Paul Feldwick is one of Adlands big thinkers. He has worked in advertising agencies for over thirty years, is an inductee the IPA Effectiveness Awards Hall of fame, and is widely recognized as an authority on how advertising works. This transcript from a presentation he gave at a Thinkbox event a few years back still remains one of the most widely read articles on this website. Here he argues why and how any organisation that wants to create more effective advertising will have to change some fundamental assumptions about advertising, communication and creativity. Paul went on to develop some of these ideas further in his celebrated book ‘The Anatomy of Humbug’.
Neuroscience and psychology; what do we do with this kind of new learning about the brain?' Because it seems obvious that there are lots of major implications here for how we create and how we judge advertising. And the main drift of a lot of these ideas seems pretty clear.
Here for example is a headline that was written by Raymond Snoddy in Marketing:
'Ads must aim for the heart, not the head.' And as headlines go I think that sums it up pretty well. In fact I think there are a lot of people, particularly outside of the ad business might be forgiven for wondering why we think that's particularly newsworthy. You might well think that, actually, it sounds like a no-brainer. The question is, how many organisations, advertisers and agencies, have really started acting on these ideas? And I suspect, I'm sorry to say, that the answer is not very many. The implications of Raymond Snoddy's headline are really quite radical for the ad business. But I think that the processes and the concepts that advertisers and their agencies are using probably really haven't changed very much.
So why do we seem to be so resistant to acting on these new learnings? The answer I think is we can't really respond to a new theory until we are prepared to acknowledge and to criticise our existing theory.
The problem is, most people in marketing and advertising (and I worked in an ad agency for over thirty years) don't really believe that they have a theory as such. They think that what they do is just common sense. But just as John Maynard Keynes wrote back in the thirties about economics, 'practical men, who have never knowingly been subject to any intellectual influence, are invariably the slaves of some defunct economist' - In other words, people always have a theory, whether they know it or not, and that theory must have started from somewhere, probably somewhere in the past.
If you don't know what your theory is, or where your theory came from, then actually you can't criticise it because you can't even discuss it - and in that sense you're going to remain a slave to it. That's why all this wonderful stuff about the brain and the emotions is going to remain just something that we hear about occasionally at conferences until we are prepared to make the effort to understand what our current theory is, where it came from and the ways in which it's built in to our language, our processes and our organisational cultures. And I think that if we're prepared to do that, then we'll find that we need to challenge some of our existing ideas about communication and about creativity as well. Those are the ideas that I'd like to sketch out for you here, so let's begin.
The pantyhose experiment
One fine morning in the late 1970's two men set up a stall outside Meyers' Thrifty Acres, a discount store in the pleasant university town of Ann Arbor, Michegan. But they weren't market traders; Timothy Wilson and Richard Nisbett were psychologists who were conducting an experiment.
On the table in front of them they had four pairs of pantyhose, labelled from A to D and displayed from left to right. As women passed by they were asked to examine the four pairs of pantyhose and say which ones they preferred, and over the day a clear statistical pattern emerged. Pair D, was the most preferred and pair A, the least.
Now one of the researchers' hypotheses had been that they would find an order effect and this proved correct. The women were strongly biased to the pantyhose on the right-hand side of the display, and the researchers knew that this must be an order effect because all four pairs of pantyhose were in fact, identical.
The more interesting finding to me was this, the respondents were also asked to give reasons for their choices and they all had no difficulty in doing so. They talked knowledgeably about the product's quality, the superior texture of their chosen pair; they were sheerer, better finished, etc. And, as the day went on, Wilson and Nisbett even started to challenge the respondents and asked whether they weren't quite sure that they weren't being influenced by an order effect. The women all looked at them as if they thought they were crazy.
Well, I came across this a while ago in Timothy Wilson's very fine book about the adaptive unconscious, Strangers to Ourselves, in which he draws this conclusion: 'The causal role of conscious thought has been vastly overrated'. Indeed, it is often a post hoc explanation of responses that emanated from the adaptive unconscious.' And this experiment shows with rather shocking clarity the huge difference that can exist between the way we think we choose things and the ways that we really choose. We are quite capable of feeling a strong preference for something for reasons which we're totally ignorant of, but we are good at disguising this to ourselves because we automatically tend to create an apparently rational cover story, which we then believe in.
Now when I read about this experiment, it reminded me of the years in which I used to carry out research groups about things like tea. I'd do groups with PG Tips users, who would all talk with great conviction about how PG Tips was stronger, more colourful, more flavourful than Tetley and how Tetley was mere dishwater. And then I'd go to a Tetley user group where the Tetley users would claim exactly the same product advantages for Tetley, while deriding PG Tips - usually as 'southern piss'.
Meanwhile, we knew that in blind taste tests most people were completely unable to tell the difference. So you know, we're used to the idea that the rational product claims that people gave us for their brand preferences, are really entirely spurious, they're mere projections. But the strength of those brand preferences is real enough. It's real enough to be worth billions of pounds or dollars to the brand owners. I mean if one looks quickly at the history of PG Tips, the first thing that should strike us about this is about how very stable its brand share remained over more than thirty years. It's easy to take that for granted, but it's actually rather remarkable. I mean that from 1964 onwards, as governments came and went, hemlines rose and fell, the world generally changed almost unrecognisably, PG Tips continued to be brand leader. And not only that, it did so in the face of other brands that were advertising and discounting heavily, of product innovations and the explosion of retailer own brands in the UK that undercut PG Tips more and more on price. So even today you'll find PG Tips is on sale at £1.55 for 80 against Sainsburys own brand at £1.15. That's a 35% premium.
So, brands like this offer a more reliable promise of future cash flows through higher margins and resilience to competitive prices as well as the potential for growth into new products. And if you work that out in financial terms as net present value, then that is the financial value of a strong brand. It's very real.
So, how is that kind of brand preference created and how is it maintained?
The pantyhose experiment and the sort of things that people say about tea both suggest that this may have much less to do with the actual product performance than we sometimes like to imagine. That would make sense because in most categories products are pretty close to parity, or it's sufficiently complex that the choice between brands can hardly ever be made on purely rational grounds. We are also much more suggestible than we think. For example, I'm told that in blind tests most drinkers reject Stella Artois because it's too bitter, but when people reframe this product experience within the particular network of beliefs and associations that they have about the brand, it remains one of the biggest selling lagers in the UK.
So if these preferences don't come just from the product, what about the role of the advertising? And let's look again at the case of PG Tips. In 1955 the brand was number four in the British tea market. Then on the opening night of commercial television those homes who were watching, that's probably only about a million homes I guess, saw this...
Within eighteen months of this campaign PG Tips had become brand leader and it remained so for over 30 years. There's an IPA Grand Prix paper from the early nineties that shows how it was the effect of the advertising that enabled this brand to maintain its price premium and its volume share in the face of increasing discounting from retailers and other brands. Here's another ad that you may remember from 1972.
Now what's going on in commercials like these?
Certainly there are some statements being made about the tea, but they're very generic. I mean good tea, fresh tea, when a good cup of tea really counts and so on. Probably every other brand of tea throughout the last four decades has been saying more or less the same things. So what is it that's making the difference? Well of course it's the monkeys, stupid.
Somehow 30 seconds of entertaining nonsense leads to a situation where people not only choose this brand but will pay 35% more for it. But somehow it seems we're not very comfortable with this. Because we all like to believe deep down that we really choose our tea on rational product-based grounds. So I think this is why, from a very long way back, ad people have invented a story about how advertising works that goes something like this.
The successful ad is like a salesperson. It has to get your attention. It then gives you some information about the product and a refinement of this is that it has to give you a single proposition or a message about the product. It has to do this in a way that makes you remember it and the when you come to buy tea, or whatever it is, you will remember this message and it will persuade you to choose that brand.
Now that story can just about be made to fit the Chimps campaign. You could say that the chimps get your attention by being different and funny, then they slip in this message about it being the best-tasting tea while you're watching and then your choice of PG Tip will be influenced because you will remember that PG Tips is the best-tasting tea when you go shopping.
Now if you believe that story, it has big implications for how you develop advertising. It means that first of all that the message or proposition will be the most important thing in the ad, so first of all you have to choose a message that is strong and convincing. Then the role of creativity in the advertising is just to support the transmission of that message. That means it has to get the viewers' attention, and you can measure this (you sometimes call it standout, or cut-through or impact) by asking them afterwards if they remember seeing the ad. And you then have to get the message into the viewer's head and you can measure this too, by asking them afterwards what they thought the message was.
Now this underlying model is taken so widely for granted that I think many people involved in marketing and in advertising find it really impossible to question it. It's built into client systems with titles such as ABC. Attention, Branding and Communication. It's supported by quite a lot of academic work, even though that's one remove from what we do, like Petty and Cacioppo's Elaboration Likelihood Theory. And perhaps most crucially, it's supported by both planners and creative people in the insistence that they have for creative briefs that look like this: 90% of agencies' own creative briefing forms have some sort of pattern which is like this model template from the APG's current book "How to Plan Advertising", where you can see there's a proposition, and then there is support and the word "message" is being used again.'
Now, I have further support that this is the case, although it could be more up-to-date but Hall and Maclay's 1991 research into 'How Advertisers Think Advertising Works' confirms that message transmission lies at the heart of the dominant model in people's minds and is almost universally regarded as an essential part of how advertising works. So as far as I can see, this is still our dominant story about how ads work and this story structures our systems, our processes and our language.
There's just one little rub as Bill Bernbach might have said - this is, I believe, the wrong story. It's wrong for three reasons.
- It's wrong because it doesn't actually fit with most ads.
You can just about torture your observations to fit it and this is what most people do. But to be honest, if you just look at successful brand-building TV commercials, it's usually pretty obvious that there either is no message in any meaningful sense, or if there is a message, it's not really the important part of the ad. What people are experiencing and responding to is a wealth of material, visuals, music, dialogue, timing, colour, entertainment, emotions. We've got used to somehow side-lining all this as if any ad could somehow be summed up in a couple of words though it must be obvious that if you could do it all in a couple of words, you wouldn't need to do all the rest of it.
- It's wrong because it doesn't fit the science.
And, of course, there are lots of references: many of you will know what they are from Damasio to Robert Heath and everything in between, we're accumulating more and more evidence about how we choose and about how we learn and also about how advertising works. We now know that ads can influence our behaviour without getting our conscious attention, and without us being able to consciously remember them. We know that more ads are more effective if they're not consciously processed. And we know that ads that appeal to the emotions tend to be more effective.
- This model is wrong because it doesn't work.
I worked in a very successful agency for over thirty years but looking back on it, the successful campaigns that we produced happened not because of this message model, but in spite of it. And I think companies that apply this message model make it far more difficult than it needs be to create successful brand-building advertising. In fact if they apply it rigorously enough, they can make it virtually impossible. I will mention no names.
Well if all this is so, there are some good questions.
- If what I call the message model is wrong, why is it so appealing and why is it so persistent?
- Where did it come from in the first place and why do we have it?
- And what other stories could we be telling ourselves that might be more useful?
Let me start with the first question. I think it appeals because it fits a general cultural myth in our society of rational man, that we make better decisions as a result of conscious thought. Now interestingly that turns out frequently not to be the case, but it's something that we rather like to believe.
It also fits very well into rational organisations where order, analysis and control are always assumed to be the best ways of getting things done. It makes it easy for us to have a process for creating ads which resembles the division of labour on a production line. So we have the strategy, and then we have the idea and then we have the execution and so on. And I think we also like it because it positions the ad business as a bunch of honest salesmen. Rosser Reeves liked to boast that 'there are no hidden persuaders, advertising works openly in the bare and pittiless sunlight.' And in fact this image of the adman as the salesman tells us a lot about where the message myth originated from.
It derives from a very early theory that equates advertising to face-to-face selling. Back in 1903 John E. Kennedy asserted that advertising is salesmanship in print. And that famous formula AIDA, was invented in the 1880's to train the sales force of the National Cash Register Company. It is indeed a useful formula for foot-in-the-door, face-to-face salespeople. It's even a good formula for a certain kind of advertising, mainly what we would today call Direct Response Advertising.
When Claude Hopkins wrote his best-selling 1923 book Scientific Advertising, he based his theories entirely on what he had learned from mail order. For Claude Hopkins an ad was all about information and the more of it the better. Then forty years later Rosser Reeves updated Hopkins's theory but with one significant change. Instead of Hopkins's long copy full of many facts, Reeves claimed that 'advertising is the art of getting a unique selling proposition into the heads of the most people at the lowest possible cost. He also claimed that, 'the consumer tends to remember just one thing from and advertisement, one strong claim or one strong concept.' Now as far as I know, there's no evidence for this at all, but this idea of Reeves seems to have started our collective obsession with having a single-minded proposition or idea in our ads.
In any case the message myth is based on a more fundamental error. It was Hopkins who asserted that, 'the sole purpose of advertising is to sell, it is not to keep your name before the public, it is not to help your other salesmen.' To which I can only reply, 'why not, Claude?' Because advertising that builds brand value, I think exists precisely to keep your name before the public and to help your other salesmen. That is how it creates competitive advantage.
As Stephen King said 40 years ago, 'advertising exists not to create sales so much as to create sale-ability.' So if the brand advertising isn't primarily about the message, and if the purpose of our advertising is creating sale-ability or the efficiency of selling rather than the sales themselves, then what alternative stories to the message myth might we tell ourselves?
Here I'd like to offer you two ideas which I personally think might be a lot more useful. One is the idea of 'associations', the other is the idea of 'relationships'. Associations are connections in the brain that link together ideas, images and feelings. The point about associations is that they don't have to be conscious, they don't have to be verbal. It's actually a very ancient idea. It was very popular in the eighteenth century. But Damasio's work on engrams and somatic markers I think has given it a new scientific respectability by at least giving us a working hypothesis as to how this happens.
Back in 1903, before the time of Claude Hopkins, Walter Dill Scott published a book called The Psychology of Advertising. Scott thought that advertising works by creating the right kind of associations for the brand. Scott also knew that advertising can influence people's attitudes to brands without them being able to consciously recall seeing the advertising itself.
One young lady asserted that she had never looked at any of the cards in the streetcars in which she had been riding for years. When questioned further, it appeared that she knew by heart every advertisement appearing on the line and the goods advertised won her highest esteem. She was not aware of the fact that she had been studying the advertisements and flatly resented the suggestion that she had been influenced by them.
Well Scott's ideas were effectively buried by the later obsessions with ad recall and message transmission, but I think we can usefully revive this idea of associations today because it's a simple everyday term which reflects a lot of what we now know about implicit learning, engrams and all of the new science of advertising.
Let's look at one successful brand-building commercial from back in 1983, 'George the Bear’.
Now I find it impossible to make sense of that commercial using a model of message transmission. There isn't a message in that. But the positive feelings that it generated through the music and the images are very effectively associated with the brand and this association may not be conscious, it may not be verbalizable.
Here's another commercial. This actually does contain some information about the product. But ask yourself how important is that information in explaining the success of this campaign?
There obviously is a message here, it's about insurance and I guess this offers some kind of rational level of reassurance. The fact is though that most brand users were already aware of this benefit before that advertising broke. And what this campaign did that was really powerful (again it's an IPA Award Winner), was to replace the previous association of the brand with the rather old-fashioned Alan Whicker with a new set of associations with a young, fashionable comedian, to sophisticated entertainment, to a contemporary style of humour and to the good feelings that you have when you watch that commercial. And one result of that was a significant growth in numbers of younger cardholders for example. So I think talking about associations instead of messages take us a long way towards making more sense of brand advertising, but we can go further.
My second idea comes from this guy, Paul Watzlawick.
He was a psychotherapist and so one of his key interests was in how human communication works in the context of personal relationships. He has nothing directly to say about advertising in this book, but I think its key ideas can be very powerfully applied to advertising and how it influences the relationship between the brand and the consumer. I want to talk about three of Watzlawick's ideas, or axioms as he calls them. They are these:
- It is impossible not to communicate.
- Human communication is both digital and analogue, and
- All communication is about relationship as well as content.
I'll explain what each of those means, I hope.
It is impossible not to communicate.
Just saying nothing doesn't mean that no communication occurs. If you say 'Good morning' and I remain silent, you might find that a powerful communication. Whatever we do or whatever we don't do, intentionally or unintentionally, we are sending out signals for others to make sense of. In fact everything we do then is communication. Psychologist Albert Mehrabian calculated that between 55 and 95 percent of all human communication is non-verbal. That means not just tone of voice and gesture, but the clothes we wear, the place we stand and many other contextual clues. So everything about a brand communicates and every aspect of a commercial communicates. Not just the words and not just the message.
Digital and analogue communication
Watzlawick secondly makes this distinction between what he calls digital and analogue communication. You have to remember that this was back in 1967, the word digital didn't have many of the connotations that it might have for us today. He was really thinking about computers. And the best way to explain his meaning I think is to consider the metaphor of a digital clock, compared with an analogue clock. One of them gives you an unambiguous replicable piece of information an abstract number. The other expresses the time through spacial, visual gestures that are open to interpretation. So we can build up a sort of pattern of comparisons.
Digital communication is precise, it's either verbal or numerical. It's logical, conscious explicit and intellectual. We couldn't do without this clarity and precision in many contexts, but it's not the only kind of communication that we do as humans.
Analogue communication might be verbal too, but it could also be visual, audible, gestural or tactile and the power of an image, or a poem or a story or a hug or a piece of music is that we're not immune to it. It's non-deniable. We experience it to a large extent through our unconscious and it implies rather than states. It has a concrete and a visceral power that's lacking from the dry digital mode of communication. And its strength, and also its weakness in some senses, is that its meaning cannot be unambiguously defined. This is an important point. Analogue cannot be translated into digital.
Now I believe that advertising also communicates in both digital and analogue modes. And I'll say later why I think the analogue mode is probably the most important.
Here's one campaign, very successful - Felix - which communicates almost entirely in an analogue mode.
So there's a press campaign, and here's a TV commercial from a different brand.
In both these cases it's very hard, I think it's actually impossible to analyse exactly how and why they make their effect. That doesn't stop us trying and people often come out with their own answers but I think really it it's what defies analysis.
But the visceral power of analogue communication sits uncomfortably with the fact that most organisations strongly privilege digital communications. In a business meeting we are usually expected to offer facts and figures, logic and analysis; often for very good reasons. But when discussing advertising this tendency strongly biases people to ignore or undervalue the importance of the analogue mode of communication.
Human communication always has two aspects.
Watzlawick's third axiom is that human communication always has two aspects. There's its content and then there's the way in which the communication helps to construct a relationship. So again, if I remarked, passing you in the street that it's a bit colder today, I am probably not telling you anything you didn't already know. What I'm really saying is something a bit more like, 'we are still friends' or at least 'we are not enemies.' If I passed you in silence, that would have a different effect. Now my choice of content is not random. But this piece of communication is not about the content, it's about the relationship. And analogue communication, which includes image, gesture and tone of voice is especially crucially in influencing relationships. That's why if we see the task of advertising as building and strengthening a quasi-personal relationship between the consumer and the brand, this is going to happen predominantly through the analogue mode.
If all this is getting too complicated, it was all explained to me many years ago in far simpler language by my first boss Martin Boase, who used to put it rather like this:
We believe that if you are going to invite yourself into someone's living room you have a duty not to shout at them or bore them or insult their intelligence. On the other hand, if you are a charming guest and you entertain them or amuse them or tell them something interesting, then they may like you a bit better and then they may be more inclined to buy your brand.
Thirty years ago I thought that sounded very plausible, but not quite clever enough to be true. I now think it's probably the most useful and truthful thing I've ever heard about advertising and I think that the research that we're getting from neuroscience and the theories of Paul Watzlawick entirely support what Martin said.
Finally, I think this leads us to a different way of thinking about that problematic word 'creativity'.
The old model which is based on the message myth positions creativity essentially as a set of tricks for getting attention and increasing memorability: so that leads towards advertising which fights for the viewers' attention. It leads to advertising which is creative in a purely intellectual way, often inventing an elaborate scenario in order to dress up an essentially intellectual idea. And such advertising very often fails in the marketplace because people in real life don't process it with the full attention that would be required to decode the message. And even when they do, the message in itself is rarely a motivation to action.
Successful and truly creative ads, I think work in quite a different way. If we pretend that advertising is predominantly digital, then we'll feel justified in thinking of any ad as being reducible to an intellectual, verbal construct, a message or a proposition or an idea. But if we understand that the important relationship building communication is taking place through the analogue mode, then we should really change our focus away from this abstract digital idea, back to the visual, visceral power of the entire advertisement; its colour, movement, music, timing and every detail.
TV ads work I believe, through the analogue mode where emotional engagement is created through the actual sights and sounds of the commercial. And great ads, like great works of art don't have to be very obviously original. Like great works of art, they don't have to have much to do with a single reductionist idea. They work as aesthetic wholes.
I will show you one now that I am particularly fond of. This was actually done in Argentina, shortly after the crisis following the devaluation of the peso, so it had a lot of resonance at that time. It's an ad for soup and I think if you look at this ad, and if you feel about it the same way that I do, you'd have to admit that it's not good because it's particularly original, nor is it good because it has a particularly clever idea behind it: it is good because it is a wonderful bit of film-making. It owes everything to the way that it's been executed.
Well, as you can tell I'm very opposed to this idea that ads are somehow based on the creative idea. I don't think they are. I don't think there is a creative idea behind a painting by Kandinski, or Bach's B Minor mass, or almost anything else you can think of. A picture may be in some sense original but I don't think that that is what makes it a great work of art.
Of course in most business contexts creativity is still a rather suspect word and art is even more so. They stand for everything that the digitally minded organisation struggles to understand and often therefore rejects. But by doing so, organisations can be throwing away one of the most powerful tools available to them for creating competitive advantage. As Bill Bernbach of course said, "Is creativity some obscure esoteric art form? Not on your life. It's the most practical thing a businessman can employ".
Bernbach understood that creativity was essentially about artistry, the tiny details which obey no logic and can submit to no analysis but which make all the difference to our emotional responses. He took as an example, the famous and successful mail order ad for a correspondence course that ran for years from the 1920's.
'They laughed when I sat down at the piano'.
Bernbach remarked "What if this ad had been written in a different language? Would it have been as effective? What if it had said, 'they admired my piano playing'. Would that have been enough? Or was it the talented, imaginative expression of the thought that did the job? That wonderful feeling of revenge.
Suppose" Bernbach continued, "Winston Churchill had said 'we owe a lot to the RAF' instead of 'never was so much owed by so many so few', do you think the impact would have been the same?" He also said "logic and over-analysis can immobilise and sterilize and idea. It's like love. The more you analyse it, the more it disappears." And I think the reason for that in Watzlawick's language is that you cannot translate analogue communication into digital.
Let me leave you with seven propositions. And I believe that if you put these into practice you would achieve more effective brand building advertising and avoid a great deal of the agony that is often involved in producing it.
- Define the advertising goal as building saleability.
- Stop talking. Mostly about messages.
- Start talking about associations and about relationships.
- Recognise the power of analogue communication.
- Shift your focus away from the abstract message or idea to the advertisement as a whole.
- Resist the urge to over-analyse and over control.
- And if all that fails, just concentrate on being a 'charming guest.