Journalism never really was detached: A case for using analytics in newsrooms to build relationships and listen

Web analytics is at the heart of a cultural rift in newsrooms. Here’s one way to think about why it might be useful.

The culture war over analytics in newsrooms is back.

A few weeks ago, Chris Moran (an editor in charge of strategic projects at the Guardian) called here on Medium for better thinking on technology, data and journalism.

Moran’s post was a reply to Franklin Foer, who, writing in The Atlantic, placed the blame for the failure of a digital rebirth at New Republic on, among other things, an oversized focus on web analytics driven by a Silicon Valley culture. Moran, critical of Foer, implied Foer’s take on analytics was simplistic and that his piece played down wider editorial failures at the New Republic.

Whatever happened at the New Republic, the exchange revealed, again, a cultural rift between those emphasizing editorial autonomy and those emphasizing analytics (summarized well here by David Cohn).

The pessimistic thinking is well summarized by Foer in his piece:

“But the pursuit of audience is their central mission. They have allowed the endless feedback loop of the web to shape their editorial sensibility, to determine their editorial investments.”

The cultural rift is centered around editorial autonomy and its (awkward) connection to the audience.

Is the editorial process characterized by an outsider’s position — of being on the outside looking in, stories driven by professional identity and an (intuitive) grasp of journalism’s core mission of adding value to public life?

Or is it intertwined with the public and its interests from the start?

Is journalism there to teach from the outside, serve from the inside, or both?

If both, in what proportion should they be applied? This is an age-old cause for a journalist’s anxiety.

At it’s core Foer’s piece is an argument that the proportions now are wrong. But, as Moran points out, it’s also not a question of either or, but a question of building better understanding around data and better metrics.

I’d add it’s about resolving some of the underlying ethical issues on a more profound level. It’s about how we talk about analytics in general.

Having worked for years on implementing analytics in a national newsroom and now looking into this from a more teaching perspective (I’m a news editor at Helsingin Sanomat now working for a year as a visiting professor in journalism at the University of Tampere), here’s a few suggestions.

1. Journalism was always about relationships. Measure those, like we always did.

Relationships always were key to journalistic enterprise.

In legacy media like newspapers or magazines a relationship with the audience was and is mainly built as a function of periodicity.

You read the paper today, send feedback in and perhaps the reporter responds or the paper prints your view tomorrow in the letters section. If you don’t like the coverage, you might phone the editor. If you really don’t like the coverage or find it irrelevant to you, you no longer subscribe or tune in. If you do, you slowly get to know the topics and writers, perhaps even got attached to a shared sense of identity. You build a relationship with the publication, and periodicity is the key.

This is what Foer is referring to when he talks about New Republic as being “close to a cult, catering to a loyal group that wanted to read insider writing about politics and highbrow meditations on culture”. Seems he really knows his audience. And yet he ends up rejecting the tools that enable this very thing in a hybrid digital environment.

But why do we need new tools in the first place? Because periodicity is going away. In the digital media environment, community and journalism are distributed in a variety of places.

The readership, by and large, does not return to a publication every day, read it through and reflect on a shared identity.

Shares, social media groups, notification streams, newsletters, native social media content and a myriad of other things work together to bring in the readership, and the ways and means of reading are a multitude. The relationship between the audience, the reporter and the publication is much, much more complicated than before. Analytics can help understand how all this works.

But it would be wrong to just point at measuring relationships and say that there really is no intellectual change. There is, in fact, fundamental change.

In a legacy print publication distribution and the reach of journalism generally speaking was not a reporter’s headache. Acquiring circulation for a magazine or newspaper was, most often, a job for the marketing and sales departments. Now, worrying about distribution and reach is front and center in newsrooms.

This, according to Foer, has taken away a sense of “detachment” that protected editorial autonomy:

“Journalism may never have been as public-spirited an enterprise as editors and writers liked to think it was. Yet the myth mattered. It pushed journalism to challenge power; it made journalists loath to bend to the whims of their audience; it provided a crucial sense of detachment. The new generation of media giants has no patience for the old ethos of detachment.”

It is true that to a reporter, the connection between content, distribution and circulation used to be muddier. You got circulation numbers, perhaps the results of a reader survey and that was it. The logistics of it all were largely hidden.

But the “myth” that drove “detachment” really was no myth at all. When talking about commercial journalistic enterprise, the fact that the publication had circulation or an audience in the first place was always inherently linked to the content. Put bluntly, it’s impossible to sell an irrelevant paper.

For example, it’s difficult to imagine a local newspaper completely “detached” from the information needs of the local residents. If they told you the local hospital is in trouble should you listen to them or not? Investigate? Act in their interest, find the facts, be fair but serve them? Then listen to their feedback, let it inform your coverage and dig deeper, still fairly? Perhaps you’d listen to them when you met them at the local market, while parking your car or at the supermarket? At what point does the “detachment” from the audience begin exactly?

Thus, running a commercial journalistic outlet and keeping it running always was (also) about the quality of your reporting, the breath and focus of your coverage.

A sense of community is at the heart of local journalism especially, and to some extent national writing as well. Listening to readers never meant “bending to their whims”, it meant taking their needs into account when setting the agenda and then doing the journalism part right. And yes, if there was “a whim” (whatever that is) you might investigate that as well, in a fair and balanced manner and show what is true here and what is not. You speak truth to power and to your audience. But that, as well, is a thing you do in a community. And analytics, if used right, can tell you if you are succeeding in that or not. If not, it can help you understand why not.

Being a part of a community also meant that if you screwed up somehow or became irrelevant, you lost your audience. And somehow, reading those letters to the editor or taking the phone calls or just by listening to people you measured how you were doing. Somehow you measured if you still mattered. It’s a different thing if you cared. You might not even care today. But pure “detachment” really was a myth.

In fact, the pessimists often have a bleak picture of not just analytics but the audience as well. They basically propose a standard critique of fast food consumerism.

Open the gates to demand driven models and everyone will just eat hamburgers and candy, the thinking goes. There are always barbarians at the gate. Opening the gates leads to a race to the bottom where the only thing left for journalists to do is to churn out more hamburgers and candy. Do not open the gates. Detach. Barbarians. Detach.

Again, black and white thinking.

In recent years, analytics have been widely used in all major news outlets in the West. The result has not been a universal race to the bottom but something of a renaissance in subscriptions in quality papers. Online subscriptions are fast gaining ground in the United States, here in Finland as well as elsewhere in Europe. Analytics is a key driver in this transformation.

In fact, the fact that distribution now is a headache for the newsroom is not necessarily a bad thing at all. One could even make the argument it gives the editorial side more control than ever before.

And yes, analytics has also been a driver of clickbait. It helps identify the latest hot topic to pound new stories on, identify groups to pander and helps “optimize” headlines to maximize the audience.

But this really is not about the tool but about the sort of publication you are making and what you are trying to achieve.

If you’re building an ad-driven business based on a horizontal maximization of audience, like the New Republic apparently was doing, then yes: chasing “whims” and the latest hot topic is a strategy you might deploy — and even see impressive short-term gains. But if you’re using analytics to see that Trump stories attract a lot of readers, you’re really not using analytics at all.

But if you’re building a vertical relationship with your readers that lasts longer than a few seconds a week, there are far more nuanced strategies to deploy. They are based on measuring relationships, one way or the other. Here’s a few examples, from the very basic to a little more advanced.

2. Look at the entry points

To understand a relationship in the digital environment, you have to understand how and why it starts.

Analytics can show you where your audience is coming from. This most probably varies by topic and can tell you a lot about the ways people are starting their relationships with your publication and why. For some, this might help identify your community on social media. For others, this might show how the publication is integrated to the daily routines of the readers.

And yes, look at if the audience in engaging with an individual piece of writing in a way they usually do. If they’re not, there might be something wrong. This where you might want to look at the headline and promotion in the places your audience usually encounters you. But stay you: overdoing the selling in the headline just weakens the relationship with your core audience and makes you less distinctive. Remember, relationship is the key.

3. Look at engagement

The simplest step away from clicks is time that the reader spends with a story or engaged time. This is in fact the basic unit of measurement in Chartbeat, the haunting tool named in Foer’s piece.

Analytics 101: to identify the things that strengthen the relationship with your readers, look at what they are reading the longest. Don’t look at what they click on the most — though the two things are surprisingly often correlated.

Yes, this does help identify stories on which to sell subscriptions or newsletter signups. And yes, it’s a decent tool in understanding the interests of your community or readership and focusing on that. But it really is just a start.

4. Look at audience patterns, not just individual stories

Foer’s description of what went down at the New Republic paints a picture of a newsroom focused on page views, real-time audience metrics and individual stories. A focus like this easily leads to a pattern of generating hits around a few trending topics, this year most likely the latest Trump controversy.

To get past that, look at more general patterns of audience interests and behavior. Look at “sub-audiences” and what connects them. Is there a community that’s interested in both culture and local traffic? Perhaps take that into account on social media. Look if you can combine your stories into new products like newsletters around these distinct groups. Are you serving them were they are?

Also, look at the paths the groups take through your product. “For the Guardian, whether or not a person’s visit includes a trip to our home page is a huge indicator of loyalty and propensity to pay”, says Moran. What’s keeping them from building a deeper relationship? Can you do something about that?

5. Look at the value in what is not clicked

News use patterns vary from country to country. In Northern Europe, for example, the dominant way to access news is to use a news app directly or navigate to a news site’s front page manually in a web browser. In other countries, newsletters or social media are bigger drivers of traffic. But to build relationships and value, having a valuable platform of your own is vitally important. Remember, front pages are important for the relationship.

But don’t just look at what people click on the front page.

We know intuitively as well as from research that news use especially on front pages is much more nuanced: the stories that are not clicked on also carry value in defining the product. While scrolling through a news site, people get a lot of information from just the headlines and snippets, perhaps only clicking through to read one story.

Yes, some of the things that are not read might not be vital at all. To some, they might be the most important thing in the relationship. Hard to tell without analytics.

But it is wrong to assume that all the stories that are not clicked on carry no value at all, and this is especially true for news. Even little-read updates on foreign events might be crucial in building a sense of a news product that brings readers back as a part of their routine.

To understand these dynamics, one might look at the paths users take through a news product, ask them what brings them back to the front page. But again, just looking at page views you really get no such insights at all, and you end up harming yourself in the long run.

6. Look at qualitative indicators

What’s the place and impact of your reporting in the community? Did today’s report lead to a lightbulb moment? Did it lead to introspection? Did it drive understanding or further questioning by the audience? Does it result in policy change?

While measuring impact is hard, it’s easier in the digital environment. One way to measure (short-term) impact is by looking at share messages on social media, a practice that’s becoming more and more common. This means looking at what people are saying when sharing your stories and what sort of context they place them in. This is much like sitting in the town square, but just hearing more.

Just don’t look at the absolute number of shares, as we know from research that the people that trust the news least and are more suspicious of the media tend to share and comment more.


Listening to people is key to relationships. Analytics helps you listen.