Thomas Stoeckle, Head of Strategic Business Development at LexisNexis BIS shares in this interview his approach to making sense of big data, and he explains the complex dynamics between humans and algorithms. News has always been storytelling and Thomas takes this further and explains what statistical or digital storytelling is about. Finally, Thomas explains the difference between earned influence and paid promotion.
Q: There is still too much noise around big data, when in fact data is not the real problem, but the way in which it is used. What’s the best approach to get there?
Thomas: Big data is big or small in the context in which it is generated and used. What is needed is the right volume and format of data to solve any given problem. It’s about the amount of information that we can cope with, and the amount of information needed to make the right decision from any number of choices. Framed that way, this is no longer merely about number crunching, about cool data sciences, or about zetta – and yottabytes. It’s about human beings, their organisations and their decisions. It is instructive to widen the perspective from the somewhat narrow data science focus to look at human sciences. That is how we get from data to insights to knowledge and even wisdom. So it’s about anthropology, history, sociology, psychology (social, organisational, evolutionary etc.). The field of behavioural economics is perhaps the most advanced in bridging the divide between data and humans.
Big data is nothing new, really. Yuval Noah Harari wrote a remarkable book, Sapiens: A Brief History of Humankind, which illustrates the big data problem of ancient times, brought about by the agricultural revolution: as our ancestors moved from being nomads and hunter gatherers in small communities, to cultivating land and manipulating a few animal and plant species in larger settlements, the amount of relevant and important data very quickly began to exceed brain and memory capacities. More complex societies generate more complex information which requires new forms of acquisition, distribution, storage. An oral system of communication was no longer sufficient. Numbers became a vital form of information. And thus humankind (as far as we know, the Sumerians, some 5,000 years ago) came up with systems of abstract symbols to share and store information. And from there, we never looked back…
Tom Standage, in his clever account Writing on the Wall: Social Media – The First 2,000 Years, tells us about ancient Rome’s equivalents of social media [and the old stone tablets looked eerily like modern ipads], and he makes the point that whilst technologies emerge and evolve, they still press the same buttons in our slow evolving and still largely Stone Age brains. And that is the point about the best approach to succeed with data: to better understand how we humans tick, and then align technology-driven approaches accordingly. The folks at IBM talk of AI as ‘augmented’ rather than ‘artificial intelligence’. Together with a healthy dose of humility, that strikes me as the right angle.
For the field of public relations research, my friends over at the Institute for PR have some excellent material, for example on behavioural insights and digital media. And of course there is the stellar work of the IPR Measurement Commission (of which I am co-chair).
Q: Humans vs Algorithm: how far algorithms can go and when media companies need a human touch?
Thomas: Expressions such as artificial intelligence, machine learning, and of course algorithms have become part of our everyday vocabulary. However, casual use isn’t the same as knowledge and understanding. Therefore, when we discuss humans and algorithms, we should start with education. I do a regular podcast, the Small Data Forum, with my dear collaborators Neville Hobson and Sam Knowles. There is hardly an episode when these themes do not feature.
On the BBC website’s science for kids section, an algorithm is defined as a sequence of instructions or a set of rules that are followed to complete a task. In other words, algorithms do what they are told to do. There is nothing good or bad about them. It’s how they are being used that makes them so. The discussion of algorithms in the context ofdigital and social media focuses on the fact that these instructions are set, tuned and adjusted to serve the commercial needs of businesses. Digital is fast becoming advertising’s biggest source of revenue, and Facebook and Google dominate the digital advertising market. The value of advertising is in reaching and getting the attention of audiences.
Today most people find their news and information through just a handful of sites and search engines.These sites make money from advertising when we click on the links they show us. And, they choose what to show us based on algorithms which learn from our personal data that they are constantly harvesting. So the result is that they show us more stuff they think we will click on – because it is surprising, shocking and attention-grabbing, or designed to appeal to our biases.
That’s a self-reinforcing mechanism with the risk of isolating and insulating people. We have seen that in the UK referendum and the US presidential elections in 2016: political debates are becoming increasingly polarised, without engaging the other side in critical and open discourse. Such a polarised media environment then becomes a perfect breeding space for fabricated clickbait and fake news.
The companies benefiting commercially claim ‘algorithmic neutrality’ and self-control, but there is also an increasing call for regulation, for ‘algorithmic accountability’ and ‘algorithmic transparency’. The European Union’s General Data Protection Regulation (GDPR) is a firm step in this direction. But I’m afraid this goes beyond the scope of this interview.
Q: Statistics are out, personal stories are in! I read recently that good news should be a storytelling. What do you consider that statistical storytelling could be?
Thomas: We humans understand the world through stories and images. Again, there is nothing new about this. What’s perhaps new is how communications professionals are taking ownership of the storytelling idea. You claim your right to be part of the community by confidently using the term narrative in every other sentence. And – wow – social media mega trends are pointing toward ever more visual and audio-visual data. Who would have thought…Well, the earliest cave and rock paintings are from the Upper Paleolithic Age, some 40,000 years ago. Instagram and Snapchat are just the latest digital outlets for a basic human need to express itself.
News was always storytelling, and it always followed clear rules and story arches: when I studied communications in the late 80s (yes, that long ago!), I learned about the inverted pyramid as the basic journalistic construction principle: have the most important information at the beginning; answer the W questions (who, what, when, why etc.). Statistical storytelling needs to achieve basically the same as the prehistoric cave paintings: first it needs to get our attention, and then it needs to impart information through telling us a story.
Sadly, earlier this year Hans Rosling, one of the giants in the field of statistical storytelling, passed away way too early. His TED talks and other appearances on YouTube are masterpieces of bringing statistics to life, of turning numbers into engaging stories. We can’t all become Hans Roslings, but we can pay more attention to that ultimate purpose of reaching and engagingour target audiences through stories, rather than drowning them in a tsunami of data.
A great and very recent example of statistical storytelling is Al Gore’s documentary An Inconvenient Sequel: Truth to Power, the follow-up to the 2006 Oscar-winning blockbuster global-warming documentary. It is weaving a compelling narrative thread from climate science data, latest developments in renewable energy, political negotiation and footage of – in Al Gore’s words – “mother nature screaming at us” (in the form of devastating storms, floods, droughts etc.).
This is also an area that my company is increasingly focusing on: LexisNexis is providing a global news tracker for an online resource centre hosted by our parent company RELX Group, where we provide data and content from across the group to drive awareness and understanding of the UN Global Goals, a set of 17 global sustainable development goals. Digital storytelling will be key in raising awareness for, and driving progress in achieving the goals.
Q: Digital influencer marketing is becoming a blunt instrument. What is the difference between earned influence and paid promotion?
Thomas: Simon Sinek, another famous TED talker, teaches us that we need to “Start with why“. When we think about our work, our work environment and our companies, we are usually quite good at explaining what we do and how we do it. But we rarely reflect on why we do it. The influencer marketing ‘hype’ falls into that category. Technical availability and expediency have created an illusion of linear causality between reach metrics and targeting. The idea of the modern marketing funnel is a sophisticated model applying the laws of physics, especially gravity: as the apple falls from the tree to the ground, so does the message get transported from the sender to the receiver through the right channel at the right time for maximum impact. And the key to this are so-called influencers with the right physical qualities, i.e. lots of friends and followers, likes, links, retweets and so on.
I wrote a piece about this recently, where I argued that paying for product and/or message promotion through influencers (be they bona fide public figures and celebrities, or YouTube, Instagram or Snapchat sensations) is just another form of advertising.
The real (and as yet largely untapped) opportunity is in more personal and intimate connections. We hear and read a lot about trust, and the erosion thereof: Edelman’s Trust Barometer, launched annually as part of the World Economic Forum in January, is showing a downward trend in the trust in institutions. This has significant consequences for the dynamic relationship between authority, trust and influence: we get our news and information from search engines and social networks. This is where the power of algorithms comes into play again: the search preferences of other users, as well as the direct recommendations from peers, friends or family, influence us increasingly more than the authoritative voices of the past. Edelman calls this the “Inversion of Influence”.
It is a general and seemingly irreversible trend that we trust our peers more than someone with authority. Someone with authority will talk at you, whereas a peer will talk with you. And that’s the crucial difference: earned influence is created through dialogue. Paid promotion is just another monologue. This then brings me back to the interplay between data sciences and human sciences: we need both, and we need humans to stay in control.
About Thomas Stoeckle
Thomas Stoeckle leads strategic business development at LexisNexis Business Insight Solutions (BIS), where he explores innovative and intuitive paths to improve client value through richer insights from content and information aggregation and analytics. Prior to joining LexisNexis, he was group director and global analytics lead at W2O Group, and managing director at Report International (now CARMA). He is also the co-chair of the Institute for Public Relations Measurement Commission.
Thomas believes passionately in story-telling through robust data evidence and compelling visualization. Forever a digital Neanderthal among digital natives, he is keenly aware that today’s constantly evolving challenges demand fluency in the three languages of business, technology, and of course humans.
Originally from Germany, he has been living and working in London for the past 17 years, working with clients all over the globe. A regular speaker at conferences and industry events, he is endlessly curious about the disruptive and transformative power of Big and small data. He hosts the Small Data Forum podcast, where recent discussion points with his collaborators Sam Knowles and Neville Hobson include the growing role of data analytics in political marketing, and the phenomena of fake news and the loss of trust in factual information. Thomas is also a jury member of Quadriga University’s Digital Communication Awards.
About LexisNexis BIS
As part of LexisNexis Legal & Professional, LexisNexis Business Insight Solutions (BIS) serves companies and organizations across the globe, offering premier news, corporate information and public records through a portfolio of solutions, including the flagship media intelligence platform, LexisNexis Newsdesk. This unique combination of market-leading content and innovative technology helps business professionals make more insightful decisions by offering them quick and easy access to the latest news, facts and insights regarding their brand, clients, prospects, competitors, suppliers and industry trends. LexisNexis BIS leading products and services are used throughout the world and in virtually every industry and area of business.
LexisNexis Legal & Professional, which serves customers in more than 175 countries with 10,000 employees worldwide, is part of REXL Group, a global provider of information and analytics for professional and business customers across industries. REXL Group is one of the constituents of the FTSE 100 Index, Financial Times Global 500 and Euronext 100 Index.