This is an essay I wrote a while ago. It is in draft form but contains some useful ideas

Marketing theory is not in a good place. On the one hand we have data analytics, whose practitioners are driven by hard functionality. This crushes creative thinking, for which the only credible theoretical opposition to this comes from behavioural science. That, however, is not always on the side of the creative angels – it is began as a media strategy and is often used as a justification solely for timeliness. It offers brilliant analysis of why people did the things they did but little help with what they might do next, simply because of its accurate recognition that reality is contextual. It is in this maddeningly vague grey area, I argue, the creativity works, and fights its battles without any secure theoretical basis.

So marketing, I argue, is in a barren place, and we see constant evidence of this in agency output.

How do we move forward? Over the years, my interest in the processes and structures of storytelling – which began during my early career as an academic – led me to look more closely than most creatives ever would into the theoretical assumptions made by our use of data. As the head of an in house agency I got to suffer first hand from many statistical misinterpretations of my work. These were often admittedly apologetic, since the work in question was some of the most effective ever run! This led me to believe that data could surely be used better to support creative thinking. It led me to look more closely into statistically analysis, and this led me to an interest in quantum mechanics. 

This science originally interested me because it appeared to be in a parallel world to the functional marketing practise I was struggling with. Scientists had been doing what worked without the burden of interpretation, and this was beginning to be challenged. What’s more, it seemed to operate with some strikingly similar assumptions about the importance of context. But quantum mechanics has been around for over 100 years and has absorbed the attention of a succession of brilliant mathematicians. It offers insight into reality at a different level. 

What really interested me, looking at it from my rather unusual perspective, was that it  makes one assumption about how things work which is both buried in behavioural economics, and also offers us a way to deepen our understanding of that nascent science. That assumption concerns the way that probabilities change over time. 

You can see why mathematicians and statistician heave tended to regard this as their property. More recently, though, philosophers have started to take an interest, and some understanding is beginning to develop into what this often startling knowledge means. 

Let’s be clear, quantum mechanics matters. The equation that formulated it and the numerical values it takes as givens have driven the development of the technology you are using to read this. It is the technology that drives the machines that produce the data we use. Understanding how and why it works gives us a much better understanding of the way we use data, and its limitations, and it shows us where and how behavioural science is right, and where it might take us next.

The ways scientists talk about and use quantum mechanics don’t always help us here. It is arguably the most useful and least understood theory in the history of science. It has informed the development of almost all current technology.  But few have attempted to explain why quantum science is so powerful, how it works, or what it means. This is why it seems dauntingly obscure and intellectual.

This reveals some game changing and authoritative insights. 

The basic principles are actually not too difficult to understand.

Quantum physics is the science of things that are too small to see or measure. So at its simplest it is as Bryan Cox describes “the calculation of the probability of the whereabouts of very small things”. Probability, because you can never know for certain. So that’s quite straight forward.

We then don’t need to worry about the calculation. Something else about this statement is very much more important. 

Full size physics depends on observation. But quantum physics deals with things that can’t be observed.

Observation is the bedrock of science as most of us know it. School science taught most of us that correct science depends fundamentally upon accurate observation. In quantum mechanics, this issue of observation has therefore raised difficult questions when it come to describing what the science is doing. After all, we are basically guessing. This is what Cox is saying. We can only calculate probability.

What does this mean? Lets say, just for the sake of explanation, that you know that a particle is in one of three positions. The probability of the particle being in position 1 is 80% and the probability of if being in position 2 is 19%. Therefore the probability of it being in position 3 is 1%. So if the probability of one position changes so will the probability relating to at least one of the other positions, because the total must always be 100%. 

So the position of the particle, the measurement of which is a single ‘event’, is split into possible events which are interdependent. This kind of event is quite different to the kind of events we are used to thinking about. Obviously, there are more than just three possible positions. Often, it is considered that there are an infinite number, but the principle is still the same. Physicists call this space, which in total is the space of a single event, both mathematical and real, ‘configuration space’. In it, each probability that can be calculated is connected to and changed by every other.

There is another very interesting aspect to this ‘configuration space’.

We all know that if you attempt to describe a thing you have made a guess about you end up describing your guess as much as the thing. A guess makes no sense unless you take into account what has led to it.

Let me give an example of this. Man Utd play Everton. Guess the score. I could guess 27-0 to Everton. You laugh. Immediately, you understand that this is a meaningless guess because I do not understand football. I could guess 5-0 to Everton. But you might still not take my guess very seriously because you might think I don’t understand the strengths of the squads and their current form. However, if you know that I am Premier League manager you might be on your way to the bookies shortly. The difference? You think I might have some information.

So in quantum mechanics, as well as the observation, you also have to have some information about how the observation was attempted. Otherwise the guess is more or less meaningless. When you conduct an experiment on things you can’t see, the equipment you use becomes part of the result. This is fully recognised by quantum scientists. They call the effect ‘entanglement’, because the observer and means of observation become entangled with the observed.

Because we are measuring probability, this entanglement of observer and observation becomes a part of the calculation of probability. The context of the calculation has to become part of the calculation. This is because guesses only make sense if you understand the parameters within which the guess was made. 

Scientist like Bryan Cox have been very good at attempting to explain quantum physics in accessible ways, but they have always talked about the science in the context of the study of things that are too small to see. However, it is quite clear that the more interesting perspective is the probabilistic one. Quantum science has been forced into probabilistic calculations by the fact that observation is impossible, and the calculations they make go far beyond any previous assessment of a probabilistic universe. Tiny objects may not appear to concern us immediately, but the laws of probability certainly do. In marketing, we are constantly making probabilistic judgements. 

Entanglement has obvious implications for how we judge the evidence we collect about the behaviour of customers, where we are again often drawing probabilistic conclusions. The degree to which data in particular is entangled with the means of collection is critical to accurate interpretation. It is an effect that has been picked up already by behavioural economists, and I will have much more to say about behavioural economics.

Once you understand that an ‘event’ in quantum physics is of this unfamiliar, probabilistic kind, many of the supposed mysteries of this branch of science become quite explicable. However, it is such an unfamiliar way of thinking that some aspects of it do appear weird. 

One of the weirdest aspects relates to the way that probabilities change over time. This is very important to quantum mechanics, as you can see, because splitting an event into many possible events is applicable to every instant of time. How do all these fractures instants come together? Does time exist at all?

Since time is a dimension we all depend on, the assumption is made that is does exist. In order to deal with this issue, an Austrian physicist called Erwin Shroedinger invented an equation which is known the wave function. This was first published in 1926 in a paper about wave mechanics, which established an equation based on a conventional wave equation to describe the movement of hydrogen particles over time, substituting  actual values for real and complex numbers called eigenvalues. This is as complicated as it sounds, and I will already be in trouble with any mathematicians who read this. The only important thing we need to know, though, is that it is a wave equation, and it is universally known as the wave function. It was found to work to describe the possible movement of all invisible particles over time.

What owe must not lose sight of is that the wave function is a description of the way that probability changes. Probability has a wave form.

This is a simple thing to say, but a very hard thing to understand conceptually, and it has even led scientists to deny that it can be expressed in a conceptual way at all. They tend to dismiss any ‘real’ interpretation because it is easier to say that the equation just applies to the very small things it was designed to describe. This is really why its implications in the real world have never been properly understood. “Just do the math”, as American physicist Richard Feynman once wrote.

However, understanding that probability is a wave is very important to understanding properly how our speculative theories work, wherever they involve assessments which are probabilistic.

Although this fact has been enshrined in quantum calculation for nearly 100 years, its consequences are still little understood. But, as the outcomes of quantum science increasingly govern the world we live in and the technology that drives it, nobody can afford to ignore them.

I will explain what the outcomes do mean, but first I will explain why we know that probability has a wave form.

How we prove that probability changes in wave form

The idea that probability might have a wave form will clearly effect the way we approach everything we do, particularly if our job is to make predictions about judgement and behaviour. Soit is worth understanding how scientists prove that probability changes in a wave form.

One of the experiments that started the journey of Quantum Mechanics is one most people have heard of – the slit experiment. In GCSE physics, it is used to show that light travels in waves. You shine a light source through two slits and you get a series of bands of light on the other side. You are told that this is an interference effect, like the one you get when ripples of water are passed through two slits.

So it won’t surprise you that you get the same effect when you pass these tiny particles you can’t see through two slits. A detector on the other sides shows them arriving in bands, like light. So they have a wave motion.

But there are two peculiarities which mystified physicists for a long time.

One. You get the these bands, showing an interference pattern, even though it is only possible to fire the particles through the slits one at a time. They are not sent in a ‘body’ like water. So what are they interfering with? Themselves?

Two. When you successfully observe them – which of course is difficult – the bands disappear and there is no wave motion detected.

The explanation for both of these facts is the same. We can’t see the particles, so the ‘events’ we are describing are of the kind I explained above – they are probabilistic, so at each moment they are composed of an infinite number of possibilities. We are not observing particles at all , but the probability of particles. And what this experiment is showing us is the way that these probabilities change.

Remember, we can’t see the particles at all. What the equipment we set up is showing us is therefore the places the particles are most likely to be. As Cox says, the experiment shows us the probability of the whereabouts. We are observing the places they are most likely to land. We are actually looking at probability. 

This is the sense in which one particle can appear to go through both slits. We are talking about the probabilities relating to the particle’s movement. Clearly, theres a probability they might go through eitherslit, and as we have seen all the probabilities add up to one event – the particles themselves will go through one slit or the other. 

What is therefore happening at the slits is that the probabilities relating to a particle passing through slit A or slit B interfere with each other, just like a body of water does. The reason these probabilities interfer with each other is this same dimple reason. The probabilities relating to any single event are never an event in themselves. All these probabilities (and there can be an infinite number of probabilities for any one event) add up to that one event.  So the sum of the probability that the particle will go through slit A and the probability that it will go through slit B for any given trajectory must always add up to one. It is a single event, so any variation in probability at one slit will create variation at the other. In terms of probability, the movement and interference of probabilities with each other is very like that of water. The particle may go through either slit. So probability (in the form of a possible track) goes through both slits, but of course the particle does not. So when we see the bars accumulating, we are kind of watching maths.

It might seem odd to be able to see probability as a physical effect without seeing present objects, but actually we are used to this idea. Think of goal on a football field. If we plot the landing point of 200 shots at goal they will be concentrated on the goal, with the pattern fading the further away we move. We can all recognise this pattern. It reflects the fact that even if the players aren’t particularly good there is a higher probability that the shot will go towards the goal than, say, the corner flag. But if we then visited the ground we would not expect to find 200 balls behind the goal! What we have in our heads is not a picture of 200 real balls but a probability map. 

What this all means is that it is a proven fact that probability has a wave form.

Why does this wave motion then ‘disappear when observed’? In order to make sense of this, we need to remember that nobody can actually observe anything directly. What scientists are talking about when they discuss observation in this context, is a further set up of equipment which is designed to show the movement of a particle, rather than where it might land after being passed through the two slits. When this movement measurement is performed, the particles detected don’t appear to move in waves at all but in straight lines. Therefore it is said that when we do manage to observe a particle in motion the wave vanishes. This shouldn’t really surprise us if we’ve understood that it’s the probability that has a wave form, not the particle. If the bands represent waves of probability, it all makes perfect sense. The probabilities vanish while we ‘see’ the particle because we momentarily know where it is, and we are able to detect that it is travelling in a straight line, as we would expect. So the fact that we have confirmed the position of the particle makes the probability wave disappear. 

What’s it got to do with marketing?

So now we have to ask this question: If this strange characteristic of probability is true of events involving small objects why would it be any different when applied to events involving large objects? We can see that it affects where footballs land, even though players do not have to face the problem of scoring goals through two slits. (If they did, we might see a very messy series of bars in their landing pattern, but to replicate the slit experiment with any precision the balls would have to travel at several million times the speed of light! Somebody did once succeed in achieving a bar pattern using bullets fired by a machine gun, but, presumably for safety reasons, this experiment is not frequently repeated!)

This is all interesting and weird, but what has it got to do with marketing?

Well, first of all we are familiar with the idea that research outcomes are influenced by all kinds of procedural factors. The idea that the observers might be entangled with the observed is not a particularly strange one. Goodhart’s law says that when a measure becomes a target it ceases to be a good measure, and that’s widely accepted. In more practical terms, whether we are talking about variations in old fashioned group moderation, where much depends on the skills and opinions of the moderator, or the ways we frame survey questions, or so called data mining, or the parameters we apply to our interpretation of it, we are all painfully aware always that there is no such thing as simple objective information.

Data…

Bayesian inference and beyond

The use and abuse of data is an issue very close to the hearts of marketers and agencies. Data analysis is often a divisive subjects. It can be simply dismissed by people who don’t like it, while others are entirely dependent upon it, but quantum mechanics offers critical and objective insight into the ways in which it should and shouldn’t be used.

If we think about it, the use of data is very closely associated with the calculation of probability. We use it as a means of understanding current and past behaviour, and predicting future behaviour. The attempt to calculate probability accurately of course greatly predates quantum science. It is a fundamental aspect of statistical analysis. For this reason, anyone who is an expert in data will have a sound knowledge of statistics and will also therefore have a good working knowledge of probability. It is important however to understand something of the history and methodology in relation to this. 

Probability seems a simple concept. We are used to thinking about it when gambling, or playing games, or calling a dice. It’s easy to calculate, up to a point. You just divide the number of times something happens by the number of times it could happen, and multiply by 100 to arrive at a percentage. That’s easy for a coin toss, where the probability of heads or tails is about 1/2, or 50%. It is simple because a coin toss is a self defined event with only two possible outcomes. The same goes for a dice throw, although there are six sides. In fact, any game or sport with rules is similar. Rules are designed to limit contextual interference, so that the outcome of the game is self contained and the possible outcomes are limited. 

Real life, though, is not like this. It is much messier, there are no tules as such, and possible outcomes are far more numerous. This doesn’t stop us calculating probabilities. In fact, we constantly attempt to assess risk in thousands of activities, from driving to climbing, to just crossing the road or making tea.

But it is clear that, for many of these events, which have no clear parameters, we can’t calculate probability the same we way as we do with a coin toss. 

How to work out more complex probabilities is a question that has preoccupied mathematicians for a long time. The first really successful formula for calculating probability in more complex events was developed more than 250 years ago, by Thomas Bayes. It was published after his death, in 1763, and offered a way of calculating the probability of an event which has became central to the development of statistics and computing.

Bayesian probability attempts to calculate probabilities where events are influenced by contextual factors. In simple terms, Bayesian probability allows a ‘given’ piece of information. Consider the risk of catching flu. It’s quite simple to calculate the probability of catching it or not using a simple ‘coin toss’ method. You can count up the people who do get it, in a specified time span or age group, divide by the total sample number and multiply by 100 to get a percentage. But let’s say you want to know the probability of having flu, given that you have a sore throat. What it is the probability you have flu now? 

This is a calculation we can do using Bayes’ theorem. The first step in a Bayesian calculation would be to understand the flu infection rate in a defined group of people to which you belong. Then we need a calculation that takes into account the different kinds of information we might have which relate a sore throat to flu. According to the Bayesian formula, the probability of the information you have implying that an event might happen (say, that a symptom of flu means you have flu) is equal to the probability of the event occurring at all (what is the the total number of people who currently have flu?) times the probability of the event taking into account the new ‘given’ information (what proportion of people who have flu also have your symptom? Note that this is the inverse of the initial proposition) divided by the probability of the occurrence of the information (how often does your symptom occur generally, regardless of flu?). 

So the probability I have flu, given I have a sore throat EQUALS

The probability I would have a sore throat if I had flu (ie the inverse proposition) TIMES the probability of having flu at all DIVIDED BY the general probability of having a sore throat.

For the sake of discussion we might say that’s about (0.5×0.4)/0.5 = 40%

This shows us that a sore throat is probably a poor test for flu. Better information – say, a hight temperature lasting more than twelve hours – might give a higher accuracy. Baysean probability is able to give a probability value adjusted by information which can be updated. In the latter case, test accuracy may increase, infection rates may decrease, and risk rates can be assessed accordingly. 

The certainty of the test is mediated by two variables: the probability of the event being tested for, and the probability of the occurrence of the information. 

But this gives some issues when the givens are more complex. If we try to work out the probability of getting run over and dying when crossing a road we run into a few issues. Here, we can use a value we already know, as the published probability of this happening by the age of 100 is calculated to be approximately 0.005 in the UK. According to the Baysean formula, working back from the answer, you might arrive at this probability rate as follows:

The probability of dying if you have been run over EQUALS the probability of you having been run over, if you have died TIMES the probability of dying before age100

DIVIDED BY the probability of being run over before age 100. We could guess might look something like (0.000051x 0.98)/0.01=0.005.

You can see how this formula could be used to give different values if you inputed different age criteria or used a different cause of death. However, we can also easily see that the Baysean calculation wouldn’t really be sophisticated enough to give you a true estimate of probability of risk to yourself from dying when crossing the road at any one time. It is not necessarily the case that you are likely to have been run over and killed by the time you have crossed the road 20,000 times. Clearly the calculation of this probability in relation to any one person will vary enormously, because there would be a number of factors which would relate directly to you and your surroundings, at the time of the accident. You would need to take into account your age and mobility, vision, power of attention and mental state, as well as the position of the crossing, the speed of the car and the attention of the driver, and there may then be further ancillary conditions to calculate varying even these variables further. 

This is a difficult, if not an impossible sum to do, because there is too much ‘information’. There are so many potential causes and conditions to take account of, many of which are closely associated, or entangled, with each other. Bayesian probability can take into account one piece of information, and variations in the one piece of information. It can calculate the probability of flu given a sore throat. But that does take into account other symptoms, such as age, location, previous health conditions, wealth, occupation, and so on and so on. Of course all of these can be calculated as ‘givens’. But the truth is that in any real context the givens are often so varied, subtle and diverse, they are infinite in number and nature. The technique of creating an inverse proposition only works if you can formulate a problem in the right way in the first place. Like a coin toss, the calculation has to have a clear parameter. The question of whether you will be killed crossing the road might have to be worked out given the rate of deterioration of your eyesight, your arthritis, your hearing, your age and mobility, the placing of the crossing point, the average speed of passing vehicles, the weather conditions, the time of day – the list goes on, and each given has variations, many of which have further variations. Together, we might call all of these factors contextual. Having just two pieces of variable information is then not enough to account for every context. Context comes into play in direct proportion to the exposure of the action to outside world. Exposures to and risks from disease can occur in a controlled environment but an action like crossing the road is entirely open to exterior factors. In fact, the more ‘everyday’ danger it is, the more closely entangled it can appear to to be in all kinds of different circumstances. 

All these variables are critical because the sum of all the probabilities attached to them must always add up to one event. Every new factor you encounter requires a recalculation of the probability of all others. And while technology might make the calculation itself possible, the entanglement of the availability, variability and reliability of the input information can make the actual data meaningless. In reality, reality is so complex that we mostly don’t bother to do actual calculations at all. You can see what I mean if we imagine certain specific circumstances in the case of crossing the road. If you were 98 and living in an old peoples home, could move at a quarter of normal walking speed and were killed on a crossing on a blind corner outside the gate of the home, it might be determined immediately that the probability of a serious accident had been negligently high, but it is the absence of any reliable means of calculation that frequently sees such cases decided in the very human environment of a court.. 

So calculation is very useful when things are simple. It enables business to make money out of gamblers, and clever gamblers to make money out of casinos. It helps health authorities to calculate general population risks from diseases, and insurance companies to provide financial compensation for obvious risks. But everyday life is too complex and different pieces of information are too entangled with each other to allow you to calculate every factor in every risk, or even most risks. As we can see, it is the unregulated everyday that poses the most complex risks and probabilities. 

This is the real issue with depending on data. It’s very useful, as long as you remember that the outcome depends on the input information, and that at some point in the process both the input information and the interpretation is wholly subjective. And the simpler and more general it becomes, the more the observer is entangled in the observation.

This is exactly the situation quantum mechanics is designed to deal with. The calculation of probability in all its complexity is impossible, but at least we can understand the way that probability forms and changes through time. Quantum physics shows us that probability has a shape. And this then gives us the opportunity to make all kinds of helpful guesses.

Making waves

In fact, we already apply some of the principles of quantum science. It won’t surprise you, either, that they have been developed, not because of any particular ideology or interpretation, but because they work, in exactly the same way as quantum science evolved. We know this set of ideas as behavioural economic.

Now, before you rush off and apply whatever you believe behavioural economics to be to everything you do, I need to add a caveat, about both quantum mechanics and behavioural economics. A  little knowledge can be a dangerous thing, and, of course, it already is.

We can see this if we apply the rules of quantum physics to some of the questions we always ask. Think in fact about the most fundamental question you want to answer when assessing communication. Something like, How do I get more customers to buy my product? Nobody reading this will, I imagine, believe that this question might be contrary to the principles of behavioural economics, but quantum mechanics would not allow it. Why? Because it completely ignores the laws of probability and entanglement. It assumes that there are simple reasons people don’t buy which can be addressed. There must in other words be something we are doing wrong. Behavioural economics does actually have a term for this assumption. It is a fundamental attribution error. I will talk about this in a moment. In the meantime, however, there is a simple way of understanding why it is a problem. It is failure based. What does this mean?  It assumes that, if you get it wrong, nobody will buy. Failure based marketing leads brand teams to ask the second wrong question, which again defies the laws if Quantum Mechanics: when and where do I put my messaging to get it in front of the people I need to reach? The assumption is that this is tge solution. What’s more, unless you try to do it you won’t do it, and it makes the fatal assumption that accurate observation can fix it. 

And this is the real problem with failure marketing. It measures reality against perfection, assuming the what is being done is failing to be perfect. The problem is that the answer to the original question requires accurate information. As we have seen you can’t have information of that kind in a complex situation, and since it is impossible to get that information you face the waste and failure we are all familiar with. 

Now lets look at it the other way round. Lets take a success based approach rather than a failure based approach. People buy your product anyway don’t they? So lets stop worrying about who doesn’t and who they are and what they think and how to reach them. We don’t need to convert anyone, unless the product is God. Instead, let’s think about who does buy and see where the footballs are going. Then we can ask the only question that matters. What do people like about your product, and how can you make people who like your product like it more? Why will this be so effective? Because people who like something tell other people about it. Then a wave forms. An information wave.

The good news is that, with this approach, if you set things up properly, you don’t need to do very much. But unlike the situation you are used to, where you are trying to find things out that you need to know, we are now going to start from a completely different place. We know what we are looking for. We are looking for waves. If we want to use them successfully we need to look for them and use them.

Although this might seem to leave a lot to chance, it doesn’t. It recognises that chance is the medium in which wit live. It is the only approach that is supported by science. 

As I have said, some people will be thinking as they read or hear this, that they already understand what I am describing as behavioural economics. I have already mentioned detection bias and observer bias. So they are right, up to a point. Or rather, back from a point. And as I’ve said, behavioural economics is indeed closely related to quantum mechanics, not least because it has evolved in the same functional way.

So what is behavioural economics?

Behavioural economics arose from the belief that decision making is emotionally driven.  In fact, its theories of behaviour identify a number of ways decision making happens as part of wave phenomena. Why is this so?

At the heart of behavioural economics is the idea that emotional decisions are faster and more instinctive, as opposed to rational decisions, which are slower and, it is often argued, influence our behaviour less frequently and less strongly. It is based on a reading of Daniel Kleinmans cognitive theories (see Thinking Fast and Slow). Richard Shotton’s book, The Choice Factory, was and is an influential attempt to explain how many of our decision making processes are of this fast, emotional kind. He usefully attempts to describe 24 different behavioural traits that relate to marketers and marketing.

The first one he discusses is the one I have mentioned above, and is arguably the most important.   Fundamental attribution error is about the importance of context. Research has shown repeatedly that a caring person in a rush is less likely to stop and help someone than an uncaring person with lots of time. From this it is concluded that behaviour is determined by situation not by personality. It doesn’t matter how kind or unkind you are. Decision making is contextual. It is the specific nature of the situation that is important. This is an extremely important observation, for this simple reason. If someone gives a homeless person sitting in the street a fiver, you immediately assume they are kind. But this action, like all action, is an event which is the outcome, not just of two, but of an infinite number of entangled circumstances, of which having time to perform the action is clearly one.

So the fundamental attribution error is an error relating to circumstances where we might overlook the probabilistic nature of events and their entanglement with each other and over-simplify the motivation for action. This leads to an unsupported belief in character and identity as the driving force behind human behaviour,. And this assumption leads directly to failure marketing, and asking the question, how do I get more people to buy my product? But here, the quantum mechanics of marketing is more radical than behavioural economics, because quantum mechanics is very clear about why it is futile to ask this question. In quantum mechanics ‘people’ do not exist. That is slightly sensational. To be accurate, they do not independently exist.

This shows us quite clearly the scientific scope of behavioural economics. It is roughly correct. We shouldn’t criticise it for that of course, we can after all say the same of quantum mechanics. But, like quantum mechanics, it is incomplete. It continually points us away from certainty towards a partial understanding of entanglement. Of course, there is no complete understanding of entanglement. 

So in the instance of the example above it is certainly true that kindness is one factor in the probability of events, but that other factors also come into play. 

It is also true that a kind personality is only one factor in the complex determination of the event.

But here is a caveat for people who want to use this information to market a product. RS talks about context as if this is easy to identify and manipulate. I am quite sure that he himself does not believe this, but it is an inviting misunderstanding. Context is complex. Context is all the information that you could ever input in order to determine the outcome of an event.

Now, context is easily restricted by rules. It is easy to work out the probability of a coin toss or a dice roll because those are strictly tule bound events. But I can’t tell you what the probability is of you being run over the next time you cross a road. There are too many dependencies. Your age, eyesight, mobility, location, environment, mood and so on and so on. Being in a rush might be a factor and then I’d need to calculate the probability of that. (I can give you a generalised figure. It is 0.05%. But in order for that to be meaningful you have to be the median of all the criteria used to work it out!)

You will probably see where this is now going. Algorithms are rules that define complex games. So putting your messaging in an environment where it is defined by an algorithm has two effects. It allows you a high degree of event manipulation. But it also restricts the event. There is a trade off. Fundamental attribution error points not to context but to entanglement, and while games define contexts they are also subject to entanglement. So it points to the limitation implicit in being in the right place at the right time, and it is an obvious limitation. You are only there. Attribution error also applies to attention. You have only been noticed because you are in the right place at the right time. Timeliness is an event in itself, and is quite independent of information. It does not mean that anyone cares.

This is this bear trap in behavioural economics. It say all the right things but doesn’t go far enough. It becomes its own attribution error. This is not only an error. It is an extremely dangerous error. Personality becomes a cult of personality and identity becomes a cukt of identity. 

However if we remember the principles of quantum physics this does not need tk be the case. 

This does not mean of course that information cannot be delivered.  

Why is this? A person in a rush has a trajectory. No human being is more single minded than one who is moving from point A to point B quickly.  The experiment demonstrates that behaviour is not determined by personality but it does not really tell us what does determine behaviour. What does having time mean? In fact it is more useful to ask, what does being in a rush mean? In RS’s example it means being with a group of people who believe that they have a reason to hurry. 

RS says this is behavioural, but we can see now that this it is also in accordance with the laws of quantum physics. If we are in a rush we are carried by. We are in fact carried along by events. It is hard to stop or slow a wave. RS sees this as so important that he says when we target work we should be targeting contexts as much as audiences. But I would go further than this. We should be working with contexts in the first place. Contexts provide our material. Targeting a context is like targeting a person. You are trying to get into a space where the audience is. That is the same thing as targeting an audience, and is just as expensive. But if you work in the right contexts you create a space where the audience might come. And then the audience will target you, which is the holy grail. If you want to use the wave analogy, you are making a pool for water to pour into. Creativity is like digging a hole.

Social proof – attraction to a crowd. The hole is a cat stuck in a tree. SWe are attracted to a forming wave of interest

Negative social proof – don’t say everyone is performing an action you want to prevent. It will encourage rather than discourage. Example sign in docs saying “30% miss their appointment times” is reassuring. Same wave effect no matter what the moral direction

Distinction. You notice someone dressed differently. Wave may start! (Requires bravery. Safety involves doing what everyone else is doing.)

Habits are hard to break. Disconcerting feeling having to walk up escalator. If people are being carried it is hard to get them to walk. Waves do all the hard work. This is the disruption problem. If you are surfing a big wave you can’t stop. Shotton says target category entrants and life changing moments – people at start of new wave

Payment interrupts behaviour. Why 99p is less than £1 and why contactless payment works for retailers. The wave continues

People lie if everyone else does. Information is misleading because it it driven by waves not facts

Mood – someone in bad mood doesn’t notice something good. Carried on wave

Price relativity. Perception is relative. It is determined by majority.  See below. Dot is same size.

Rory on why video conferencing never took off – seen as poor mans air trip not rich mans phone call. Then there was Covid!!!

Primacy effect. You remember first things in list first. This is wave tail behaviour. Shotten says memory saturated and first impressions always stronger. Quotes interesting explanation by Asch – first terms set up a direction. Once direction is established and stabilised later terms become less distinctive and memorable. 

Expectancy bias. Customers guessed price of brownie high low depending on whether it was on napkin, paper plate or china plate. A different version of social proof by means of presentation. This taken to tedium by restaurant dish descriptions. As Shotton says product is change by writing.

Confirmation bias. Once you are riding a wave its hard to get off

Overconfidence. Your own impressions of yourself used to make your own wave

Wishful seeing. Perception reflects desire. We see what we want to see – another version of confirmation bias

Media context. Paid for media more valued. Shotton is media guy!! Wave effect of contexts

Curse of knowledge. You assume everyone knows what you know. Everyone is on your wave. Why you should think like a customer. Eno – if you want to listen come out of the studio

Goodhart’s Law. When a measure becomes a target it ceases to be a good measure. QM wave/particle. You seek to meet target rather than understanding underlying objective which you will fail. Target destroys wave. Short term destroys long term.

Winner’s curse. The winner of an auction pays over the odds. You’re at the top of a wave…. Payday spend is 70% higher

Power of group. Canned laughter makes you believe something is funny. This is social proof

Veblen goods. More expensive is better. Value and Finest. Crowd behaviour for value. Belief that brands need to shun promotions.

Replicability crisis. Professions shadowing names. Mr McClean the dentist. But nominative determinism isn’t backed by evidence. It isnt replicated reliably. Its just felicity. But it might be true. Rory says failures die, successes live on. In wave world a ripple might go nowhere, but it might be the start of a big one.

Variability. Social proof affects some people more than others. Your reaction depends on context. You might be on a totally different wave. Social proof works when people are scared. Eg Should’ve. People are very scared of losing their sight. But scarcity – being unique and different – works better when they are feeling good… 

Cocktail Party effect – picking your own name out in a crowd. A version of hearing what you want to hear. Localise messages – your country needs you. Part of national wave of feeling.

Scarcity. The less there is the more you want it. There are only the rib eye steaks left.  A big wave is hard to catch

Ethics. Is it manipulative to create or ride waves. Argument about confirmation bias and social media algorithms. Social media is engineered to exploit confirmation bias. Tim Minchin. This is instantly political. Hitler identified, amplified and rode a wave.

You can see that all of these behavioural descriptions assume that information travels in wave form. When a ripple of an idea already exists in somebody’s head it will tend to amplify as a wave builds. If communication has an amplification effect it will be highly effective.

I’ve missed out from this list the effect which I find most interesting. RS calls this the Pratfall Effect. It describes the consequences of clumsiness.

Pratfall effect. Mistakes make people more appealing. Admit your brand has a flaw. Think small. Hand stuck in tube. Self deprecation. The product is small or awkward or has had its spelling abused and been turned into an animal. But the real reason this is powerful is because when you bump into something and knock it over it makes ripples. In fact the only way to start a wave is to bump into something and make a ripple. RS says that the best chance of growing brand is to flaunt flaws, but I’m not sure if this is quite right. The best way to grow a brand is to crash it into its context so it looks different from everything around it.

Behavioural economics give interesting insights into the ways we behave and are influenced. However, it can be tricky to use effectively as science of persuasion, because it is generally retrospective. The reason for this is that although it is good at recognising waves it does so from a point of observation. In other words, it restricts new opportunities by observing them in the old-fashioned way. 

We can see just from the things brands say in the press, how unfamiliar waves effects really are to us when they appear.

At present, brands frequently report themselves to be staggered when something ‘takes off’. Even in an industry as self promoting as marketing I almost never see anyone boasting about their actual skills at launching products into the stratosphere, for the simple reason the nobody knows what they are. 

Why?

For the simple reason that you don’t make it happen. You are just part of a sequence. You don’t have that kind of control. You can’t project.

So how do we get around this time issue?

(Although having done it myself I know that there are some techniques which are both essential and almost unknown.)

Of course, aside from the practicalities of doing it, owning the fact you made it happen is still important. The good news, for advertisers and marketers, is the fact that judging risk correctly gives us authority, not just with each other but with our actual customers. This means that behaviour is still influenceable if you go about it in the right way, by ignoring rules and seeking out the improbable. Improbability is highly detectable precisely because it is highly influential, and it is detectable immediately if you are on the lookout for it. Better still, if you can work out how to make an improbable event you can start a wave of influence yourself. 

How do you do this? There are of course no rules. If you want to do something exceptional, though, you do have to understand two things. You have to be exceptionally aware of what is going on, not just around you but around the lives of the people you want to influence, and you have to have an exceptionally sensitive understanding of how those people feel. 

This is not rocket science. In order to really influence behaviour you have to be just ahead of events as they happen. Empathy is as important as data. To succeed, you have to think like the people you are talking to, and you have to use your imagination and creativity to see what might happen next. I said before that events influence other events like waves, so maybe we can think of it as surfing. Data and research can spot and even predict waves, up to a point, but they are just observers. By the time you have reported a wave it is history. To catch a wave properly you have to be in the water, and just in front of it.