Netflix has your number
Insights from social media analytics are helping UOW researchers transform business education.
Every time you click play, streaming services are capturing data about who you are, what you like, and what you are likely to choose in the future.
American sports writing icon of the mid-twentieth century, Walter Wellesley "Red" Smith, is commonly believed to have described the process of writing with no little irony when he said: "You simply sit down at the typewriter, open your veins, and bleed." For decades, filmmakers and scriptwriters have sought to discover and quantify the magic that leads to screen hits. If they found it, they'd no doubt bottle it.
Dr Shahriar Akter, senior lecturer with UOW's School of Management, Operations and Marketing in the Faculty of Business, says that by 2020, there will be more video content consumed on Netflix in the US, and Australia than any other provider, including YouTube.
Dr Akter says the secret to success lies, not only in the sheer convenience and the shifting viewing habits of a younger generation that is accustomed to having their entertainment cake and eating it too, but in the stream of data each user provides about who they are, what they watch, how they watch and when. This is business analytics - or "big social data analytics" - at work. Researchers like Dr Akter see it as a huge opportunity, if companies can learn how to embrace its potential.
Dr Shahriar Akter. Photo: Paul Jones
Analytics for insights and competitive advantage
"Most people think Netflix is an amazing content provider of films and television series," Dr Akter says, "Behind the content production are extraordinary analytics capabilities, such as meta tagging of data, meaning, when you log onto Netflix, they know what you watch, when you watch and from where you watch." It's this type of insight that Dr Akter says is the "ultimate" in customer analytics.
"Netflix can measure customer by customer what it is you're watching. But here's what they're also doing. Every show you watch gets what's called meta tags, or if you like, attributes or descriptors. So they know if Thomas Quinn watched House of Cards in 2015 in a politically vibrant city like Canberra. Imagine having that massive amount of data from millions and millions of customers.
"Rather than saying, 'What show could we create in 2020?', the director of House of Cards sitting there is saying, 'I see what the data is telling me. People really like House of Cards that take place in politically vibrant cities like Sydney or Canberra'."
According to his recent research, in the still-developing field of social media analytics, Dr Akter and his colleagues found that despite the vast amount of available social media data, core business process that drive profitability have not changed.
On-demand streaming service Netflix now claims more than 81 million members in more than 190 countries.
And many firms don't know how to reconcile this treasure trove of data from more than 2.2 billion worldwide users with value creation. "Social media enables companies to get immediate access to such information," Dr Akter says. "However, information is only useful for companies when meaningful insights are generated from it and actionable strategies can be developed and implemented to address these challenge and generate greater business value."
Their recent papers, 'The primers of social media analytics' and 'How does social media create value', dissects the big social data and into specific uses to help establish a path to value. It builds on previous work that helps broaden the understanding of big data and its role in capturing value in the context of local emergency services. The massive penetration of social media paves the path for firms to derive value by capturing, understanding and presenting information on customer tastes and preferences, relationship managements, promotions, new product developments, crisis management and competitive intelligence.
The well-used tools of the trade for social media analysts continue to be measuring engagement and monitoring trends. How many likes, shares and comments as a proportion of your audience tells you how engaged they are. And keeping an eye on trends both on social media and in the outside world can help business managers or decision makers predict the future behaviours or trends.
In a similar vein, topic modeling allows analysts to detect dominant themes from social media that help paint a picture of the customers' interests. Using historical data is one level, turning it up a notch is to use data to predict outcomes. For example, how does a brand, say a bank, or even a political campaign strategist, understand how people feel about what their product or person represents? The answer is opinion mining or sentiment analysis.
"Amazon, e-Bay, Pandora, Last.fm, iLike, and many others predict customer demand based on the positive recommendations and reviews," Dr Akter says. "The research literature tells us that an overall positive or negative sentiment based on comments and reviews does have a real impact on sales. This was always believed to be the case and now we have the data from social media to prove it." And as illustrated at the outset, Netflix is most definitely using predictive tools to inform what shows it invests in and creates as Netflix 'originals', as well as continually defining the customer experience to take on the big production studios at their own game.
The big data network for business analytics
Imagine this scenario: you go into your favourite department store or other retailer's website and you start clicking around various products. The store can log your IP address or because your browser saves your password, you've automatically logged into your account and they know you are in the store, electronically speaking. The next step is to have on hand your purchase history data, which tells them your value as a customer.
Are you a frequent visitor who is always busy or are you a rare visitor but you always make purchases? Either way, the retailer can assign you a value based on measures of recency, frequency and monetary (RFM). "Google has been working for the last five years on driverless cars," Dr Akter says. "Now you might say, 'Wow, that's kind of cool and innovative'. Yes, but marketers view it as an opportunity to monetise it. Imagine Google partners with your favourite store whose site you're browsing.
"They know your customer value and they could send a driverless car over to your house to pick you up and drive you to the store for free. If you think by the way that this is distant future, I would say the future is now because Queensland has approved the trial of driverless cars in the last week. Now imagine if a store teams up with Google to provide this analytics driven service to you as a loyal customer, I am sure you would stick with that store."
Insights become education
Those insights are not only research-based, but are changing the way marketing, business analytics and related subjects are taught, closing the gap between research and the classroom and enabling students to keep abreast of the rapid pace of technology-driven change in the business world. "When I share my research findings of social media analytics with students, the can instantly relate and engage with them in the learning environment," Dr Akter says.
"Research methods and findings encourage creative thinking and problem solving, which are essential graduate skills for business students." Dr Akter has introduced Instagram based class projects for his first year students, such as new product innovation project, a corporate social responsibility campaign, and an advertising campaign to bring social networks into the classroom and he encourages engagement with students through the class Twitter account.
The connection between research and teaching is a two-way conversation, with students providing Dr Akter with insights from classrooms filled with mostly young digital natives who think about branding and advertising in a very different way to older generations. "In order to understand consumer behaviour in digital markets, discussion of digital behaviour with students who regularly use Amazon, eBay or Netflix, for example, is really helpful in understand the black box of Generations Y and Z.
"Research-driven education has helped me to develop key theoretical ideas or analytical tools which are usable in multi-cultural environments. Discussion of social media analytics in the classroom has provided me rich insights of the cultural perspectives in analytics, such as how Chinese consumers use Tencents QQ for shopping, Sina Weibo for micro blogging or WeChat for micro messaging." This interaction, and its new concepts, methods and practical insights, is playing a critical role in maintaining momentum in social media and big data driven e-commerce research.
Separating sense from noise
Social media is fast becoming a living laboratory for researchers wanting to answer critical questions about social interactions among consumers and the implications for society and markets. The challenge will be to make sense and meaning from the volume, variety and velocity of data social media creates.
"Big social data usually are stored in two forms: the structured data comprises user profile characteristics and social behaviour - likes, retweets, comments and so on - and the unstructured data, which relates to user-generated content from micro blogs to rich content with audio and video.
"One of the biggest challenges ahead of managers is to bridge the gap between these two sources and connect the attitudinal and behavioural data on individual consumers to study them in a holistic manner. More sophisticated data analytics techniques are need to achieve this goal, but when that happens, it will enable marketers to communicate and engage with the consumers more meaningfully."