Struggling to understand the role of data analytics in your newsroom? Come to #NRS14 http://www.wan-ifra.org/events/13th-international-newsroom-summit …
In part one of this feature Q&A, published yesterday, Tom Betts explained the FT’s mobile-led weekend evolution, which now sees 75% of weekend subscriber traffic coming via mobile devices. In part two, he talks to WAN-IFRA Research Fellow Julie Posetti about the role and ethics of data analytics in news publishing.
Can you tell me about the approach that you take, particularly in relation to the Financial Times, with regards to gathering data from your readers and subscribers, the attitude you have to them, and the way in which you balance editorial responsibilities and ethics against that role?
BETTS: It’s very well understood among the editorial community [at the Financial Times] that we should be data-informed, but not data-led, so I think we believe very strongly that it’s very important that the editorial community are aware of and have the numbers, but that the numbers don’t actually dictate what we do. Because when we talk to our readers, our readers tell us overwhelmingly that, more than anything, they value the judgment of our editorial teams and that’s what they pay for. If anything, they’re pretty terrified of algorithms taking over, and dictating what they read.
The other key principle is that while we have enormously rich data on our customers and because of the subscription relationship that we have with them, we are very clear that the access to that information is very carefully controlled, and that the editorial teams have access to only very, very high level aggregate statistics about their content.
So your journalists won’t necessarily know whether the subscriber ‘David Cameron’ is focused entirely on the food section of the Financial Times, for example?
BETTS: Certainly not. No, everything is aggregated and anonymised.
In terms of your regard for your audience, what kinds of ethical parameters do you operate within when you’re collecting this data? I’m mindful, as I’m sure you are, of the backlash against recent controversial audience research by Facebook, for example. This has lead to debates about the regard that Facebook may or may not have for its audience in terms of disclosure and permissions and so on. So, I’m therefore interested in whether those issues weigh on an organisation whose main work is editorial – the way in which you regard your readers – your subscribers – in terms of your responses to them and your requirements of them, and what kind of standards you’re operating according to, in regards to data collection.
BETTS: I think it’s an incredibly important area, not only because it’s something which is widely debated but because there’s an enormous amount of sensitivity in the information you can collect. I think what’s important to us is that we are as transparent as possible.
When you register or subscribe, in the terms of service for reading, it’s very clear what we collect and why. I think we want to be as honest as possible. What’s really important is if we are to retain the respect of our readers, we need them to appreciate the value exchange which is taking place and that they can get something back. It’s about transparency; it’s about respect.
So transparency you’ve said, and respect – obviously they’re two critical words, and we should probably add the word ‘trust’ – but how do they play out in reality in terms of the relationship between news publishers and consumers?
BETTS: We’re clear in the fact that we’re very heavily reliant on first party data. I think one of the biggest challenges in any media organization that’s running advertising is the question of third party data and enriching the data statistics that you’ve got, and tracking people across various different sites outside of your portfolio. That will become an increasing challenge for publishers. One of the principles we follow is that the first party data is the data that matters. We’ve got unique information that other people don’t have and that is why transparency is so key – because people know that they’re surrendering information because they’re giving it to us at the point of registration. We can be very certain of the information that we are collecting, whereas if you’re entirely reliant on third party data sets, that’s a lot more challenging.
Moving to current debates about the role of data analytics in newsrooms, such as the issues of AB headline testing and clickbaiting, what’s your view on attempts to manipulate the way in which news is being received by audiences?
BETTS: It’s something that we don’t do at the Financial Times but…it’s important to draw a distinction between that kind of testing and being manipulative. With social media for example, you could write a really salacious headline in order to get a click, but if it’s not actually representative of the story then ultimately the reader has a very poor experience. From an editorial perspective, it’s crucial that you’re retaining that integrity.
One of the things which is really key is actually empowering journalists to have access to information about how their content is performing. There’s quite a lot of fear in data. There’s fear from the newsrooms that it’s used as a management tool, and then there’s fear from customers about how their information’s being used. The more people become exposed to it being used in a very rigorous and trustworthy way, you build trust.
I see the way in which journalists are debating analytics as a debate related to social media – it’s all about the way in which we connect with our audiences. What’s your take?
BETTS: It’s a fascinating domain, which is full of fear but I think it’s better to know these things than to not know them.
There are plenty of newsrooms or news organizations where people will just chase volume but because ours isn’t [based around volume], ours is about quality, there is no really obvious clear-cut measure of success. So actually, it’s about allowing people to feel how their content performs. It’s very action-oriented, and focused much more on the operational end of what we do, rather than the commissioning process.
Graph: supplied to the author by the Financial Times. Around 60% of FT subscriber consumption is now via a mobile device, and it is being driven by increases in activity at weekends, and outside core working hours. This is supported by heavy usage of mobile devices during the working week (but outside core working hours), when mobile accounts for more than half of subscriber consumption. At the weekend, subscriber consumption is now overwhelmingly via mobile devices with around 75% of consumption coming from devices.