#1 Big Data; It’s About Size AND How to Handle It

Listen to The DesignSpark Podcast

#1: Did Facebook Try To Sell Me A Bubble House?

“50 percent of our advertising spend is wasted. If only we could work out which 50 percent.” One good thing about the modern era of Big Data is that we can probably say goodbye to that tired old joke. On the otherhand, there is a sense that many are not comfortable that their personal preferences are being mapped and analysed in minute detail. How did this come about, and how does it work?

Clearly, the Internet has marked differences, compared to broadcasting, as a content-delivery mechanism. Traditional television or radio broadcasting is more or less “fire and forget”. It’s impossible for broadcasters to know which receivers out there are tuned to their frequency. Information about who is watching, or listening, has typically been captured through organised surveys such as carrying out direct questionnaires or installing equipment in the homes of selected, consenting audiences to monitor their viewing; the UK-based Broadcasters’ Audience Research Board (BARB) has been monitoring viewing patterns this way since its inception in 1981, beginning with a panel of 3000 homes.

The Connection is the Key

But all this is so last millennium. Today’s media consumers are kind enough to request their preferred selections to be delivered to their personal IP addresses. BARB still uses metering equipment in the homes of its panel members to monitor viewing patterns, and now fuses this with device-level data from any PC, tablet or smartphone that connects to registered channels to monitor viewing trends and provide information of value to content makers.

Information about shopping habits is similarly eagerly sought. For years, retailers have offered inducements such as discount coupons and loyalty schemes to harvest data about their customers. That data has been expensively acquired and is useful to guide marketing decisions such as purchasing and pricing. More recently, with the spread of data-driven marketing, the data has acquired a monetary value too and, with customers’ consent (or lack of unconsent) has been marketed to interested third-party organisations.

So, data-driven marketing is not new. Marketers have been moving in this direction, arguably, since the beginning of marketing (whenever that was). What is new is Big Data and – perhaps even more important – the advent of tools and techniques powerful enough to extract valuable information from it.

Why is it Big?

The information now being exchanged over the Internet, in forms such as online registrations, login credentials, and details of purchasing transactions, is creating oceans of data for marketing companies to capture and mine.

And all this is without even mentioning the personal information people give away about themselves willingly on social media platforms. Browsing, shopping or watching while logged into these platforms provides the link between the device and the user that allows messages such as advertising to be targeted and personalised. There is also the spread of the Internet of Things (IoT) and the popularity of personal connected devices such as home digital assistants, which are capable of recording user interactions for analysis. And more and more of us are wearing fitness trackers that can monitor information such as location, current activity and even health-related data such as eating patterns and heart-rate trends. People are putting more and more personal information onto Internet servers all the time – sometimes without being fully aware or even properly understanding their privacy settings or the permissions the settings grant to the service provider (e.g. Amazon Echo – you agree for your data to be used to train speech-recognition algorithms, but how that training is done and what access human agents have to the information is not described).

Still more contentious is the suggestion that organisations could be using smartphones to eavesdrop on individuals’ conversations to serve them advertising relevant to their current concerns or needs. This article by BBC technology reporter Zoe Kleinman describes how easily software can be created to relay background utterances to a desktop or server from anywhere in the world, which could be used for several purposes – targeted advertising perhaps being among the most benign. On the other hand, it may not be necessary to listen in to the microphones we all carry with us each day. This piece on digitaltrends tells how social-media platforms maintain connections to huge numbers of websites and can recognise logged-in subscribers to send them related advertising.

As the data ocean becomes ever larger and deeper, the range of uses for the data is becoming wider and wider. Some of the more unpleasant among these are political surveillance and election tampering, which has made headlines recently in the aftermath of the 2016 US presidential election and UK’s Brexit referendum.

Monetising Big Data

While it can be difficult for individuals to understand how unknown parties’ knowledge of their data will impact them personally, some questions we can be sure everyone would like to know are how do companies monetise Big Data, and what exactly is it worth?

Arguably, the companies making the most money from Big Data are the Internet search giants and social media platforms, who can exploit that power to link connected devices with users and serve advertising – helping advertisers (their clients) save that wasted 50 percent.

On the other hand, information from Big Data is valuable across the spectrum of marketing activities. However, Big Data is unlike traditional marketing data; by definition, the data sets are too large or complex for traditional data-processing tools to handle. Companies such as retailers or credit-card providers have the technology to collect more and more data from their own stores or payment terminals, for example, but then need powerful tools to extract useful information from it.

This presents opportunities for specialised data analytics consultancies to position affordable services that can help companies derive value from the large quantities of data they already hold and the even larger quantities of data they are likely to collect in the future as the number of connected devices and sensors in the field continues to proliferate. Subject to data-protection regulations, they can also aggregate data from multiple sources and market the data – or insights derived from it – to third-party organisations.

The ultimate democratisation (in the corporate sense) of Big Data analytics will likely come in the form of Analytics as a Service (AaaS). AaaS lets companies access heavyweight software in the Cloud on a pay-per-use basis, saving the huge up-front investment in compute power and software development needed to create a bespoke platform. Ultimately, this will enable companies to take advantage of cutting-edge machine-learning platforms to capture insights lurking deep in the data ocean that would otherwise never come to light. Discovering “what lies beneath” is both an exciting and a frightening prospect.

Conclusion

Big Data is different from old-fashioned data, and the world is still learning about its true power and how to handle it. Despite privacy concerns, many are captivated by the opportunities for monetisation; various stakeholders stand to gain by saving ad spend, growing market share, and generating revenue from data and the insights that can be derived from it. The financial prospects mean we can say one thing for sure: Big Data is here to stay and it’s probably going to get bigger.

For a more light-hearted take on Big Data head to Apple Podcasts, Podbean or Spotify and download the first episode in the brand-new series of The DesignSpark Podcast.