How do you teach an old dog new tricks? By subjecting it to experiments, of course. Perhaps the best known experiment in behavioural science – and the oldest trick in the book – is the Pavlov experiment, now considered standard in generating conditioned responses.
Nowadays, behavioural economics, behavioural insights, behavioural tech are all the rage. And we human beings are the new guinea pigs, dogs, mice, pigeons, i.e. we are under constant experiment. It has been taken to such ridiculous levels in information technology that, as Shoshana Zuboff writes in her book, Surveillance Capitalism, technology companies claim to know more about us than we know ourselves. More about her book, when I will be writing a review of it soon on my blog.
The question is, has behavioural science run ahead of itself and all of us, and is it too late to put the genie back into the bottle?
As an advertising and brand communications professional, I know from my long experience in the industry that we do depend on consumer behaviour insights when creating campaigns for our clients and building their brands. As someone who has also studied economics, albeit not behavioural economics which must be an entire new stream of the discipline now, I am aware that plenty of the economic choices and decisions ordinary people make are based on their socio-economic conditions and upbringing as well as their demographics. Which is why even in the advertising business, we consider consumers as socio-economic segments of individuals/households, for purposes of simplification and classification while developing strategies and campaigns. Each is motivated by a different set of parameters while making their brand choices; what’s more, there is a certain process by which consumers make their decisions, often known in industry parlance as the consumer decision-making journey.
Behavioural economists justify the need for their discipline on what is quite erroneously considered a new theory, that man is not always known to make rational decisions. To advertising and marketing professionals, that is old hat, as old as the industry itself. We have always known that when deciding between brands, consumers respond to rational as well as emotional brand benefits. Equally important, consumers are guided most in their decisions by need, household budget/product price, tastes and preferences as well as their self and social images. These factors would change in order of importance, depending on the consumer segment.
In his book, Thinking, Fast and Slow, behavioural economist Daniel Kahneman writes about how we human beings apply two modes of thinking, but under different circumstances and conditions. Basing his work on the study by two psychologists, Keith Stanovich and Richard West, he writes:
“I will refer to the two systems in the mind as System 1 and System 2.
- System 1 operates automatically and quickly, with little or no effort and no sense of voluntary control.
- System 2 allocates attention to the effortful mental activities that demand it, including complex computations. The operations of System 2 are often associated with the subjective experience of agency, choice and concentration.
…The automatic operations of System 1 generate surprisingly complex patterns of ideas, but only the slower System 2 can construct thoughts in an orderly series of steps.”
There has long been a variation of this theory in advertising as well. Advertising and marketing folks have always recognised that there are two kind of purchase decisions that consumers make: those on an impulse, and those that require considered deliberation; these are usually differentiated by purchase value/price, but occasion and sensory considerations can also play a role.
If behavioural economists are restating what is already common knowledge in advertising and marketing circles; they do so to emphasise their role in helping governments shape policy decisions. Since governments don’t necessarily understand the pulse of their citizens, this might seem in order. However, what is worrying is the growing trend of combining behavioural science with information technology, especially in the digital age. The problem is not so much in the idea or principle itself, but the methods employed to garner those behavioural insights and the uses that these are then put to.
It is by now well-established that both behavioural science and information technology grew at around the same time in the US, and have accelerated their pace of growth and importance since the start of this millennium, especially after the dot-com boom going bust. They both owe their importance in the economy, to government. Especially the American government, which since the Cold War era, and more recently 9/11, has invested billions of dollars in research into behavioural science and information technology to develop improved surveillance and defence capabilities.
That is a world far removed from advertising and marketing or consumers. Except that in the digital age, internet-based tech companies have almost all built their businesses increasingly on an advertising-based business model and that is where we folks come into the picture. As I wrote in my blog post on building brands in the digital age, Google, Facebook, Amazon are media heavyweights in the advertising industry now, and their business depends entirely on generating behavioural insights from our cyber footprint trails. Often without our knowledge or permission to do so.
How are these behavioural insights different or superior to what we in the advertising industry depended on, all these years? For one thing, they are based on our browsing activity on the internet, so they are more relevant to the higher socio-economic classes of people, especially in a country like India. More importantly, I believe that the data that they are able to harvest off our browsing, purchasing and emailing activity are based on past behaviour. These companies are said to be developing even more powerful algorithms that have predictive capabilities, but that is not quite the same as understanding what makes different segments of consumers aspire to different goals, be motivated by different needs, their changing lifestyles and preferences, their product usage and attitudes, etc. At least not for now.
The reason for this, and the fundamental difference therefore, is the fact that behavioural insights mined from our internet activity are purely transactional in nature, not relational. On the other hand, brands are about relationships between companies and their products and consumers and other stakeholders. This very transactional dimension is also connected with the other point I make in my previous post on brand building in the digital age, regarding digital campaigns being downstream, sales-driven activity while brand-building is inherently an upstream investment. Therefore, I find the sudden craze to join the digital bandwagon (or digital brandwagon) and spend millions on digital and social media campaigns quite inexplicable.
Coming back to the area of concern regarding the methods employed by behavioural scientists and internet tech companies,
- both cite the importance of experiments, in order to be effective
- both are based on a system of incentives and rewards as ways of making people behave in certain pre-determined ways.
- finally, because they are effective only in experiments, users’ responses are always fodder material for further experiments.
These are all quite distinct from consumer behaviour as we understood it in the past, and as you can see, they are all transactional. And because so much of it uses digital technology and the internet, it is also insidious and treacherous in its aims and practices. Behavioural data harvested off the internet ostensibly uses it to enhance our experience, and in the process uses it to further future research, manipulate our emotions and thoughts, follow us wherever we are, etc. It is a creepy world, with eyes and ears everywhere.
Hal R Varian, chief economist at Google is the purveyor of just such a world. In a talk he gave at a NABE (National Association of Business Economics) Annual Meeting in 2013, Beyond Big Data, he said:
“But tools for data manipulation are only part of the story. We have also seen significant developments in methods for data analysis that have emerged from the machine learning community. Nowadays we hear a lot about predictive analytics, data mining and data science. The techniques from these subjects, along with good old-fashioned statistics and econometrics have allowed for deeper analysis of these vast data sets, enabled by computer mediated transactions.”
Whereas previously, conventional consumer or marketing research required research organisations’ expertise in framing the research parameters and objectives, developing areas of probing and questionnaires, managing focus group discussions, nowadays behavioural insights can be generated through clicks on an advert and hits on a website. Again, formerly, research was relational and based on marketers’ and advertising agencies’ requirements in understanding their consumers better, while the so-called behavioural analytics of the digital world are an hourly or daily stream of at-your-fingertips information which is often not a very good guide to how consumers feel, think, live and make their decisions in the real world. It is equivalent to missing the wood for the trees. One should not expect it to be otherwise, since behavioural insights in the new age are based on experiments, and experiments, as we all know, are conducted in highly controlled conditions.
I am of the view that merely because behavioural science and technology enable a new possibility hitherto unimagined, it doesn’t necessarily make them the right or ethical choice always. For example, Nobel Prize-winning behavioural economist, Richard Thaler, who, along with Cass Sunstein, came up with the idea of Nudge (also their best-selling book) advising the UK Government on private sector pension plans might make sense and the programme has indeed been successful. What about using Nudge theory to the Covid-19 pandemic asking citizens to use tracer app technology? Would it be justified? What about issues around privacy, surveillance and basic civil liberties? In a country like India, use of a tracer app could also lead to heightened stigma and ostracization of Covid patients, negating the benefits of the idea. And if its use is justified on ethical grounds or for the common good, what can we say of human rights in China that is building a mass surveillance society?
Speaking of Nudge theory itself, Daniel Kahneman admits it is a product of libertarian paternalism, the idea that people are not rational enough to make the right decisions. In Thinking, Fast and Slow, he writes:
“Thaler and Sunstein advocate a position of libertarian paternalism, in which the state and other institutions are allowed to nudge people to make decisions that serve their own long-term interests. The designation of joining a pension plan as the default option is an example of a nudge. It is difficult to argue that anyone’s freedom is diminished by being automatically enrolled in the plan, when they merely have to check a box to opt out.”
In advertising too, we have had the concept of the nudge; all of what we write and create as campaigns for brands is nothing but nudge. However, we use the gentle power of persuasion and do not assume that individuals are incapable of making sensible choices for themselves.
It is strange that System 2 of the human mind that Kahneman refers to, which enables agency, choice and concentration should be given short shrift in Nudge. Then again, as we observe in digital communications and technology, more and more of our decisions are being made for us. Increasingly, we are losing our agency and freedom to control our own lives. Because the gods and evangelists at Google, among others, have decided it must be so. In the same NABE paper, Hal Varian gives us an insight into Larry Page’s thinking:
“Larry Page used to say that the trouble with Google was that you had to ask it questions. He thought Google should know what you want and tell it to you before you ask the question. We all thought he was joking, but Larry’s vision has been realized with Google Now, an application that runs on Android phones… the issue is that Google Now has to know a lot about you and your environment to provide these services.”
I thought technology was supposed to be at the command of man; now we are increasingly hurtling towards a world where tech will act, command, guide on your behalf. If this isn’t ridiculous enough, Hal goes on to compare sharing information with Google with talking to your lawyer, doctor, accountant… people you know and trust.
The next big thing or episode in this digital journey is the Internet of Things, when every gadget and gizmo and perhaps even us (with wearable tech) will be connected. Elon Musk, founder and chief of Tesla, has already expressed the desire and vision to connect the human brain with machines, to make us more like them. We will have everything answered for us, done for us, even thought for us.
Tom Friedman, in an article in New York Times, warned of the dangers of “the deep” in all the deep machine learning that is going on. According to Shoshana Zuboff, we are being herded into a world of surveillance where powerful companies and governments will wield total control. Yuval Noah Harari has also alerted us about such a world in his books, notably Homo Deus and 21 Lessons for the 21st Century. I agree with a lot of what they write and still have plenty of reading of my own to do on the subject.
For now, let me sign off by saying that it doesn’t have to be this way. Behavioural science and digital technology, both have their place in a consumer society. Only, balance must be restored and choice handed back to citizens. And for that we need governments to do a better job of regulation. But when government and the information technology industry have been in a symbiotic, parasitic relationship for decades, it’s hard to see what can change it now.