“Forget artificial intelligence – in the brave new world of big data, it’s artificial idiocy we should be looking out for.” (Tom Chatfield)
Rather than focusing on artificial idiocy, let’s think about how we can be naturally intelligent in our use of any size of data to be better marketers. While data is a powerful tool to provide evidence for better decision-making, too often it misses the human empathy, cultural perspective and creative thinking that make us wiser.
Emotions are key to human lives, because they tell us what is important and what is not important. As Benjamin Hoff wrote in The Tao Of Pooh, “Those who have no compassion have no wisdom. Knowledge, yes; cleverness, maybe; wisdom, no. A clever mind is not a heart. Knowledge doesn’t really care.”
Emotions are critical for successful marketing, and the evidence in the data suggests that they increase marketing ROI and build price elasticity. In the IPA’s 2013 study of Advertising Effectiveness: The Long & Short of It, an analysis of advertising campaigns submitted for industry awards showed that those which were emotional had almost twice the ROI of rational campaigns (based on the number of large business effects observed).
The analysis also showed that emotional campaigns led to very large reductions in price elasticity for 7% of brands (compared with 5% for mixed campaigns and 0% for rational campaigns). However, these effects had to be measured over three or more years.
Emotional campaigns work by priming people’s minds, gradually making the brand associations stronger and stronger in people’s minds. Although this takes time to be effective, three or more years according to the IPA’s analysis, it works because we all make decisions based on “does this feel right?”.
Much of TapestryWorks’ work is helping clients better understand which are the most relevant feelings for people who are doing a specific customer job (or shopping in particular category).
In a recent self-funded survey, we explored the motivations for female beauty in Australia, Indonesia, Thailand and UK using the StoryWorks framework and visual and verbal stimuli based around 12 human values.
Using verbal stimuli, confidence (or a local translation) was seen to be the single greatest goal of beauty for women, chosen by 63% of women in Indonesia, 55% in UK, 53% in Australia and 49% in Thailand. Confidence dominates the responses, and even when aggregated across the 12 motivational segments (or aggregated further into 6 segments), it severely skews the results.
Confidence is commonly associated with beauty. Does that imply that this is just the rational response to the question? Does confidence mean the same thing to women in the four countries, or does it have multiple meanings, making it a popular answer?
Octavio Paz famously said, “I’m not saying a literal translation is impossible, only that it’s not a translation”. His point was that you can translate words, but you cannot translate meaning. For example, in Bahasa Indonesia, confident is usually translated as percaya diri, which literally means “believe in yourself”. Is that really the same meaning as confident?
How do women’s beauty motivations translate into images? In our online survey, we also tested visual stimuli designed to match the verbal concepts behind each motivation. Confidence remains important for women, with a visual version of confident first or second most popular choice in all four countries, but now chosen by around one third of women.
Overall, the responses to the question using visual stimuli were more consistent (the top 5 motivations are the same across all countries, although appearing in a different order). Confident remains the most important single motivation in Indonesia and UK but a visual version of powerful was most popular in Australia and a visual version of hopeful in Thailand.
Moreover, if we focus only on women who chose the verbal concept of confident, we get exactly the same result. Their visual preferences are not consistent with their verbal preferences. But we know that the sensory and visual perception is very closely tied to our emotional world, and both are mostly done in the unconscious brain.
In our survey, we tested a different set of images in each country, and in Indonesia these were local images with Indonesian people and contexts. Using these images, responses were even more nuanced and diverse, as the local context made interpretation a little more rational and literal. However, the overall pattern of choices matched that of the main image set very closely, showing the same differences from the question based on words only. The responses in both cases make complete sense based on other research on Indonesian beauty conducted by TapestryWorks.
What the results show is that human decision-making is implicit more than explicit, and emotions are best measured in non-rational ways. They also show that culture plays a big role in shaping the relevance of emotions.
The expression, “fish can’t see water”, is a good description of the way that much of our cultural landscape is hidden from us, influencing behaviour in ways that we don’t even notice. For example, in China the standard greeting doesn’t just mean “hello” or “how are you?”, it literally means “have you eaten rice today?”, having a profound but largely unseen influence on perceptions of rice and what it means to people.
Culture and context play a large role in shaping perceptions of beauty too, and going back to our example, let’s compare women’s choices in three different contexts. From a list of 36 words and phrases in Bahasa Indonesia, the most popular choice to describe the goal of beauty was percaya diri, a standard translation of confident although literally meaning “belief in yourself”. The most popular choice from 36 “Western” images designed to match the verbal concepts was the image below, matching the verbal concept of charismatic (a form of confidence, and in the same motivational segment). And the most popular choice from 36 “Indonesian” images, matching the same concepts, was the second image below designed to match the verbal concept of hopeful.
Does percaya diri mean confident, or self-belief, or charismatic, or hopeful, or all of these? Even when we look only at those women who selected the verbal concept percaya diri (189 out of 300 women), the most popular images from the two sets remain these two.
The answer may be that verbal concepts drive people to respond in a very rational and predictable way, focusing on answers that are the most popular or most familiar. The answer may also be that verbal concepts are slippery, meaning different things on different occasions, and that images provide more context and therefore more specificity.
In fact, both are probably influencing women’s responses. In other work conducted by TapestryWorks earlier in 2017, we looked at women’s beauty rituals at different times of day and on different occasions. We found that daily routines are more strongly associated with natural looks, and involve fewer beauty products.
This everyday beauty was different in the morning where it was associated with giving energy and impetus to the day, reflecting each woman’s individuality, during the day where touching up make-up gives a break from routine, and in the evening where the focus is on cleaning off the grime of the day and reconnecting with loved ones.
By contrast, beauty routines for special occasions like parties and weddings require more products and more extensive routines (sometimes including salon, spa or gym) and involve bolder colours too. Beauty goals for such formal occasions are very much about being charismatic and seeking to be the centre of attention. “Belief in yourself” is an important goal of beauty for Indonesian women, but with a very different meaning between these occasions and the normal routine of every day.
Human behaviour is complex and artificial intelligence cannot even come close to replicating our complex lives (yet). However, artificial intelligence and big data can work well within domain specific tasks (like playing chess). This limitation leads me to the third weakness of big data which is creative thinking.
Steve Jobs is often quoted as saying that “creativity is just connecting things” although most people forget that he went on to talk about the importance of different experiences and self-reflection. Perhaps a better quote is from Maria Popova (who runs a truly creative website called brainpickings.org). Her take is that, “in order for us to truly create and contribute to the world, we have to be able to connect countless dots, to cross-pollinate ideas from a wealth of disciplines, to combine and recombine these pieces and build new castles.”
This is something that big data cannot do, but small data (or thick data) can. If we go back to the difference between short-term rational advertising campaigns and long-term emotional priming, we must remind ourselves that long-term priming has much better ROI, but that ROI cannot be measured in the short-term.
In a new analysis of campaign data published in 2016, IPA alarmingly demonstrate the rapidly decreasing effectiveness and efficiency of campaigns, in tandem with increasing numbers of data-driven and short-term campaigns. Efficiency halved from 2010 to 2014, and their most recent data also show that while short-term campaigns show stronger effects on short-term sales activation than long-term campaigns, they have much weaker effects on market share and profit, while long-term campaigns almost triple impact on market share and lead to significantly higher profitability.
Are we focusing on the right things? Indeed, is big data actually encouraging us to focus on the wrong things or at the very least the things that don’t matter? Nate Silver writes in The Signal and The Noise that, “One of the pervasive risks that we face in the information age … is that even if the amount of knowledge in the world is increasing, the gap between what we know and what we think we know may be widening.”
Nate Silver also warns that “most data is just noise, as most of the universe is filled with empty space”. By contrast, the smallest piece of data can bring you to a much deeper understanding, allowing you to connect multiple dots that previously seemed disconnected. One small example of this is work I conducted for Facebook on Asian Beauty (you can read an article based on the work here).
Although big (or should that be medium-sized?) data from an online survey of 4,158 women across four countries explained some of the differences between the countries, it was a single, simple observation that explained another key dimension of difference. Women in some countries would sit quietly in the focus group facility until asked to start a discussion, while in other countries they would immediately introduce themselves to each other.
This difference between friendly and more formal cultures, combined with the differences between cultures that are more passive or active in the pursuit of personal beauty, led to the overall cultural codes attributed to the four countries:
- Indonesia – “Accepting Beauty”
- Japan – “Pure Beauty”
- Malaysia – “Aspirational Beauty”
- South Korea – “Perfect Beauty”
Without empathy, perspective and creativity, big data will never lead to wisdom. Cynthia Ozick wrote that “data is memory without history”, pointing to the impotence of data that lacks meaning and our inability to turn information into knowledge without such meaning. Or as Albert Einstein put it best, “Not everything that counts can be counted, and not everything that can be counted counts”.
Advertising Effectiveness: The Long and the Short of It by Les Binet & Peter Field (IPA)
Selling Creativity Short: Creativity and effectiveness under threat by Les Binet & Peter Field (IPA)
The Signal and The Noise: Why So Many Predictions Fail-but Some Don’t by Nate Silver
Unveiling the Beauty Secrets of the Modern Asian Woman by Facebook IQ https://insights.fb.com/2016/12/20/unveiling-the-secrets-of-asian-beauty/
[This is a longer version of a SoundBites talk for Data-Driven Marketing Association of Singapore on 29 November 2017. You can access the presentation slides here.]