Jeff Bezos, as CEO of Amazon, has access to more data on customers than almost everyone else on earth. He wrote in his 2016 letter to shareholders: “Good inventors and designers deeply understand their customer. They spend tremendous energy developing that intuition. They study and understand many anecdotes rather than only the averages you’ll find on surveys.” [1]. Why would Bezos care about the anecdotes when he doesn’t have just big data, he has enormous data?
Pivotal to a great strategy is a great insight. From ivory tower pontification to innovation workshops to immersive field studies, everyone has their pet method for deriving insights. And, increasingly big data analytics and artificial intelligence are being employed to find nuggets of insights in heaps of data.
Broadly speaking, our efforts to find insights can be categorized on dimensions such as: directed (i.e. hypothesis-driven) or non-directed, quantitative or qualitative, objective or subjective, etc. In this article, we argue that complex problems require the use of complementary methods, such as big data (generally large sample, quantifiable and objective) and thick data (often small samples, descriptive and subjective).
Business executives can learn from social scientists. In human-based or social sciences research, a common strategy is to generate insights after collecting data from a variety of sources in a comparatively non-directed manner on a topic or question of interest and then analyze the data for themes. It relies on moving from a position of not knowing to a position of developing theory and insights on a topic. It often presumes that these insights are context specific and subjective and thus hold the data provided by participants as truthful and relevant in their own right.
Tricia Wang was working for Nokia in 2009. In her wonderful TED Talk, Wang [2] uses such a strategy, known as ethnography, to collect data on the lived experience of a potential customer base. She relied on her observations and the responses of people she interviewed and observed to develop a contextually relevant narrative before generating any new insights. Importantly, she did not approach the topic with her own hypothesis which might have prevented her from learning new and essential insights from the customers themselves [2].
The data pointed to an emerging popularity for affordable smartphones, which big data used by Nokia had failed to identify. Nokia’s big data analytics was dependent on historical data, of which there was little or none regarding affordable smartphones. Moreover, their surveys were hypothesis-driven and informed by their existing business models. Nokia interpreted the market from their hypothesis-driven, quantitative, big data -- and missed the boat entirely. In the four years from the start of 2009 to the end of 2012, Nokia’s market share fell from 50% to less than 5% [3].
The fundamental issue with big data, irrespective of how much good analytics or even artificial intelligence is applied to it, is that one can only find, or not find, what is in the historical data set. And the fundamental problem with hypothesis-driven research is one can only prove or disprove the hypothesis. What is also required is to surface insights that are not in the historical data and have not yet been conceptualized by the strategist or researcher.
In Tricia Wang’s words: “I concluded that Nokia needed to replace their current product development strategy from making expensive smartphones for elite users to affordable smartphones for low-income users. I reported my findings and recommendations to headquarters. But Nokia did not know what to do with my findings. They said my sample size of 100 was weak and small compared to their sample size of several million data points. In addition, they said that there weren’t any signs of my insights in their existing datasets.” [4]
Tricia Wang’s ethnographic research is a cornerstone of design thinking: an increasingly popular method adopted by corporates to explore and conceptualize design challenges [5]. It is optimistic and empathic, focusing on how individuals interact with services and products [6]. The designer/ researcher participates in exploring design challenges, as well as co-creating sustainable services with the users [7]. This increases the chances that users actually desire and use the services.
So, is design thinking or other ‘thick data’ methods the panacea for insights? No, not on its own. Big data analytics is powerful and complementary. For the best insights, we need both. Wang describes a highly successful combination of research strategies by Netflix which essentially introduced the world to binge watching. They achieved this with the use of thick data insights that they added to their already well-established, big data recommendations algorithm [2].
Admittedly these are expensive strategies. But not as expensive as getting it wrong. Ask Nokia or many of the companies that have spent $122 billion on a big data industry of which only 27% of projects deliver profitable returns [2].
Nate Silver, the renowned statistician with five honorary doctoral degrees, writes in [8]: “Data-driven predictions can succeed – and they can fail. It is when we deny our role in the process that the odds of failure rise. Before we demand more of our data, we need to demand more of ourselves.”
Acknowledgements to my co-author for this article: Dr. Philippa Skowno.
References:
[1] “2016 Letter to Shareholders”, Jeff Bezos
[2] “The human insights missing from big data” TEDX Cambridge, September 2016, Tricia Wang
[3] http://www.tech-thoughts.net/2012/12/smartphone-market-share-trends-by-country.html#.WceszcgjFPY
[4] https://medium.com/ethnography-matters/why-big-data-needs-thick-data-b4b3e75e3d7
[5] “Change by Design,” in Change by Design: How Design Thinking Transforms Organisations and Inspires Innovation, New York: HarperCollins, 2009, T. Brown
[6] “Design thinking for social innovation,” Stanford Soc. Innov. Rev., vol. Winter, 2010, T. Brown and J. Wyatt
[7] “Thinking beyond the cure: A case for human-centered design in cancer care,” Int. J. Des., vol. 6, no. 3, pp. 27–39, 2012, T. Mullaney and H. Pettersson
[8] “The Signal and the Noise”, Nate Silver