Penguin Random House Publishing Direct Justin Ractliffe recently wrote a new, in-depth report for The Copyright Agency which examines the evolving consumer marketplace for books in Australia, and how publishers can align with modern reader shifts.
The report provides some interesting perspective on the utilization of data-driven insights as a driver of publishing success, and repeatedly refers to Netflix as an example of a modern, content-focused business which has built consumer data into its product decisions, leading to greater success.
Which is not surprising – Netflix has repeatedly been held up as an example of how modern data analytics can be used to maximize traditional creative processes.
You may recall the story of how Netflix used ‘big data’ to create ‘House of Cards’, which all of its indicators showed would be a hit.
As per SOFY:
“Netflix identified that the British version of House of Cards was watched by many subscribers. Those members who watched the British version of House of Cards also seemed to favor movies starring Kevin Spacey. This was one of the patterns that led to Kevin Spacey being cast in the lead role – in fact, big data was instrumental in how most of the characters were cast. It had a role in how the script was finalized and how the overall narrative progressed.”
Brilliant, right? Tapping into consumer data, from Netflix’s 155 million subscribers, to create shows which align with their interests, thus driving viewer loyalty. Great.
Except that’s not really true.
Definitely, Netflix, as a case study, provides some interesting considerations, as noted by Ratcliffe, but the correlation/causation elements of this argument don’t play out as strong when you dig a little further. Yes, Netflix has achieved success – but how much of that, really, comes down to its original content?
Here’s a simple counter – Netflix published some 700 original shows in 2018, and 80 original films. How many of them can you name?
Sure, Netflix has had some big hits – the platform has struck gold with shows like ‘Orange is the New Black’ and ‘Stranger Things’, and movies like ‘Bird Box’. But on balance, Netflix’s production output is still fairly hit and miss. If Netflix’s data-driven approach were as spot on as implied, why are so many of its offerings failing to gain momentum, and attract big audiences?
But Netflix has 155 million subscribers, which many use an indicator of the success of its original content. Is that an accurate correlation?
Not necessarily. Netflix benefits from brand equity as much as it does from its content – the platform is cheap to access (in relative terms), provides access to a heap of movies and TV shows (Netflix-created an otherwise), and there are also, every now and then, cool, original shows that people want to watch. Given these various elements, it’s impossible to attribute how much of Netflix’s overall success comes down to its’ own productions.
To make it an even more difficult comparison, Netflix doesn’t actually reveal any viewership numbers, so we have no idea what’s working and what isn’t on the platform.
Netflix did, just recently, publish a list of its best performing originals between 2018 and 2019, which provides some insight:
But as noted by The Guardian, it’s interesting, here, to note that only two of the shows listed weren’t a first season. So, again, despite Netflix’s data insights, it’s clearly not fueling ongoing success – which suggests that, potentially, the user analytics-driven approach isn’t as effective as suggested.
Arguably, the Netflix programs which have seen success have been driven to higher levels of viewership by the social media hype train, as opposed to simply aligning with consumption data. Indeed, ‘Black Mirror’, ‘Stranger Things’ and ’13 Reasons Why’ were all among the most tweeted about TV shows in 2018, while the Bird Box challenge became a trending meme on social media, which undoubtedly increased views of that film. In each case, an alternative view could suggest that it’s not consumer data that’s lead to their eventual success, but influence – slotting into the zeitgeist of the day and prompting expanded discussion, which then leads more people to take a look.
As such, maybe it’s not so much data-driven decision making that publishers should be investigating, but viral dynamics – studies like this one by Steve Rayson which dig into the reasons behind why such trends occur, with a view to fanning your own waves of success.
But that path, which requires understanding modern communication dynamics, aligning with pathways that lead to increased social media traction, and igniting those first sparks in just the right spots, is definitely not easy. It’s worth noting too that Ratcliffe doesn’t definitively suggest that publishers should adopt a strictly data-driven approach, but the risk in adopting such a strategy is that you end up narrowing the field too significantly. Basically, my response after reading Ratcliffe’s thesis was ‘I hope people really liked ‘Boy Swallows Universe”.
At core, my concern with Ratcliffe’s overview is that it looks to cater to the existing audience of readers, as opposed to seeking ways to win more readers back. And of course, the latter goal might not even be viable, but if you’re looking to make data-backed decisions, and the data you use is limited in scope, because the reading market is already narrower than before, then that seems like more of a path to eventual obsolesence, as your audience will thin further and further the more you refine your targeting.
But then again, the alternative is not a lot better – the truth is that people’s reading habits have changed, and with so many distractions available to us 24/7, asking someone to take the time to read a full novel these days is asking a lot. People aren’t reading as widely as they once were – which, as to Ratcliffe’s suggestions, points to the benefits of maximizing those who are.
As an author, and someone who loves to see new voices, that concerns me. But from a marketing perspective, I see the logic.
But I would argue that the Netflix data-driven model is not as solid as many suggest. Again, Netflix created 80 new shows in 2018, and relatively few of them resonated. And you don’t have to take my word for it – go watch them, a lot of them just simply don’t work.
At the rates that Netflix publishes, its track record is not much better than any other publisher, even with its added consumer insight. Which would suggest that the existing publishing models, which rely upon the knowledge of publishers – their ‘gut feel’ as it were – are probably just as good.
It seems less about the data in the acquisition process, and more about the marketing push, using elements like influencer marketing and similar dynamics, to amplify messaging.
But the market is narrowing, as are opportunities.