The Financial Times recently ran a special report entitled Decoding Big Data. Authors of the report, Paul Taylor, Caroline Nevitt and Katie Carnie, write, “As more people connect to the internet, the volume of information generated by our digital lives rises.” To help people understand the issues involved with big data, they prepared a 4-minute “explainer” narrated by Paul Taylor. The explainer can be accessed by clicking on this link. In a separate piece, Taylor writes, “The ‘digital universe’ continues to expand rapidly and will double every two years through 2020, but only a small fraction of this ‘big data’ is being analysed, which leaves a huge untapped opportunity for companies and other organisations.” [“Lack of data analysis worrisome,” Financial Times, 12 December 2012] Taylor reports that the market research firm IDC asserts that the growing mountain of unanalyzed data is falling into an ever-increasing “data gap.” A report issued by IDC states that “data holds potential analytic value that could have major business, health and societal impact” and yet “only 0.5 per cent of that data are actually being analysed.” That leaves an incredible amount of untapped intelligence locked inside potential data gold mines. Taylor continues:
“The reason [that the data remains unanalyzed], IDC suggests, is that the majority of new data being generated is unstructured – for example video surveillance footage, audio files or social media – meaning we know little about the data until it is characterised and tagged. The industry currently faces a shortage of the talent and technology needed to tag and analyse unstructured data.”
For students just entering college, the field of big data analytics is a promising career field to consider. Taylor reports that the study claims that “the digital universe in 2012 has taken on the feel of a tangible geography.” Unfortunately, it is a landscape filled with “vast, barely charted [places] full of promise and danger.” Taylor concludes:
“This untapped value could be found in patterns in social media usage, correlations in scientific data from discrete studies, medical information intersected with sociological data, faces in security footage, and so on. However, even with a generous estimate, the amount of information in the digital universe that is ‘tagged’ accounts for only about 3 per cent of the digital universe in 2012, and that which is analysed is half a per cent of the digital universe. ‘Herein is the promise of “big data” technology – the extraction of value from the large untapped pools of data in the digital universe,’ says the report. The report’s authors conclude that ‘our digital universe in 2020 will be bigger than ever, more valuable than ever, and more volatile than ever.’ Not surprisingly, IDC predicts that big data will be a big boon for the IT industry.”
Among those hoping to benefit from better big data analytics are marketing firms and companies for which they work. Emily Steel reports that digital era marketers bear little resemblance to cast members of the hit television series Mad Men. The new stars are data scientists. “A new generation of executives, armed with millions of terabytes of data, are taking over today’s advertising world,” she writes. “They are schooled in creating sophisticated automated systems for buying and selling ads, searching for patterns in the data to tell stories and tapping algorithms to evaluate the effectiveness of marketing.” [“Data scientists take byte out of Mad Men,” Financial Times, 12 December 2012]
“Across the industry, advertisers are tapping big data in hopes of reaching marketing’s ultimate goal: targeting personalised ads to the right person at the right time. ‘There is a spectacular change occurring,’ says Quentin George, chief innovations officer at Interpublic’s Mediabrands advertising buying group. ‘What big data is trying to do in the ad business is harvest consumer signals to deliver something more relevant and more meaningful.’ Marketers have long mined consumer information – ranging from public records data about how much a person’s house is worth to surveys about whether they are married or have children – to send direct mailings and make telephone pitches to people most likely to buy their products. Even the ‘mad men’ drew on panel-based research about the television shows people watched, the radio stations they listened to and the newspapers and magazines they read.”
It’s clear why advertising agencies that watched traditional advertising campaigns dwindle would jump on the big data bandwagon. The industry, however, still has to convince clients that the bandwagon is going to lead a parade of consumers who will purchase their products. Steel continues:
“Big data’s renewed heft in the advertising industry … came partly as a result of a concurrent disruption of the advertising business as smartphones spread and consumers digitised their lives. This not only unearthed a treasure trove of real-time data about individuals, but is also forcing the advertising and marketing industries into new ways of doing business. Marketers, eager to prove the return on investment of their ad dollars, suddenly could track the ads consumers saw and clicked on. They are discovering, for example, that nationwide US television advertising campaigns are failing to reach a large portion of their target audiences.”
The reason, of course, is that traditional television ad campaigns reach only viewers who are watching the programs during which the campaign airs (and, if they have digitally recorded the program, they may skip over the ad). Marketers certainly have data that tells them the demographics of the typical audience, but the approach still reaches lots of people who are not likely consumers of the product being promoted. That’s why other, even more targeted, approaches are being explored. Steel continues:
“The industry is, therefore, creating dossiers about individuals based on everything from the information revealed in online dating profiles to the pictures people posted to social networks, the items they put in their online shopping carts and the television programmes viewed on digital video recorders. These profiles are now the nexus around which an increasing part of the $518bn advertising industry operates. The rise of online ad exchanges, which operate much like stock trading platforms, means that marketers can now buy ads based on the profile of the person who sees the ad, rather than the site where it appears. Within the blink of an eye, marketers can evaluate the performance of an ad and alter the message, the way it looks or the audience targeted to see it. Most of the main advertising holding companies operate units to buy ads via these new programmatic ad-buying models and report that their clients are devoting an increasing share of their budgets to them.”
Although Steel makes it appear that most of the $518 billion spent on advertising is targeted, the fact is that the vast majority of that money is not spent that way. That shouldn’t come as too big of a surprise based on Taylor’s discussion concerning how little analysis is really being done using big data. Steel admits as much. She writes:
“While the advertising industry is dazzled by the profusion of data and the possibilities it wields, many marketers struggle to sift through the information to track the most relevant details about consumers, let alone mine the data for insights and deploy sophisticated targeting models. Having to deal with too much information, and not all of it accurate, is a common complaint among ad executives. A recent survey of 120 top marketing executives, conducted by Publicis’ Razorfish digital ad agency, revealed that many businesses continue to deploy age-old strategies of tapping historical sales data instead of real-time information.
‘The business needs to catch up with the hype of the technology at this point,’ says Ray Velez, global chief technology officer at Razorfish.”
Steel next raises an issue I’ve written about quite a bit in relation to big data — privacy. She writes:
“Even larger loom privacy issues, which threaten to limit the amount of consumer information marketers can track and use for ad targeting. This is becoming a bigger area of concern as companies marry information collected both online and away from a screen. MasterCard, for instance, recently started to make available an analysis of transaction data for ad targeting. Facebook, meanwhile, has worked with Datalogix, a data company, to track whether people buy products after seeing an ad on the social networking site.”
For more on this topic, read my posts entitled Big Data Dilemmas and Targeted Marketing: Understanding the Individual. Steel concludes her article by noting that even more intrusive data collection efforts are in the idea stage. For example, “A Verizon patent application published in November details plans to create a digital video recorder that watches television viewers and listens to their conversations for ad targeting. A Visa patent application published in 2011 describes tapping information from DNA data banks.” Steel discusses such “intimate” data collection in a separate article entitled “Advertising gets the power of intimacy.” [Financial Times, 12 December 2012] She writes:
“For marketers, the fast-growing business of tracking intimate details about consumers’ lives presents unprecedented opportunity. Lawmakers, federal regulators and privacy advocates, meanwhile, increasingly worry it is leading to a ‘database of ruin’.”
She borrows the term “database of ruin” from Paul Ohm, an associate professor at the University of Colorado Law School. He writes:
“Many businesses today find themselves locked in an arms race with competitors to see who can convert customer secrets into the most pennies. To try to win, they are building perfect digital dossiers, to use a phrase coined by Daniel Solove, massive data stores containing hundreds, if not thousands or tens of thousands, of facts about every member of our society. In my work, I’ve argued that these databases will grow to connect every individual to at least one closely guarded secret. This might be a secret about a medical condition, family history, or personal preference. It is a secret that, if revealed, would cause more than embarrassment or shame; it would lead to serious, concrete, devastating harm. And these companies are combining their data stores, which will give rise to a single, massive database. I call this the Database of Ruin. Once we have created this database, it is unlikely we will ever be able to tear it apart.” [“Don’t Build a Database of Ruin,” Harvard Business Review Blog Network, 23 August 2012]
Companies need to take privacy issues (and concerns like those expressed by Ohm) very seriously. Targeted marketing will only create value for companies and consumers if they agree on what data can be collected and how it can be used. As I’ve noted in other posts, there is good reason to believe that such accommodations can be reached; but, if Ohm is to be believed, the time for reaching a compromise could be running out.
