When data trumps journalism judgment
A story jumped out at me last weekend. It was placed in the most prominent position in all of British news: at the top of the BBC website. Its headline read: “Mandelson should never have been ambassador, says Epstein’s victim’s family.”
It referred, of course, to the appointment of Peter Mandelson, an old friend of Jeffrey Epstein, the dead American financier and sex offender, as Britain’s ambassador to Washington by the government. The victim in question was Virginia Guiffre, who took her own life in April.
While her story is a tragic one which deserves to be told, I would argue that Sunday’s story was not one that would have passed muster if editors had been using their news judgment. Indeed, I think it reflects a trend that is harmful to journalism.
That trend is an over-reliance on data ahead of journalistic judgment. I never thought I’d say this. I fought – no exaggeration – for years to get a newsroom to focus on the user data that was available to them. I would count myself as an enthusiast for data in journalism, even an evangelist. But only when used in the right way.
First, let me deal with the story. For me, it fits neatly into the “dog bites man” category. What did you expect Giuffre’s family to say? Of course they didn’t think that a friend of her abuser should have been appointed as an ambassador. It’d have been a story if they did, but they didn’t.
This is where the journalistic judgment comes in. I worked in a newsroom for more than 20 years and there is no way this story would have even got beyond the first pitch to the news desk – and this was a newsroom that was very much not friendly to Mandelson.
Journalists do not report everything that is said to them; they are not stenographers. Instead they work out what is important and report that. The fact that the family made these remarks to a prominent BBC show shouldn’t really have affected the decision.
What I think encouraged the BBC team to give the story such prominence was data that will have shown that stories about Epstein, especially if they mention Prince Andrew (which this one did), perform well, as I know they do everywhere. So the odds were that this story would bring good traffic to the site. On a Sunday morning that might have felt like an easy decision and the site’s editors could be pretty sure the data would back it up.
But ultimately this is corrosive. Readers are not stupid; they can spot a wafer thin story when they see it. And this makes them less likely to trust a site’s judgment over time. In the long run, this is bad for journalism. (This is a topic that Thomas Baekdal alluded to in his excellent – and free – email earlier this week.)
I am picking on a single story here and I acknowledge that my supposition might be wrong. The fact that it was a “BBC exclusive” may have outweighed everything else. But this is a trend I am seeing more often in wider story choices.
For example, British news providers have been consistently overindexing on American stories for a while now. Yes, it’s a big and lucrative market and Trump’s second presidency is a very compelling story, but it doesn’t demand the prominence on UK websites given to stories coming out of the States.
The BBC is doing this too. On Thursday, three of its top seven stories were about America, not including Trump’s state visit to the UK. For the British national broadcaster it feels excessive.
My argument is that you should be informed by data, not slavish to it. You never give up your editorial judgment. I remember an editor objecting to reporters being able to look at story data because “they’ll just want to write about Harry and Meghan”. They might, I replied, but that’s why we have an editor. I think news publishers need to remember that.