Tagged: research

How metrics change journalist behaviour

There is a fantastic piece over at Nieman Labs by the very talented Angèle Christin, a ‘newsroom ethnographer’.

She has just spent two years studying the habits of journalists at a publication in the US and a publication in France, comparing their claimed vs. actual behaviour when writing copy for websites. Specifically, she looked at the effects of web metrics within editorial departments.

It’s a fascinating read, touching on some characteristics of metrics that are true for all of us (As Karl Pearson said, “That which is measured improves”) and some that are specific to journalists (I’m not sure if there’s a specific term yet for ‘the temptation to knock out a quick listicle when your weekly traffic is not looking great’).

That said, it is not clear from the article whether the news sites Christin worked with generated revenues purely from advertising, or whether they also had a subscription or pay-wall element to their business. This matters, because it changes how journalists’ effectiveness is perceived within a business. (full transparency: I have not read the full academic paper, only the summary article)

On a site where revenue only comes through advertising, the temptation is to see the journalist’s role as simply generating content against which advertising can be sold. On this model, the content itself matters very little, as long as it is attracting plenty of people (ideally of the right demographic). The MailOnline is arguably the most extreme example of this model.

On sites with a more balanced revenue model (e.g. the FT, or The New York Times), the business is less dependent on ad revenue. This means that articles can be judged not just on how many people read them, but for example on whether the content was of a high enough quality for readers to think their subscription was worthwhile, or want to come back again next week. If a site is writing about very specialist subject matter (anything ranging from complicated hedge fund analysis to esoteric model railway design), the value that some readers will place on the content far outweighs the possible advertising revenue generated, so the dynamics could shift significantly.

While these are the main business models in modern news sites, you could also imagine other metrics being used to judge performance of journalists, for example a new publication trying to establish itself in the market might evaluate journalists on how many times their articles are picked up by other sources, or quoted by other influencers, rather than pure page views.

Indeed, some publishers such as the FT are now moving to selling advertising to a cost-per-time spent basis, which will arguably change the pressures on journalists again.

It would be fascinating to see how editorial teams responses to these metrics, and the pressure they place on journalists, vary according to the business model of the publication.

None of this is to say that page view metrics have no place in the editorial process – they are the easiest way for the publication to keep on track with what readers are interested in, they help keep things profitable, and ultimately contribute to many journalist’s feelings of accomplishment. No one wants to write ‘just for clicks’, but at the same time every professional journalist I’ve ever met has wanted their work to be read by others. Otherwise they would never publish.