If there is one industry that knows what it means to be digitally disrupted, it's the media. With massive ongoing changes in the way people consume content – online, mobile and through social media channels – the media industry has had to quickly transition to digital, while also figure out new business models.
CIO Australia looks at several examples of media companies that are making this transition and the technologies that are driving their businesses forward.
Content recommendations and targeting
For any content driven website, analytics rule. The longer a reader stays on the website clicking through to other articles or content items, the better, along with a high frequency of returning to the website, and sharing its content with their social media networks.
Going beyond monitoring traditional aggregated data in Google Analytics, recommender systems have started to come onto the scene for editorial businesses, meaning it's not just applicable to e-commerce sites like Amazon and eBay. Better targeted content and personalising user experience does not only mean more audience engagement, but also makes a more compelling proposition to advertisers, which are a major source of revenue for media companies.
Australia's Bauer Media, a magazine publisher with more than 60 websites, has a recommender system for creating more detailed viewer segments, to better predict which piece of content to serve to the viewer based on what he/she is likely to resonate with.
“We've got a project that we are working on right now to automatically tag all of our content so that we can get a better insight into what people are actually reading about. We have a lot of sites on different platforms, and they'll have stuff categorised as a celebrity news story, for example. That's really as much insight as we can get from that, so it's not really that useful to an advertiser," said Simon Rumble, head of data, analytics and CRM at Bauer Media Australia.
“If we are able to dig deeper into who those people are, which celebrity they are looking at and what kind of news story it actually is, we can start building up more detailed segments.”
Rumble is using natural language processing to pull out key terms in an article, and then look up the meaning of those terms in an external database. If a viewer is reading an article on actress Jennifer Lawrence, for example, the system would scan the article and pull out the key term 'Jennifer Lawrence'. It would then notice it's capitalised to identify it as a proper noun, and then look it up in an external database that states the term refers to a person, the fact that she is an American actress, and a list of the movies she has starred in.
This is not only used to recommend other articles tagged with the actress's name, but also build up a segment for people who have viewed more than three stories on Jennifer Lawrence in the last 30 days.
“That could be called the 'Jennifer Lawrence enthusiast' segment. So when the next Hunger Games movie [starring Lawrence] comes out, instead of just pitching to the distributor, 'Hey, we've got this segment of people who are interested in celebrity stuff', we've actually got a much more detailed set of segments that we can also throw at them," Rumble said.
one giant category [celebrities] is limiting for us. So by being able
to automatically tag our content, we can segment our audiences better
but also get a better insight into what people are actually looking
Detailed segments is also a focus for News Corp Australia, one of the largest media companies in the country. Chief technology officer, Alisa Bowen, who recently took over from former CIO Tom Quinn, said it's silly to serve up the exact same home page, site navigation and layout to every single one of News Corp's 4 million monthly viewers.
“We are determined to break out of this out-of-date view, so can we know how the site navigation should be structured or the content should be presented for all our users,” she said.
Bowen and her team have been pulling together data signals directly from viewers when they interact with the content, when they fill in their details for entering a competition, when registering with the websites, and when commenting on articles.
The team has also combined that with data from its partners through Quantium, which includes retailers and bankers, so they can tell not only what viewers are interested in reading about but also what they spend their money on.
“Mary Meeker [known as 'Queen of the Net'] has a wonderful expression, she says: 'An app should never ask a consumer what it ought to already know'. And that's our guiding philosophy when it comes to the work we are doing on personalisation and the application of big data insights for product experiences,” Bowen said.
News Corp Australia is also leveraging Unruly's ShareRank technology, which it acquired last year. The proprietary algorithm measures the emotional response of a viewer when watching a particular video, and predicts the likelihood of that video going viral as a result.
“It is a mobile advertising network for video ads that play in stream and are very specifically targeted for different segment types based on how likely that video is to engage certain emotions and responses from the audience,” Bowen explained.
The algorithm was fed 10 million streams of data, hundreds of millions of videos, along with people's emotional reaction to them online and how many times they've been shared on social media.
“They can provide guidance to video editors and producers around how to re-cut the clip to get the most emotional reaction,” Bowen said. "They can also provide information to marketers and advertisers about whom to target their clip to."
For example, the Queensland Government created an awareness campaign targeted at young male surfers on how to stay safe in the water. But when Unruly did its tests it identified that the strongest emotional reaction actually came from fathers with young men who are out surfing, rather than the intended demographic.
“So what Unruly advised the government to do with that campaign was to take a portion of their media spend and allocate it towards this additional target market with the expectation that they would be the ones who would virally share it, because they have the strongest emotional reaction,” Bowen said.
Continues next page.
Yahoo7 goes down virtual and wearables path
Yahoo7 is investigating how virtual reality (VR) could apply to its business and enhance the audience's experience when interacting with content.
“I think VR will change every aspect of the digital world and media will have a part to play in that,” said Paul Russell, director of technology at Yahoo7.
An example of how VR headsets could be used is for major events that Yahoo7 reports on at the scene. By having a 360 degree camera for the journalist to take footage of the surroundings, viewers at home can feel like they are also experiencing it at the scene when viewed in their VR headset.
Yahoo7 has also developed a wearable app for the Apple Watch, which came out last year, to allow sports fans get notified of the latest sport game results on their wrists. Users can view live scores and upcoming matches in AFL, tennis, cricket, A-League, and English Premier League.
"At the moment wearables are fairly focused on health and fitness, but I imagine in the coming months that will be something a lot more applicable to media," said Russell.
Continues next page.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.