In 2012 it was clear the world of news publishing was undergoing a second revolution just as profound as the great online audience migration of the early 2000s: the rise of social media as a distribution channel. Andrew Hunter, Dom Filipovic and I were working at Australia’s biggest internet portal, ninemsn, and from 2009 we had seen traffic ramping up from Facebook in particular. That year saw the introduction of the “Like” button, which publishers could place on their news articles. Although we didn’t know it at the time, we were still in a period of relatively modest ad supply, and general advertising CPMs (cost per thousand impressions) were holding up. We knew from other changes we had experienced that the source of traffic – the distribution channel – has a tendency to influence and change the nature of the content, and because we saw up to a third of traffic coming directly from Facebook we decided it was time to investigate.
That investigation took the form of “Share Wars”, an independent research project that analysed sharing data from hundreds of publishers around the world. Out of the discoveries made from that data we wrote All Your Friends Like This, our book published by HarperCollins in 2015.
In that book we explain why sharing is important to news and how it changes the nature of news: the fundamental distribution mechanism on social networks is the recommendation, and recommendations carry not only an implicit approval of the content, they reflect on the recommender. This is completely different from a broadcast model or a first-generation digital model, and as a result different kinds of content come to the fore. We created a framework to explain this difference, based on what we saw in the data. In general we saw much more content being recommended and forwarded that reflected the identity of the sharer, a process we called “Teaming”. Topics that touched on personal identity – such as bicycling, pet ownership and politics – were much more likely to be shared on social networks and therefore were becoming more likely to appear in news outlets as those news outlets tuned their content mix to maximise traffic.
The key to the insights we made was having original data: no one else has access to the Share Wars database, because no one else had set up a site-scraping engine that collected the URLs of the English-speaking news media and then queried Facebook for the sharing data associated with that URL. Back in those days, Facebook allowed public queries on URLs, simply returning the number of Shares, Likes etc on a given address. We collected thousands of story URLs every day and pinged Facebook tens of thousands of times in a day, collecting the returned data so we could analyse it later. None of this is possible any more, as Facebook changed their terms of service and turned off this kind of access. You may be interested to know it was the original “Like” button mechanism that we used to query the sharing counts on any given URL.
Our 2012 discovery that the most-shared stories were mostly about “Teaming” – that is, reinforcing group identity – is interesting in light of subsequent revelations about the use of Facebook by foreign agencies to sow discord in democracies. We also investigated what we called “hoax news” at the time, which subsequently became known as “fake news”, and highlighted the vulnerability of the news/social media environment to the influx of untruths. The killer thing about news that isn’t true is that it has both a cost and content advantage over real news, and will naturally rise in any social network: it’s suppression relies on the audience and the trust that audience places in institutions. That trust, as we have seen in recent years, has been greatly eroded.
As Editor-in-chief of ninemsn and later as Chief News Officer at Mediaworks, I used the insights generated from this multi-year research project to direct content resources. For example, our dissection of highly-shared stories had led to the development of the “SENT” model: Simple, Emotional, New and Triggered. “Triggered” is the only criteria in that model that you won’t immediately understand: it relates to the relative commonness of the content subject. While people pay attention to unusual things, what we discovered from looking through that data is that if I can tell you something simple, emotional and new about something common then that will be a sure-fire sharing success.
The Share Wars project and writing the book has left me with an abiding conviction that original data is the key to useful insight. There is very little you can do in an information vacuum, which is exactly where many content businesses are still operating.
To read more about All Your Friends Like This and access two of the book’s chapters written by me, see Go Inside News:
Chapter 2: The Birth Of Share Wars
Chapter 9: Arminland (republished as We Choose To Know)