just in case something important
happens. So our minds remain open enough to allow reality to over-ride our predispositions as long as a stimulus is unusual and strong enough. It is this peripheral vision that lets through stimuli that are not a focal point of attention and that are often never activated or recalled without special techniques.2 So while it is true that we absorb only some of what we confront, it is not true that this involves conscious sorting. It is more a result of the way our attention jumps, steps, or meanders through the day, dragging along with it, its broad fuzzy edges. In the process, modern urban humans are immersed in brand experiences. Most brand experiences are probably traversed in a largely inattentive and unconscious way. Think, for example, of what happens when you sit in a traffic jam. What does the experience do to whats stored in your brain about your car; or your car radio; or your iPod; or to the music to which you may be listening? You may think youre just sitting (and perhaps cursing). Yet all the while your brain is accumulating brand experiences; and the sum of what sits in your brain in relation to each brand is constantly shifting its shape. Marketers inject their interventions into this landscape in an attempt to control or at least manage the shape the brand takes in the mind. It should be obvious that they face a challenge. iv). Consciousness and the unconscious If you believe in the unity of science, then you ought to believe that theories of mind and brain should be aligned. Up to now Ive been using the languages of mind and brain interchangeably. The question of why it helps to have two languages (i.e., of mind and brain) is not trivial Occams razor dictates that we should posit no more things than are necessary3. But you cant use words like the fuzzy edges of consciousness without recognising that youre dealing with both mind and brain; and without asking a question as to how they fit together. The question matters because there is a renewed interest by marketers in the idea that much of what accounts for human behaviour is buried in the unconscious. With that interest comes a natural interest in methods that get at whats buried. Usually, qualitative methods or their quantitative derivatives are invoked. But the theories that underpin such methods dont always fit comfortably with the theories that underpin more conventional quantitative methods. There is a danger then, that we slip too easily into simplistic generalisations and end up saying things like quantitative measurement cant get at the true motivations behind brand purchases; or qualitative research is just unreliable, impressionistic nonsense, etc. So how should we connect the two languages? My starting point, like that of many other people, is recent brain research (Ward, 2002). Heres what it seems to show: Patterns of brain activity associated with reports of consciousness are identifiable and different from remaining brain activity; All parts of the brain are involved in consciousness. To quote Ward: The theatre of consciousness is the entire brain (p.317).
I dont think many people would question the reality of consciousness. Were conscious; and thats that. By definition, the unconscious is harder to find and so there may be people who have difficulty believing that there is such a thing, particularly when faced by quotations like that from Ward. Yet, a lot of brain activity happens without being present to consciousness. Moreover, there is plenty of evidence that unconscious brain activity helps drives behaviour. How can this be so and Ward be right? The torch-light metaphor due to Crick and Koch (1990) works for me. If you think of consciousness as a light with fuzzy edges, then what Ward is saying is that there is no place in the brain that the light cant shine. But its also clear that the torch doesnt capture all of the brains current activity at any one moment. It only shines on a piece. Though it still has effects, the rest shades into darkness. When it comes to what we do, all brain activity plays a role; i.e., the stuff in the light, in the shadows, and in the dark.
Certain parts of brain activity, for example, the bits connected to the autonomic nervous system (e.g., cardiovascular regulation, etc.) are mostly in the dark. And some activity (e.g., reflex actions) bypass consciousness, at least temporarily. Even so, the bits that are mostly in the dark connect easily to consciousness when necessary (as in when we become aware that our hearts are pounding). Wendy Gordon recently published an article in which she suggested that most of whats important happens beyond the light of consciousness (2006). This, in my view, is typical of the over-simplification that traps us if we dont think the issue through. Lets try a simple question: could she have typed her article in a state of unconsciousness? I mean: its surely ironic that we cant even study the unconscious, without making it conscious. The problem I think is this: because consciousness is so central to our daily experience and organisation of life (like the sea is to fish), we tend to take its magic for granted. Reverting to the metaphor of the torch light with fuzzy edges: consciousness appears to have an evolutionary function and that is: To put the focal point of relevant brain activity into the mind. By that I mean: whats present to your consciousness at any moment, is what the brain has determined is most important if youre to achieve your goals. Take a concrete example Tiger Woods. When Woods hits a golf ball there is a phenomenal amount of complex brain activity going on. Most of it is surely at the edge of or beyond consciousness. But lets not forget the hours of conscious practice he put in to develop his swing. And lets remember that if he didnt keep his eye on the ball (the most conscious bit), he would surely make many more mistakes than he makes. So it goes for the rest of our daily activity we have to put our mind where the action is if we want to achieve our goals. Even those who are researching what is not present to consciousness, have to do so. Yet, there is no question that the life of the mind (brain) is richer than whats easily present to consciousness. In other words, there are bits that the torch doesnt get to easily. Experience teaches us that we need specialised evocative, projective, or nudging techniques to get the light to shine on those bits. What about theorists like Freud and Jung? Well, if you read Freud (i.e., in the original), then you will be struck by the extent to which he was trying to construct a mechanical (i.e., Newtonian) physics of the brain. Jung less so. With the benefit of hindsight, the theories of both look like theories of the mind rather than brain. A discussion of their views, informed by a commitment to the unity of science, deserves a separate appendix. v). How much does the brain record, but the torch leave in the dark? If it is true that consciousness is connected to most parts of the brain as long as the brain directs it there, but that what comes to mind is always only a part of whats going on; and if it is also true that special techniques like hypnotism or depth interviews seem to bring more to light than unaided introspection, then one of the more interesting questions is: just how much does the brain record? Does it, perhaps, record everything? I might as well express my view upfront: I agree that all of experience may leave some trace in the brain. But I disagree with people who say we remember everything, its just that we cant always find our memories. Lets see why. The more we learn about the brain, the more we understand that memory doesnt just play back the past; it reconstructs it. For example, if I cast my mind back to events that I remember clearly, I almost always see myself in the event as if I were watching a play. And yet we never actually see the whole of ourselves in situations while were living through them. To take a simple illustration, we dont see our faces. The hard evidence for the reconstructive model of memory comes from research done in the United States to test the veracity of eyewitness accounts in court cases (Schacter, 2001). Based on this research it is clear that memory is reconstructed, not played back; and that the nature of the prompt that is used to recall the memory, can play a fundamental role in what gets reconstructed. This leads to the question: how do we know which kinds of prompting activate the brains powers of invention rather than the brains powers of recall? How do we separate recreated memories from real events which have been brought back?
I do not need to be convinced that people can be prompted to recall long lost memories; and that they can do so reasonably accurately. The evidence suggests therefore that experience does leave traces in the brain which we dont always access readily. So Im comfortable agreeing that we remember more than we think we do including a lot more brand exposure than is easily recalled. But we cant use anecdotal accounts of the apparently miraculous recovery of particular experiences, to prove that the brain records everything. The difficulty with validating recovered memories is that we cant go back to check. And even if we can, we can usually only go back partially. Unfortunately for the humans remember everything school, the weight of current evidence points away from the thesis. The brain is an incredibly plastic instrument. It is constantly reshaped as experience goes on. And as it changes its shape, so can memories. Eventually, like the rest of the body, it loses some of its plasticity. Real physical stuff gets into the brain and clogs it up then memory is lost forever. Its looking increasingly unlikely, therefore, that we will one day discover all of our histories buried somewhere in our brains; and that under the proper prompting, those histories would be available for recall. Id be happy to change my view if evidence to the contrary could be brought forth. vi). Emotion In 1969, when Ekman showed that there are six universal human emotions, some academics on the social sciences side of Western campuses called him a fascist. Since then, people have calmed down: there are at least six universal human emotions and they show up on our faces in the same ways, worldwide. If we were to stumble across a smiling head-hunter in the jungles of North Borneo, we would know whether the smile was genuine and whether he was trying to be friendly. The universal emotions are: happiness, sadness, anger, fear, surprise, and disgust. Ekman has since added contempt to the list. And others have shown that while certain emotions may not be present at birth, they are present surprisingly early in most (not all) peoples lives: guilt, shame, jealousy, envy, and pride. The core emotions are inextricably linked to the autonomic nervous system. To quote Levenson (2003): Human emotions evolved to deal with situations of great significance to the individual and the group in which a rapid, multi-system response is required. To continue: The core system (of emotion) has all of the capabilities necessary for processing incoming information continuously (my italics) and for detecting a small number of prototypical situations that have profound implications for the organisms immediate well-being and long-term survival. In other words, the core emotions are largely involuntary. We are in the 21st century partly because our ancestors had reflexes which enabled them to cope with deadly circumstances fast enough to live (e.g., when stumbling across a poisonous snake on a path). The associated emotions showed up on their faces because of their value as a signal to others. But the signals would have had no value unless the others were able to interpret them. Many of the others would have been relatives. Our ancestors survived if they acted fast enough and signalled efficiently; and their relatives survived if they were able to interpret the signal. So, fear = danger attack, jump, run, or whatever. Disgust = horrible, do not touch, do not pop into mouth. Guilt = I still want to be friends, I understand that I shouldnt do that again. Smile = Im not a threat, Im a friend you do not have to kill me because Im not planning to kill you. And so on Lets summarise some of the universal aspects of the core emotions: They are found in all cultures. They show in peoples faces in identical facial expressions. People are expert at interpreting the associated facial expressions. Across the world, people know the emotions. Numerous situations evoke the core emotions; e.g,. death of relatives evokes anguish (amplified sadness), snakes evoke fear. The core emotions activate the whole body: appropriate parts of the body jump to attention so that action can follow.
Speed is of the essence in emotion. Thats one of the reasons the emotions are hard to control. They happen before we know it. We see a friend and smile without thinking. Emotions are an instance where consciousness catches up a while after the event. I see a puff-adder on a path (a lazy, poisonous South African snake that doesnt flee the approach of humans). I leap backwards in a way which is humanly impossible. I run and then I stop and discover that my whole body is shaking4. The word engagement has become very topical in the world of branding. Emotions matter in part because most people in our industry suppose that emotions are implicated in engagement. We agree. But were also critical of the frankly undisciplined and intellectually lazy ways in which the word emotion tends to be used. What we mean by emotion is informed by what scientists working at the interface of the human, neural, and biological sciences are telling us about it. vii). Brand equity and accountants From the intangible edges of consciousness to the very tangible: accountants and the stuff that excites them money! I think of brand equity in a largely classical, financial way: it is the net present value of the profit stream that we can expect a brand to generate over time. This being so, businesses have to answer the question: how much investment should go to marketing or branding? Marketers have to answer the additional question: what use of marketing resources will create the greatest brand value over time? The marketing research profession is not known for its understanding of business systems. This is a weakness. Every business has options with respect to resource allocation. Although marketing should be at the heart of brand-owning businesses, there may be times when the greatest value can be added, not by marketing, but by financial engineering (for example). And value can be created in many ways: e.g., by money management (the task of the treasury); by driving productivity (the task of leadership along with business process engineers like the IT department); or by higher margins (facilitated by strong branding); and so on... Marketings task within a business system, is to build a case for the resources it believes are needed to create brand value; and then to allocate spending across different marketing options in a way that maximises value creation. Marketing research should help by providing intelligence that justifies brand investments; and by providing further intelligence that helps allocate those investments across the range of marketing possibilities. How much money should go to marketing and how much to financial engineering this is a question marketing research should help answer. And how much money should go to TV advertising, to buzz-marketing, or to opening up new territories again, marketing research should help provide answers. If we cannot show that brand investments are worth making, or what marketing investments to make, then we shouldnt be in business. This means: the information we as marketing researchers supply to marketers must be linked to the creation of brand value. Profitable sales over time are the ultimate dependent variable. Some might argue that shareholder value should be the ultimate dependent variable. In my view, it depends on what you mean by shareholder value. If you mean market capitalisation, then I dont agree. The problem with market capitalisation, as people like Mandelbrot have been pointing out since the 1960s (Mandelbrot, 2004), is that the price people pay for a share in a company is often disconnected from the value of the company. And I mean often. What statisticians call fat-tailed distributions describe changes in asset prices. If share prices were connected to brand value, then price changes would not be so littered with drama. Market capitalisation has too much to do with the psychology of the investment community and not enough to do with the underlying value of the businesses. So lets start by keeping an eye on whats real and by trying to keep the accountants happy: the purpose of strategic brand management is to drive profitable sales over time. The greater the profits, the greater the brand value. As long as this is what we mean by shareholder value, then shareholder value is acceptable as the ultimate dependent variable. Our framework must illuminate the process of profit and shareholder value creation.
viii). What is a market One last thing before we get into brand equity: what is a market? In almost all attitudinal marketing research we ask people what they regularly buy. We seldom follow up, however, with a question about what they would buy if so-and-so were not available. If we did, the answers people gave would probably depend on the category. If we were talking about Ariel laundry detergent, people would probably respond by naming another brand of laundry detergent. But if we were talking about Coca-Cola, people could respond by saying tea, Snapple, Pepsi, coffee, bottled mineral water, crisps, or even ice-cream. The question of market definition has always bothered me. Like so much else in our industry, markets are fuzzy. Even laundry detergents arent straight forward. People have a choice to buy a premium brand detergent or to buy a store brand and spend the savings on something else, say, a Coke. Our clients are often quite myopic about the nature of their competition. Usually they (and we) think of competition as something thats alike. But Ive done studies where 30% or more of the true competition was excluded from the brand list by a combination of definition and client myopia. What can we do about this? Fortunately, there is a general principle that applies: people divert time and money away from the things that are less important to them; and towards the things that matter more. The higher a persons involvement in a market, whether it includes apples and oranges (the fruit market) or only apples (the apple market), the more they will be prepared to spend in it. The way to divert money from Coke back to a premium laundry detergent brand, is to lift involvement in laundry detergents. This has implications for both marketing and research: From the marketing point of view, creating brand involvement helps to drive resources in a brands direction against all comers (and I mean all). From the research point of view, a properly phrased question about the importance of the brand decision, should be in every study.
There is more that we can do. We can recognise that markets have poorly defined edges. And we can design our questionnaires in such a way that we include the choices that are relevant to a respondent. In other words, we can allow respondents to define the market. So: we know how to lift category spending just lift the involvement. And we know that we need to pay attention to what people switch between, when defining a competitive set. Lets move on now to consider the challenge of brand value creation against the backdrop of all the above words and definitions. ix). Brand value creation I divide the challenge of brand value creation into two components. First, theres the communications challenge; i.e., the challenge of shaping what people believe about brands. And second, theres the sales challenge; i.e., the challenge of pulling marketing levers that will help sell brands. The first is about what gets into the mind. The second is about what people do with what gets in.
The Brand Equity Framework: Understanding communications

i). Part One: The nature of brand exposure The following graphic shows a picture of a person. She is bombarded by stimulation (though note that she doesnt necessarily feel bombarded)5.
Most of what our lady faces doesnt have to do with brands. Ive called that everything else. Then theres Stuff about brands which is about brands. As Ive already suggested, this can be decomposed into three types: first, the experience that comes from brand use i.e., transactions; second, the experience that comes from what the client does (including communicate) to try to induce sales i.e., marketing interactions; and third, the exposure that comes from other ways in which brands are encountered i.e., incidental interactions. The structure of exposure varies in interesting ways as a function of product category. As an example, compare cars and fruit juices.
I: The nature of brand exposure
Everything Else
Stuff about brands A Typical Brand Brand Use
Stores visited TV watched Cell -phone Packaging Etc. News stories Word of mouth Channel owners Etc.
What Marketers Do
Total Potential Exposure
What Others Do
I own a Toyota. Transactional interactions are by far the most importance source of whats both in my mind and my brain about Toyota. Moreover, if I think about it, it comes down to two main kinds of transaction: first, its about what goes through my mind has I walk towards the car and get into it; and second, its about how I feel when Im driving it. Toyota has an excellent service and maintenance plan but that only kicks in about once every six weeks. What role does advertising play? Not much. Because Im committed to Toyota, I dont need much persuasion or reassurance, so I dont pay much attention to their advertising. I see their ads. And it might damage my sense of Toyota if the company ever came up with a campaign that offended my sense of the brand. But they havent so far; and so the advertising registers in a mildly reinforcing way. I see lots of Toyotas on the road because it is the market leader in South Africa. That plays a bigger role than advertising because it reinforces what I think about the look of the brand. The other possible role that market leadership could play i.e., that of reassurance is irrelevant to me. I dont need to see how many people drive Toyotas to be reassured about my choice. A final note: because Im so committed to Toyota and generally disinterested in further information about the category, I pay very little attention to information about other brands. Which means: for other brand marketers to get anything into my mind about their brands, they would first have to climb the mountain of my commitment to Toyota. In an important sense: what I think about other brands is caused by my commitment to Toyota and low category involvement.
Now consider fruit juice: I hardly ever drink fruit juice so unlike cars, transactional interactions play little role in my image of fruit juice brands. Im not interested in the category and so marketing communications play a very small role. Some people suggest that you would probably find quite a lot about fruit juice brands sloshing around in my brain if you hypnotised me. But its almost certainly sitting in some cabinet called irrelevant for the time being because its not having much impact. My conscious sense of fruit juice is dominated by the sight of fruit juice packaging in stores. Its boxy or in bottles. Its relatively colourful, but otherwise rather flat and dull. Its next to butter or bottled mineral water or other semi-cooled stuff. My perceptions are rounded of by a lifetimes worth of propaganda about how good fruit is for you. It doesnt seem to have had much effect. At this point I want to draw attention to just two implications of these anecdotes: The structure of exposure varies as a function of the differences across product categories. Its worth thinking about this when designing research. It would be worth creating a data-bank of learnings about the differences. The anecdotes have implications for measurement: if we want to understand whats in someones mind about a brand, then we have to measure everything thats relevant to how it got there. This simple graphic tells us: o o There is a lot that we have to take into account, especially brand encounters that are usually not measured; e.g., day-to-day transactions Its not only own brand encounters that are relevant. Other brand encounters also shape what we think about our own brands
Finally, note that the life-stage of a persons relationship with a brand is important. In the early days of my relationship with Toyota, I paid quite a lot of attention to Toyota advertising. In those days I was more involved. Now that Im less involved I pay much less attention.
ii). Part Two: What gets into the mind Perception is a process. It starts with what gets picked or slips through, continues along nervous mechanisms; and then hits the brain which reverberates. Though its experienced as seamless, we can characterise it as a two- stage process: first, theres what gets through; and second, theres the sense we make of it (remembering that that sense is never static because of the dynamic nature of memory). Take a look at the graphic below by way of illustration and lets return to Toyota and fruit juice for further examples: Commitment to Toyota coupled to low category involvement dramatically dampens my interest in communications about motor cars. Low category involvement dominates my sense of fruit juice brands. I hardly register advertising; and if I do, its marked as currently irrelevant. Implications: we cannot understand what gets through perceptions gates without measuring category involvement and prior brand relationships.
We know other things about how the brain/mind works. People can be highly involved in a category, yet loyal to nothing in it. For example: Im highly involved in and committed to the red wine category, but highly promiscuous within it. This even goes with multiple commitment to some of the brands within it (in the same way that Im committed to all four of our children). But my behavioural loyalty to any one brand is low. And it results in indifference with respect to all other alcoholic beverages.
10
II: What gets into the mind?
This set of personal characteristics has very strong effects on both what I see of brand communications for alcoholic beverages; and on what I do with what I see: I pay very little attention to communications about, for example, beer or whisky brands. I pay a lot of attention to information about red wine and red wine brands. Some beer and whisky brands are so heavily advertised that I cannot help seeing and absorbing it. Like much else, the information goes into a place in my mind marked irrelevant for the time being.
The configuration of my mind and brain in the car market is different. In this case, Im very committed to a brand, but almost completely disinterested in the category. The combination of high brand commitment and low category involvement dampens my interest in category communications. But as with beer and whisky, its a heavily advertised category. And so, as with beer or whisky, most of what lays tracks in my brain whether conscious or unconscious - goes into that place called irrelevant for the time being (theres a lot of stuff in there). If I do see Toyota advertising, it tends to go to a place called oh, okay, good... Recently my old Toyota (a Camry) finally ground to a halt at the roadside. Id had it for eleven years. One of the effects of my low category involvement is that I have no interest in replacing my cars until they stop. I got out, got home, and had a new Toyota within two hours: a Corolla I traded down. You can tell that I didnt spend much time agonising over this decision. Its probably fair to characterise most people in the motor car market as involved. People think quite carefully about what car to buy, even if theyre committed. Lets consider the process a person goes through when they own a car, starting with new car ownership. We know theres a tendency for new car owners to indulge in post-purchase rationalisation. This is driven
11
by the importance of the category decision and their need to feel that they have made a good choice. In this stage people may consume quite a lot of marketing communications, including for cars they didnt buy. Information consumption is high its part of the rationalisation process. Yet it tends to be pressed into service to support the choice theyve just made. In other words, as far as possible, people will use competitor communications to reinforce (what may be temporary) current commitment! At some point the honeymoon wears off and people settle into the mode of day-to-day ownership. Heres what we know about how they will behave: even if they remain committed to their brand through this stage, they will still monitor quite a lot of marketing communications as long as they stay category involved. Moreover, they will continue to use competitor information to reinforce their current brand commitment if they can. So, for example: committed owners of BMW may actively watch Mercedes (or VW, or whatever) advertising with a view to continuing to build arguments as to why BMW is better. To the extent that they succeed in such rationalisation, the effect of the Mercedes adverts will be to reinforce commitment to BMW (Hofmeyr, 2001). Its worth dwelling on this point: we know that committed involved people look at marketing communications for both their own and competitor brands (unlike me in the car market who is committed but uninvolved). We also know that the marketing communications they see for competitor brands dont seem to make much impression on their views of the competitor. The strategy of a committed involved person seems to be to monitor other brand messages in order to check that their own brand is better. Mostly the evidence suggests that they succeed. It is possible, however, to beam communications to people that puts a question mark in their minds. If they then interrogate their own brand and find no answer, it may become possible to dislodge their commitment. But research over some 15 years on the psychological dynamics of commitment suggests that it will take time; and frequently committed people manage to rationalise the question mark away. When a car owner starts thinking about buying a new car, they shift into information search as long as theyre involved enough. At this point, competitor information may be considered more carefully. If theyre committed however, then their existing commitment is likely to steer them into this phase in important ways. For example, theyll be more likely to go to their current brand dealerships first; and theyll be more likely to pay attention to communications about their current brand. iii). A summary: implications for marketing and marketing research Our task is to help marketers position their brands in the best possible way. To do this we have to understand the connection between what marketers might do and what people end up thinking. Some form of modelling is always involved, even if its a simple two-way table of the claimed advertising exposure by brand image. Marketers are increasingly realising that our simple approaches are too simplistic. Many influences go to make up what a person ends up thinking and feeling about a brand. Its time to pull this together and spell out the implications for measurement: The image a brand has cannot be understood without measuring all the potential sources of that image. Leaving out a source may lead to mistaken conclusions about whats driving image. Any brands image isnt affected by information about that brand alone. How people feel about competitor brands directly influences how they interpret information about their own brands. Category involvement plays a critical role in terms of a persons overall disposition to consume information. The higher the involvement, the greater the likely consumption.
Marketing communications are so prevalent in some categories (e.g., motor cars, beer) that even those who want to, cant avoid it. What effect does it have? Our research suggests that the effect will depend on the brand life-cycle. When it comes to really well-known brands, communications can hardly have an effect on awareness because awareness of both the brands and their communications tends to be high already. In addition, unless communication is unusually creative, its unlikely to affect brand image that much. If anything, its in these situations that Ehrenburgs model of the nudging effects of advertising may be most accurate; i.e., the ads dont increase awareness, but they maybe nudge the brand closer to
12
the front of the queue on any given day. This in turn leads to a temporary increase in purchase likelihood and accounts for the day-after lift in sales thats been found in split cable tests of TV advertising exposure. Given all this, it may seem that you cannot be sure youve picked up whats driving a brands image unless you measure everything about everything. Knowing how the mind sets itself up to be more open to some kinds of information than others, can help us to design questionnaires which ask each respondent mainly about whats relevant to them. Our strategy should therefore be to exploit the fact that everyone has a big place in their minds set aside for stuff thats currently irrelevant. As a minimum, we can immediately improve the quality of our testing and tracking questionnaires by including measures of: Category involvement: moderates both the propensity to consume information about the category and the likely method of dealing with whatever gets registered; Existing brand relationships: attention skews towards favoured brands. Communications are therefore more likely to be seen; responses to favoured brand communications are more likely to be positive.
The Brand Equity Framework: Predicting what people will buy

i). Working with the brains shorthand; i.e., summary evaluations In satisfaction research we draw a distinction between transactional satisfaction i.e., how you feel about a particular experience (with, for example, your bank) and overall satisfaction; i.e., how you feel (about your bank) in general. Although the former influences the latter, particularly at the beginning of a relationship, the latter can become quite stable over time so that a persons mood after any transaction tends to drift back to an underlying and probably slowly changing equilibrium level (Bolton, 1998). We measure both levels with questions about satisfaction. The answers people give at both levels are what I call summary evaluations (Hofmeyr, 1990; Hofmeyr and Rice, 2000). For example, the answer to a question about an experience youve just had at, say, Pizza Hut, is a summary of your feelings about the detailed event-stream youve just lived through; e.g., the waiting time, the food, the service, and so on. The answer to the broader question about Pizza Hut as a place to eat, is a function of the cumulative experience resulting from all your visits. While the answer to the first question is a summary of a particular event stream, the answer to the second is a summary of summaries. Whether particular or general, people dont have too much difficulty answering these questions. Thats because the questions seem to be tapping into something real. And what theyre tapping into is the brains talent for shorthand. Summary evaluations are the way that the brain lets us know how confident we can feel about a decision. When were satisfied, for example, we know that we dont have to think about the decision too much. By contrast, dissatisfaction is the brains way of motivating us to continue thinking. Without this shorthand, life would be far more challenging than it already is because wed be forced into reworking the consequences of every decision. The brains shorthand is yet another illustration of the magic of consciousness; i.e., of the way the brain organises awareness to be (mostly) in the right place. Now heres the paradox: while people tend to have no difficulty answering questions about these summaries, they often have difficulty answering questions about the detail. So people make choices about which car to buy or which soft drink to drink; e.g., BMW or Coca-Cola. And they quite easily rate both the last consumption experience and the brands overall. But speak to them about the details of what the brands mean to them and theyll only be articulate to a degree. Good qualitative researchers almost always lift more to the motivational surface than people do by themselves. So there are two levels of cognitive processing: first, the summary with respect to which we tend to have no difficulty answering questions even if it is to say Im undecided; and then the detailed with respect to which most of us are more or less articulate, depending on our introspective abilities and capacity not to lie to ourselves (See Note 6 for a the questions we will use to measure equity).
13
From the marketing research point of view: To understand whats in someones mind about a brand, we have to measure both the summary and the detail; It is the job of the more qualitative or projective techniques (depth psychology, social anthropology) to help fill in the detail; The key to economy of measurement and prediction, however, is to identify valid and reliable quantitative summary measures.
ii). Brand experience, brand memory, brand characteristics Ive already argued that whats stored in the brain about brands involves a dynamic network that can be evoked in multiple ways as a function of the entry point (Cokes name, bottle, etc,). Ive also drawn attention to what the latest brain science tells us; namely, that each evocation will be new as a function of changes in the pattern of activation. Over time, however, the scale of the variation across particular brand experiences should get smaller as long as there is consistency in the way that the brand presents itself. Over time, then, each evocation may be dynamic in smaller and smaller ways and so we develop an abiding sense of what a brand is and does both in its details and in general. The greater the consistency and the greater our accumulated experience, the more inert this sense is likely to be so that, in time, it may become almost impossible to change what someone thinks about a brand. Eventually we get older and our brains degrade then brand evocation dims. Some nice research areas flow from this: How do the transactional and the overall levels of evaluation interact over time? What are the conditions for the development of a stable, long term equilibrium brand evaluation? What is the relationship between what people believe about a brand and what they want? To what extent is brand switching driven more by changes in what is relevant to people than by changes in what they believe? Is there a place in marketing research for the psychological equivalent of Shannons measure of information (1948); e.g., by treating psychological surprise as an analogue for probability? o o Shannon defined the amount of information contained in a signal in terms of its probability. The less probable a signal, the more information it contains. Perhaps we could look at brand exposure in the same way: the more surprising it is, the more information it contains and the more cognitive reorganisation it will initiate.
In the following graphic, Ive represented the levels of detail and general by what Ive called brand characteristics (details) and brand relationships (summaries).
14
III: How what gets into the mind, is processed
Lets deal with the detail first. Many researchers define brand equity in terms of a list. Here is a generic list adapted from Keller and Lehman (2004): Awareness first mention, unaided, aided Associations characteristics or attributes: differentiation, relevance, popularity, quality, trust, value, price sensitivity, etc. Relationship loyalty, commitment Behaviour recommendation, share of wallet, repurchase rate.
I do not believe in trying to develop a universal list for the detail brand characteristics column. Here are the reasons: Product categories differ from each other in terms of the attributes and characteristics that may be most relevant to brand strength; Over time, new brands or category innovations may change whats relevant to brand strength within a product category; Attributes like quality/trust arent helpful by themselves. The question when, for example, quality is an issue, tends to be okay, so tell me what to do7.
15
In any case, rather than be prescriptive, Ive listed families of attributes or characteristics that might form part of a typical brand equity study. Our guiding principle: this is one of the most important parts of the questionnaire. Without good detail measures, we cant help marketers with what to do. Every study must include measures which uncover the details of what people feel about brands. In developing the list we need to retain the flexibility to be guided by current market realities and to allow for continuous adaptation or improvement to our methods as necessary. iii). A word about differentiation One of the most deeply embedded ideas in marketing is that a brand must have a unique selling point or, in its modern form, that it should be differentiated. But think for a moment about what differentiation means. It means: delivering on a benefit or property that others dont. This may involve functional attributes; or it may involve the successful evocation of feelings and emotions. Either way, what it means, in quantitative terms, is that the brand scores well in relation to what it does for a person and that other brands dont score. In other words, its a special case of being better than other brands. The word I use to describe this is advantage the brand is seen to have an advantage over others with respect to this characteristic. If differentiation is nothing but a special case of advantage, then why do people use the word? What are the psychological characteristics that make people say that brand is unique or special, rather than, that brand is all about <property x> no other brand offers that? Perhaps its a matter of whether or not we believe we have a choice of brands. When we think that two or more brands offer a particular property, then we compare and choose the one we rate best; and we feel uncomfortable using the word differentiated. We prefer the word better. But if we feel that a brand does something no other brand does, then it comes more naturally to say this brand is different its got something none of the others have. Whether better or different, a phrase that captures this is relevant advantage. In other words, whether other brands are seen to offer a property (in which case we tend to say that the brand is better) or whether they dont (in which case we tend to say the brand is different), a brand must be seen to have an advantage in terms of a property and the property it offers must be relevant. iv). The measurement of brand relationship The Conversion Model, a competitor product marketed by TNS, but which I developed in 1988, arrives at remarkably accurate measures of brand connection and relationship on the basis of very few questions. To do this it exploits what Ive referred to above as the brains shorthand; i.e., it is based on the identification of very few core metrics which accurately get at the way the brain summarises how it feels about any thing overall. The theory is in the public domain (Hofmeyr, 1990; Hofmeyr and Rice, 2000). It is based on the butterfly cusp catastrophe (you will find a picture of the butterfly cusp as it applies to brand relationships on the next page). Its important to distinguish between the theory of brand relationships from the Conversion Model itself. The latter is just one way to operationalise the theory. Synovates The Momentum Engine also draws on the theory. Most existing methods (e.g., classical customer satisfaction approaches) implement pieces of the theory.
16
IV: A Theory of Brand Relationship based on the Butterfly-Cusp Catastrophe
What elements of the brains shorthand does the theory exploit? There are three main ones: The extent to which there are both negatives and positives associated with an option; i.e., ambivalence (See Cacioppo and Berntson, 1994). The overall sense of the current or potential impact of that option on your life; i.e., the extent to which the option matters to you. The summary sense when all is said and done, of how an option fits with your needs and values; i.e., classical satisfaction.
Do not allow the picture to intimidate you. Lets unpack it. The bottom horizontal axis is labelled Overall Evaluation. Its based on the answer people give to a simple question about how they feel about an option overall; e.g., a bank or a laundry detergent. This is the summary of summaries mentioned before. The axis which goes back into the page is called Importance of Relationship. This is also a summary. Its the answer to a question about the overall relevance of an option. The surface on top of the graphic splits in the middle into something vaguely resembling a butterfly shape (hence, butterfly cusp catastrophe). It captures the effect of competing negatives and positives. To the right is the vertical axis; i.e., the outcome variable: Nature of Relationship. Now look at the back of the top surface. This represents an area where involvement is low but people still rate brands. Lets take an example: laundry detergents. Many people do not think of laundry detergents as an important part of their lives. Still, they have feelings about which laundry detergent brands are best; and happily rate them for us.
17
The theory predicts that people in this region will feel a mild attraction to brands they think are good; and a mild aversion to brands they think are bad. They will be indifferent to brands they rate neither positive nor negative. Because theyre uninvolved, all their feelings will be shallow; and they will drift from the positive, through the indifferent, and towards the negative in a relatively linear way, as a function of their overall satisfaction ratings. Change happens very casually in this region of the butterfly cusp. So: the theory outlines the exact nature of the relationship a person will have with a brand (whether shallow positive, indifferent, or shallow negative), as a function of variations in overall rating at low levels of involvement. Using the word commitment in this region, is inappropriate. The commitment regions, portrayed at the front edge of the top surface, are far more complex. Theyre activated when involvement is high. So lets take something we know many people are passionate about in the United States; e.g., abortion. In the United States, strong lobbies represent both a womans right to choose (pro-choice) and an embryos right to live (pro-life). Many people are intensely committed to one and against the other. What drives both states of commitment is high involvement coupled to positive evaluations of the one and negative evaluations of the other. In this region of the cusp both options evoke involvement, but one is judged negatively and the other positively. Commitments formed at high levels of involvement tend to be sticky and polarising; i.e., people in these states of mind tend to find it impossible to be neutral (thats the reason for the fold at the front end of the cusp); and it tends to be hard, if not impossible, to change their minds. As an example, lets take someone who is pro-choice. How do they feel when shown graphic images of an embryo that is recognisably human, being destroyed? Or take someone who is pro-life: how do they feel when shown pictures of poor single mothers being intimidated by picket lines, or of botched illegal abortions? The point is that the more important something is to a person, the more resistant they will be to any information which could move them in the opposite direction. Human beings have multiple methods for coping with negative information in situations like this. What involvement does, is activate those methods (Aronson, 1988). In our book (Hofmeyr and Rice, 2000), we use the analogy of personal relationship to help explain the front end of the graphic. For example, take a couple whose relationship has gone sour. These days people break up or divorce much more readily than they used to. Yet most people still make some effort before giving up on a personal relationship. Why? Our answer: involvement. Its a function of the extent to which that relationship is important to them. Look again at the front edge of the top surface. Its a state of mind in which people stay committed even as they become less satisfied. When people really care about a relationship, they dont just give up. The preferred strategy of a committed, involved person, is to try to save the relationship. Many researchers question whether this theory applies, for example, in business-to-business relationships, or in service categories or infrequently bought goods. I find this perplexing. Think, for example, about our business. We have clients (obviously). Over time, we develop relationships with our clients. In fact, we put a lot of effort into developing such relationships. We invite them to conferences. We organise special events. We get together semi-casually every now and then. We try to encourage a state of mind on their part in which they will be forgiving when we make mistakes. And it works! Of course, not every relationship can be abused forever; and no client is going to tolerate an endless series of mistakes. But as long as a sense of partnership has developed, clients will be more likely to be forgiving than to switch. Dissatisfaction will be tolerated, at least for a while. And its not just physical obstacles to change that cause this (e.g., the potential for discontinuity in tracking data; or the problem of being trapped by legacy systems). Even in business (perhaps even more so in business), we have a tendency to form relationships that matter to us. And when we do, we become more forgiving. One last observation: notice that it would be very difficult for a person to start off a committed relationship in an involved, dissatisfied state of mind. Usually though not always, intense commitments to something are formed from a combination of high involvement and positive evaluation. As long as involvement remains high, the theory then predicts that a person will remain committed, even as their evaluations start to change. What this means is that most people who are found in a state of dissatisfied commitment, are likely to have got there via a state of satisfied commitment. It is possible for people to jump straight into a committed relationship while in a dissatisfied state, particularly (as the theory makes clear) if they start off by being ambivalent but are then pushed to a decision. But this is rare.
18
Ive dealt with the theory in much greater detail elsewhere (Hofmeyr, 1990: Hofmeyr and Rice, 2000). Those who are interested should look at my paper or our book. Synovates core equity metrics are informed by this theory and will continue to be until something better comes along. v). Misconceptions about involvement Since involvement is so central to our conception of brand relationships, it may be worth devoting a moment to clearing up some misconceptions about involvement. Every now and then I come across marketers who say oh mine is a low involvement category (or market) therefore, brand building is impossible. Any marketer who holds that view should be fired on the spot. Over a fifteen year period in over 7000 projects, the Conversion Model team have shown that there are no categories or markets in which there arent at least some involved people. While it is true that categories may vary in terms of the extent to which people get involved in them, every market is characterised by some people who do care about the brand choice and some who dont. What matters therefore is: what is the distribution of involvement and how can it be increased. For a great example of the creation of involvement in a category, look no further than the intel inside campaign. Before the campaign, people were unaware that micro-chips were branded. In 1992 Intel initiated the intel inside campaign and everything changed. By the mid-90s, not only had Intel turned the market for micro-chips into a firmly branded market, but theyd also ensured that a significant proportion of people in the market cared about which brand was inside. Involvement varies in another way: according to circumstances and personal dispositions over a life-span. As Ive already indicated, when I first bought a car, it was an involved decision. Now I no longer care. Or take another example: I went through many years not caring about which brand of toothpaste I used. But as I got older my teeth got permanently stained. When toothpaste marketers launched brands with whitener in them, I suddenly took notice. A decision which had been relatively unimportant to me, suddenly became more important. What did it was the manufacturers ability to identify an unmet need; and so, raise the relevance of the category to me. Even a cursory look at the history of advertising yields many examples of great marketing which created involvement where before there was none. In my experience, there is nothing that people cannot be induced to care about. All that matters is that marketers take their brands seriously. vi). The pressures and barriers in market circumstances If brand relationships are one set of variables that determine what people buy, then the other set is what Butch and I called market factors (Hofmeyr and Rice, 2000). Market factors consist of all the external circumstances that amplify or dampen our natural propensity to spend against brands. Some create pressures which favour spending. Others constitute barriers which depress spending. Typical factors include: Sheer visual and aural presence (visibility, word-of-mouth, spending on broadcast advertising the night before shopping) Relative or absolute cost (temporary discounts such as coupons or special offers; the unaffordability of premium brands in developing markets) Distribution (some brands may struggle to get listed, or they may not be stocked by small retailers) Regulatory factors (not all markets are free; e.g., by law, South Korean retailers had to give local cigarette brands more prominence than imports) Contracts (marketing agreements e.g., Coca-Cola and Disney or, in business-to-business situations,, preferred supplier relationships)
19
Other factors (for example, Mom doesnt buy it, Its unavailable in our office, I cant get it at parties)
The relationship between market factors and brand relationships is subtle. Broadly, it has the following characteristics: The more committed someone is to a brand, the less relevant market factors will be: o o When faced with hurdles, a committed person will attempt to overcome them; e.g., they will pay more, or go out of their way, etc. Committed people do not need to be pushed to buy the brand. Thus, positive market factors make buying easier, but do not amplify it.
Market factors come into play in relation to brands to which people are uncommitted or indifferent. But the direction of spending will depend on the configuration of a persons brand relationships: o If a person has a brand to which they are committed, spending heads for that brand. Neither hurdles (which dampen) nor pressures (which amplify) help the brands to which such people are indifferent. If a person is uncommitted to all brands, then spending heads in the direction of whichever brands have a local market factor advantage whether due to pressures or to the absence of hurdles. An interesting case is people who have no brand relationships because theyre uninvolved. Their spending tends to go to the local market factor winner often the market leader.
In general: decreasing commitment to any brands (remember multiple brand commitment is possible) increasingly brings market factors into play. The interaction of the brand equity and market factors terms also impacts on how people will behave at each purchase. People who are committed to at least one brand are consistent and predictable: if possible, they buy the brands to which they are committed most of the time. By contrast, people who have low strength relationships to all brands behave in an erratic and unpredictable way at each purchase. This is because of the unpredictable way in which market factors fluctuate from purchase to purchase. Yet, as long as we know the relative strength of a brand in its market over time, the distribution of uncommitted peoples spending will be predictable over time8. One or two brief illustrations may help: Ive suggested that positive market factors play no role when there is a brand to which a person is committed. To illustrate the point, lets consider Coca-Cola and Pepsi. Both are well distributed and highly visible. Relative to most CSDs, therefore, both are characterized by strong, positive forces which amplify the likelihood of them being bought. Now consider someone who is committed to Coca-Cola: does Pepsis market strength amplify the likelihood that this person will buy Pepsi? If theyre indifferent to Pepsi, not at all. With this configuration of relationships i.e., commitment to Coca-Cola, indifference to Pepsi pressures which favour Pepsi will be ignored. By contrast, market factors will be very much in play when someone is indifferent to all CSD brands. If they have to buy a CSD, theyll follow the line of least resistance, buying whatever is most convenient or on special. And so, when it comes to Coca-Cola and Pepsi, they will be likely to drift back and forth between the two as a function of the day-to-day fluctuations in circumstances which favour first the one, and then the other. What predicts purchase propensity, is the configuration. vii). Going from brand relationships and market factors to brand sales Lets suppose that weve taken care of what should be in a competitive set; and lets suppose that we understand that higher market involvement will lift overall spending. In addition, lets suppose that weve
20
established the configuration of a persons brand relationships; and that weve measured the extent to which market factors amplify or dampen a brands chances of being bought. What should our brand equity framework look like? I started using the core of this framework in about 1992. In 2000, Butch and I published the framework in our book, Commitment-Led Marketing. The general idea is simple9: what people actually do in a market is a function of what they want to do coupled to what market factors direct. More formally: spending in a market is a function of brand relationships and market factors no matter how that market is defined. Notice that its the configuration of relationships that matters. And as you will see as we do more and more work of this kind, the function is not linear. Ive tried to capture the idea in the graphic below.
V: Predicting which brands people will buy
Brand Relationships Brand Characteristics - functional delivery - emotional resonance - personality/character - tie to deeper needs Attribute Association Brand Relationships Category Involvement Overall Evaluation Importance to Person Relative Appeal Theory of Commitment 0 - 100 .70 .09 Attitudinal Equity
Actual Purchases Market Factors Predicted Brand Share Actual Share of Wallet
.20
.18
??
.03
This brand would get more share if it were ?? easier to buy This brand picks up share from both brands where hurdles depress consumption
.78
??
.01
.01
?? This brand has as much presence as the one above, but it doesn t help Respondent > .88 Aggregate >.99
Brand Scores
Correlations:
The column on the far right represents how a person is actually distributing her spending across a relevant set of products or brands. We seldom observe this in single-source data. The panel folks think they see it but their percentages are based on assumptions about market structure and they dont have attitudinal data which backs up those assumptions. For instance, when someone saves money by buying a cheap laundry detergent, the panel folks have no idea where the savings go! We have attitudinal data and we often have claimed share, but we know that claimed share needs to be treated with caution. To flag our general ignorance therefore, Ive populated this column with question marks. The heart of the system is the first two columns. The attitudinal equity is our prediction of the way a person would spend her money under neutral market conditions. In this illustration, her natural disposition to spend is in the proportions: 0.20; 0.09; 0.70; 0.01.
21
We calculate these proportions using measures of the brains summary evaluations and casting them within a theoretic framework based on the butterfly-cusp catastrophe. The market factors column is a measure of the extent to which external factors amplify or dampen her natural propensity to spend. Barriers inhibit spending against some brands (e.g., brands one and two), and a strong brand relationship coupled to positive market strength diverts the resulting spending towards brand three. Note that there are also strong market factors driving spend for brand four (it could be Pepsi while brand three could be Coca-Cola). Yet it doesnt get any increased spend because there is no brand relationship. The scores for the two measures come from two sets of questions: Questions which measure the strength of brand relationships as outlined in terms of the butterfly cusp catastrophe: o o o An overall rating in terms of needs-values fit for each product or brand in a respondents relevant set (the TME question is fine); A measure of involvement for each product or brand in the respondents relevant set; A measure of respondent level category involvement
Questions which measure the obstacles against or the pressures in favour of brand spending; e.g., share of voice, affordability, distribution, etc.
Its important to remember that the theory of relationships allows people to feel strong attachments to more than one brand (as with me in the wine market). In fact, multiple commitment is a natural human condition. The disposition to spend as captured by the equity score is therefore a function of the configuration of brand relationships, not the just the absolute relationship. Moving to the third column: its our prediction of the way this person will spend her money, given the equity and pressure scores. Tell us how much she is spending and this column predicts how much each product or brand will get; i.e., in this case, 0.18; 0.03; 0.78; 0.01. The first two brands get less and the third gets more. The equity and market factor scores give us robust, valid variables against which to model. We know that if the scores go up, brand share will go up. We therefore know that creating models to determine what will make the scores go up, helps us determine what real-world marketing actions should improve brand sales. The scores are therefore ideal dashboard or scorecard scores. The range of potential marketing actions is enormous and takes us back to the early parts of this paper. There I made the following point: we cannot know whats impacting brand value creation unless weve inputted all the potentially relevant factors into our models. Here I make the point: we cannot know how to impact brand value creation unless weve got valid dependent variables against which to model. The equity and market factor scores will be our dependent variables. Separating the two measures in this way is necessary for accurate modelling because we need uncontaminated measures of each one if we want to be able to model what will impact on each one. So, for example, the equity score should be free of any hint of behavioural measurement. If it isnt, then any driver analysis aimed at telling us how to improve desire for a brand, would be contaminated by behavioural factors.
22
viii). Seamless scalability Ive argued that our approach should predict share of spend no matter how the market is defined; and even if its defined differently for different respondents within a study. Lets illustrate the point using the beverages market.
VI: The Scalability of the Method
Competitive Landscape CSD s Equity Score .26 Market Factors -.04 Share of Throat .22
Competitive Landscape Everything liquid Equity Score .33 Market Factors +.09 Share of Throat .42
.15 100 .58
-.08
.07 100
.25
+.07
.32
+.12
.70
.29
-.16
.13
.01
irrelevant
.01
.13
-0-
.13
! Coca -Cola gets more than its fair share: - Stocked in small shops - Available at parties - Constant visual presence ! Smaller brands tend to get less
! Coca -Cola gets less than its fair share: - Its expensive - It doesn t get into the home often enough - There are no vending machines in offices ! Tea and coffee get more than their fair share
Imagine that a respondent has been asked questions about everything they drink from tea, coffee, and tap water to carbonated beverages and fruit juice. In the course of the questionnaire, theyve answered our key equity and market factor questions. Suppose we want to model her behaviour within the CSD market. Ive created an illustration of what the situation might look like on the left hand side of the graphic. She buys four CSD brands regularly in the proportions: 0.22; 0.07; 0.70; 0.01. The third brand (it could be Coca-Cola), gets more than its fair share. It should only be getting 58%, but market pressures favour Coca-Cola when considered from the point of view purely of CSDs. Its always stocked. It gets bought for parties. And it has a very strong visual presence. Two brands get less than their fair share because they are neither as visible, nor as available. Now look at the right hand side, a hypothetical picture for the same respondent, but now including a broader range of beverages. Coca-Cola is still third on the list. But imagine that the first two beverages are tea and coffee and that our respondent is very fond of both. Pitting Coca-Cola against tea and coffee in her broader beverage world, depresses its equity score. Moreover, market factors amplify tea and coffee consumption because they are both cheaper and more readily available at consumption points; e.g., in the home, in offices, when visiting friends. And so, whereas Coca-Cola gets more than its fair share from this hypothetical person when its competing against other CSDs, it gets less than its fair share when competing against tea and coffee. Note that she brings additional volume into the picture when we include her tea and coffee consumption. Hence: the volume of Coca-Cola that she drinks remains the same. Its just that the share drops. When translated into share within the CSD market, her Coca-Cola volume amounts to 70% and is amplified by market pressure. In the beverages market, that same volume is 13% and is dampened.
23
Equity Surface Plot
Brands in an isolated space: satisfaction is the driver
Brands in a cluttered space: satisfaction loses traction
Share of Throat
10
Satisfaction Clutter
0
How do we engineer this seamless scalability? Take a look at the graphic above. It comes from a paper I gave at the INFORMS marketing sciences conference (2006). This is whats known as a surface plot. The vertical axis represents share of throat. The two horizontal axis represent elements that are used in the creation of the attitudinal equity. The plot tells us what a brands share of throat will be for a respondent as a function of two of the inputs to the attitudinal equity score. The inputs are: the overall rating a beverage or brand gets plus what I call the clutter [Note: clutter is calculated as a function of the number of brands that get similar or identical ratings]. Clutter is a relative rating. Its value depends on the values of other options. The more other options there are with the same rating, the higher the density will be. Now, imagine that there were 30 CSDs in this study. We plot each brands position on the horizontal surface for our consumer as a function of the brands overall rating and clutter. The surface with its plotted positions for each brand represents her mental landscape when it comes to CSDs. We can then read off the share that each brand will get by looking at the vertical axis. The total adds up to 100% share. Now imagine that we introduce the wider world of beverages. Our respondent is fond of both tea and coffee and its therefore likely that she will have given tea and coffee ratings which are at least as good as Coca-Colas. Coca-Colas overall evaluation doesnt change. But introducing tea and coffee increases the score it gets for clutter. We can see what will happen from the graph: the point representing Coca-Cola will move commensurately along the clutter axis and so attitudinal equity will be lower. Take tea out, and the clutter scores of both coffee and Coca-Cola will be higher shifting their points up and resulting in a seamless adjustment of their equities in the world of slightly lesser competition.
24
New shares are implied with each move. The above are two of the variables we use in the calculation of attitudinal equity. The fundamental drivers of equity are relative, not absolute ratings. They are: positive outranking, negative outranking, and clutter; and are borrowed from the world of operations research. Because theyre relative, whenever new competitors are added or subtracted, all products or brands may change their positions. The changes lead to changes in attitudinal equity scores and concomitant changes in predicted share. The system is seamless and dynamic; and readjusts whenever new options are introduced or existing options change their scores because a persons attitudes have changed. It is this that allows us to model seamlessly within one study. I imagine this world of equities as a swarm of dots on the surface. Each dot represents how a person feels about a brand in a market. Where a dot is, is a function of where all the other dots are because relative ratings drive the positioning. Since the positions of all the dots are linked in this way, when any dot moves, so do all the others even if only marginally. Change can come for a number of reasons. Introduce new dots by extending the market definition (to include tea and coffee, for instance); and the rest will adjust to accommodate the newcomer. Take dots away (tea, for instance), and the rest will swallow up the vacated space. Change the evaluation of a brand (she decides shes bored of Coca-Cola she really likes the new, lightly flavoured bottled mineral waters); and new relative evaluations lead to changes in all the brands. In this way, a persons feelings about brands in a market are represented by this constantly shifting set of dots in a mental landscape adjusting perfectly to predict new shares whenever change takes place. ix). A word about price Price is obviously a key characteristic of offers in markets. As with so much else covered in this document, there is a ton of literature on price. For our purposes, what matters is the relationship between price and purchase propensity. In a simple view, the higher the price, the lower the demand. As prices go up, the likelihood that a brand will be bought, falls. This, however, is not how people work. The reason is that prices may also signal quality and scarcity. And generally, people have quality preferences; and some people like to own whats rare. This creates a reverse logic in which higher prices may help to make things more desirable rather than less. For marketing research purposes, it helps to start with a simple attribute: value for money. This is a core characteristic of an offer and is about quality and price being matched. Broadly speaking, brands that are judged to be value for money have a better chance of being bought than brands which arent. But unqualified, the attribute value for money is ambiguous and predicts poorly. There are a number of reasons for this. Some people attach the attribute only to brands which marry the quality they want to a price theyre prepared to pay. These are then the brands they buy and call value for money. Others take the survey task seriously and marry the attribute to any brand which is fairly priced in terms of its quality across the quality and price spectrum. They then buy brands at a price theyre prepared to pay. And some cannot afford to buy the brands they believe are value for money, particularly in developing or emerging countries. They buy the brands they can afford. We can disentangle this mess by improving our attributes. Fairly priced, Worth what it costs, or Correctly priced for the quality are attributes which measure value for money in its conventional sense. A combination of a quality brand and a brand Im prepared to pay for tell us which brands appropriately combine a level of quality and price for a person. And affordable tells us which brands people can buy. To pick up expensive but value for money brands in poor countries we can use an attribute like is expensive, but fairly priced given the quality. In wealthier countries, an appropriate attribute might be is cheap, but correctly priced given its quality. An altogether more subtle situation involves people who buy down rather than up. In other words, they intentionally buy cheaper brands even though they can afford the premium brands. Why is this? Some might say: well, they clearly dont think the premium brands are value for money. I think the situation is more complex than that. Ask some of these people what they think about premium brands and they may well say that the premium brands are better quality and not unfairly priced. Yet they dont buy. Why not? The reason is: low involvement they just dont care enough about the category to want to own the best.
25
It perfectly describes my behaviour in the car market. If you ask me whether or not Toyotas Lexus range is fairly priced I will answer yes. But if you then ask me well, why did you buy a Corolla? my answer will be: because I just dont care enough about cars. So there are three important aspects to price. The first is the classical relationship between quality and price as captured in attributes like fairly priced or worth what it costs. The second is affordability, which has to do with the sheer amount of money that people have relative to the price. And the third is the value that is put on the category, which relates to how involved people are in thinking about the category. x). A word about validation There are many kinds of validation. Im interested in only one: when we predict how a person will distribute their budget in a market, that is what they should do. Confirming this is trickier than it sounds. One method would be to look at the next purchase and to treat purchase probability at the next event as a surrogate for share. But purchases are yes no events which may vary in terms of how much is purchased at each event. At each level of predicted probability, some would buy and others wouldnt; and the amounts they bought could vary relative to the average they buy. To validate our probabilities we would have to show that the aggregates for people at each equity score, corresponded to the probabilities that were predicted. Even if we did this, it would not be a strict validation because predicting aggregate purchase probability across a group of people is not the same as predicting how one person will distribute their spending over time. A second option would be to observe a persons spending for some time and then to compare the predicted brand share against the observed brand share. But a question is: for how long should we observe? Over time a persons dispositions may change. Give people too much time and our predictions are less likely to be accurate. Cut the time short and you make it hard for people to acquire a track record of spending. A far easier method is to ask people what they regularly do or intend to do. Any number of questions are available for this: e.g., constant sum methods (what brands are likely to fill your next ten purchases); past 24 hour consumption methods; purchase frequency or intention; and so on. The problem is that we know that people have a limited ability to answer these questions in ways which correlate with what they do. Memory seems to stymie people when they look back; and either lack of self-knowledge or circumstantial interference seems to stymie their ability to predict going forward. Our approach will be to recognisze that validations based on claimed behaviour are not as good a those based on real behaviour. So we will look for single-source validation wherever possible. But we will validate from every possible angle. Validations based on claimed behaviour will therefore also play a role. xi). Pulling it all together Ive identified three areas of measurement that are key to the brand equity framework when thinking about how whats in the mind translates into sales: The area of detail; i.e., everything thats stored in the brain about a brand in a dynamic web of images, memories, emotional impressions, etc. The area of summary evaluations; i.e., the overall sense of the brand that appears to be the brains shorthand method for storing brand impressions. The area of brand market presence; i.e., all the non-psychological factors that amplify or dampen the purchase likelihood for the brand.
In order to predict sales, we propose two key metrics for each brand in a study: Attitudinal equity derived from summary evaluations and expressing the appeal of the brand in a single number between 0 1.
26
Market factors a number which captures the extent to which market circumstances amplify or dampen purchasing. It varies between -1 +1.
Adding the two numbers together gives us a purchase probability for each person for each brand. In other words, it allows us to predict how a person will distribute their budget in a market, no matter how the market is defined. Both scores correlate remarkably well with behaviour; and their sum correlates almost unbelievably well. In other words, these are two robust and valid metrics that can be used as dependent variables when modelling marketing inputs and brand impressions. When our task is to model the effect of brand information (no matter what the source may be remember, there are three types) on the strength of the desire to buy the brand, then attitudinal equity will be the appropriate dependent variable. Similarly with attempts to model from the detail level of brand impressions, to the overall evaluation of the brand. What the market factor measures do, is assist with tactical and strategic planning so that the brands chances of being bought can be increased independently of what people think of it. Historically, however, the greatest brands have been those that combined a high equity with an appropriate market presence. Ive argued that establishing valid and reliable measures of the detail about brands is one of the most important things in any project design. The most common way of doing so currently is through attribute association questions. In my view, we need better measurement in this area. Ive suggested that Synovate Censydiam have a key role to play in this regard. Measurement of the equity score will be based on the butterfly catastrophe theory of brand relationships. Ive said that we need valid economy of measurement. In my experience, this theory successfully directs us to such measurements. Although the theory requires four independent variables, we only need two measures. They are: An overall evaluation of the brand in terms of the respondents needs and values in the market in question; i.e., brand satisfaction. An overall measure of the importance of the brand to the person. In other words, a measure of the extent to which the brand matters to a person.
The remaining dimensions (ambivalence and competitor appeal) can be constructed from combinations of the above two measures. At the category level, we need a measure of category or market involvement. These are the criteria our framework fulfils: All summary brand scores, in particular the attitudinal equity and market factors scores have a very strong correlation with actual behaviour. The slope of the relationship between behaviour and the scores is steep meaning that there is a significant pay-back for an improved score. The equity score is a purely attitudinal number. o o Behavioural measures (e.g., most often brand) are partly influenced by market factors, therefore to establish the effect of communications on the strength of a brands appeal, we need a purely attitudinal dependent variable
27
Notes 1. For a comprehensive review of different approaches to branding and brand equity, see Kellers and Lehmanns (2004) Brands and Branding: Research Findings and Future Priorities, a review conducted for the Marketing Sciences Institute. 2. Note that while the brain stays open to peripheral stimulation, we are only conscious of an over-ride when strong and unusual stimuli come along. Mostly we go on our way, taking in information through the periphery with very little if any awareness. This is why brand communication can take place without explicit awareness or conscious recall. Its also partly responsible for whats known as low involvement processing; i.e., the absorption of brand information without thought. Unlike some, we do not believe that marketing communication works either through low or high involvement processing. It is a matter of degree. The topic is important enough to deserve a separate paper. 3. Occams razor is the popular term given to the principle that theory construction should involve the fewest assumptions possible. It is named after the English logician and Franciscan monk, William of Ockham (c. 1295 1349). Others before him had expressed the same idea e.g. Thomas Aquinas (c. 1225 1274): If a thing can be done adequately by means of one, it is superfluous to do it by means of several (quoted in Wikipedia). In other words, if we can get away with a language of the brain, drop the language of the mind for scientific purposes. Although it sounds like a reductionist approach to science, it isnt. 4. Emotions research is advanced. It is a complex area. I hope that I have done justice to the main ideas. Readers who may doubt the power of an evolutionary perspective to account for emotions should read Ekman et al (Editors, 2003) and Levenson (2003). 5. This may seem like a throw away line, but its quite important. The point is that people actively structure their experience in the course of a day; and that active structuring reduces exposure to easily manageable proportions. In spite of the very information rich environment in which we live, our normal state is not to feel overwhelmed. Other things may overwhelm us from time to time, but stimulation seldom does. 6. Questions: a. Overall rating: Please indicate your overall feeling about each brand. Use a 10-point scale where 1 means you think the brand is extremely poor and 10 means you think it is outstanding. Scale: 1 = extremely poor 10 = Outstanding
b. Brand involvement: Please read the following statement and indicate whether you agree or disagree with it. Statement: <Brand> has a great deal of personal meaning to me Scale: 1 = Completely disagree 7 = Completely agree
c. Category involvement: Please read the following statement and choose a number which reflects your feelings. Statement: Deciding which brand to buy is Scale: 1 = Completely unimportant 7 = Extremely important
7. This is why we stress the importance of research to establish what drives perceptions of attributes like quality or trust in different product categories. Our industry doesnt know enough about this yet. It is also why questions about attributes like quality or trust should be contextualised when we do know their dynamics. In banking, for example, instead of the bare inclusion of an attribute like trust in the attribute association, ask: To what extent, if at all do you trust <bank> to care more about service than about profits? 8. There are two points implicit in this paragraph. The first is: the weaker a persons relationship with all brands or the more ambivalent a person is about brands, the more unpredictable their behaviour will be at
28
each purchase occasion. Predicting when and why behaviour becomes unpredictable is as much an advance in understanding as being able to predict precisely and its a considerable advance over what we have now, which is poor prediction. The second point is: just because we cannot say what a person will buy at each purchase occasion does not mean that we cannot predict their purchase propensity over time. Take a person who is equally likely to buy one of three brands at each purchase occasion. Next time they buy we may not be able to predict which of the three brands will get bought. Yet, we should feel confident predicting that they will buy each brand a third of the time, over time. 9. This formulation harks back to Fishbeins original formulation of behavioural choice as an outcome of a linear combination of attitudes to social norms and evaluations of the behaviours outcomes in relation to personal goals. Fishbeins model is an optimising model in which all possible outcomes are evaluated according to a weighted additive function. I dont think this replicates how people make decisions. Im of the satisficing rather than the optimising school of human behaviour, initiated by H.A. Simon but perhaps best represented by the heuristic approach of Gigerenzer (1999). Im aware that this opens up a considerable literature. Im also aware Ive expressed this point quite clumsily.
29
References
Aronson, Elliot (1988): The Social Animal (Fifth edition), W.H. Freeman and Company Bolton, Ruth (1998): A Dynamic Model of the Duration of a Customers Relationship with a continuous Service Provider: The Role of Satisfaction, Marketing Science, 17:1 Cacioppo, John and Gary Berntson (1994): Relationship between Attitudes and Evaluative space: A Critical Review with emphasis on the Separability of Positive and Negative Substrates, Psychological Bulletin, 115:3 Crick, Francis, and C. Koch (1990): Towards a neurobiological theory of consciousness, Seminars in the Neurosciences (2). Ekman, Paul, Joseph J. Campas, Richard J. Davidson, Frans B. M. de Waal (Eds. 2003): Emotions Inside Out, Annals of the New York Academy of Sciences 1000. Gigerenzer, Gerd and Peter Todd (1999): Simple Heuristics that Make Us Smart, Oxford University Press. Gordon, Wendy (2006): Out with the new, in with the old, International Journal of Market Research, 48:1 Hofmeyr, Jan (1990): The Conversion Model A new foundation for Strategic Planning in Marketing, New Ways in Marketing, 3rd EMAC/ESOMAR conference, Athens (2001): How people consume advertising and its implications for marketing, Advertising Research Foundation Conference, Philadelphia (2006): A Unified Framework for the Measurement of Brand Equity, paper presented at INFORMS Marketing Sciences conference, Pittsburgh and Butch Rice (2000): Commitment-Led Marketing, Wiley and Sons Keller, Kevin and Donald Lehmann (2004): Brands and Branding: Research Findings and Future Priorities, monograph for the Marketing Sciences Institute Kuhn, Thomas (1970): The Structure of Scientific Revolutions (Second edition), University of Chicago Press Levenson, Robert W. (2003): Blood, Sweat, and Fears: The Autonomic Structure of Emotion, in Emotions Inside Out (Eds. Ekman et al) Mandelbrot, Benoit (2004): The (Mis)Behaviour of Markets, Basic Books Pinker, Steven (1999): How the Mind Works, W.W. Norton and Company Shannon, Claude (1948): A Mathematical Theory of Communication, The Bell System Technical Journal Schacter, Daniel (2001); The Seven Sins of Memory, Houghton Mifflin Company Ward, Lawrence (2002): Dynamical Cognitive Science, The MIT Press
30
About the authors Jannie, formerly a doctor of religious studies, is arguably the worlds leading expert on brand commitment. He joined Synovate in April 2006 as Global Head of Insights for the Brand and Communications practice. Jannie has worked with leading companies around the world including CocaCola, Levi Strauss, Nokia, American Express, Procter & Gamble, IBM, Intel, HP and McDonalds. He is the inventor of The Conversion Model, which has now been used in 8,000 projects in over 120 countries. Ged Parton joined Synovate in January 2006 as CEO of the new Brand and Communications practice. For the past 5 years Ged has been running research businesses in New York and London. Prior to entering the research arena he worked in advertising and consultancy and was also Managing Partner of Ogilvy, London.
31
32

Synovate White Paper - Brand Value Creation - Communications and Equity

Uploaded by

Copyright:

Available Formats

Synovate White Paper - Brand Value Creation - Communications and Equity

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Synovate White Paper - Brand Value Creation - Communications and Equity

Uploaded by

Copyright:

Available Formats

Brand Value Creation, Communications and Equity

Copyright Synovate 2006

Brand Value Creation, Communications, and Equity

To start: Some words and definitions

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

The Brand Equity Framework: Understanding communications

Copyright Synovate 2006

I: The nature of brand exposure

Total Potential Exposure

Copyright Synovate 2006

Copyright Synovate 2006

II: What gets into the mind?

Copyright Synovate 2006

Copyright Synovate 2006

The Brand Equity Framework: Predicting what people will buy

Copyright Synovate 2006

Copyright Synovate 2006

III: How what gets into the mind, is processed

Copyright Synovate 2006

Copyright Synovate 2006

IV: A Theory of Brand Relationship based on the Butterfly-Cusp Catastrophe

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

V: Predicting which brands people will buy

Copyright Synovate 2006

Copyright Synovate 2006

VI: The Scalability of the Method

.15 100 .58

Copyright Synovate 2006

Equity Surface Plot

Brands in an isolated space: satisfaction is the driver

Brands in a cluttered space: satisfaction loses traction

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

Copyright Synovate 2006

You might also like