Paul O’Mahony is Managing Editor at The Local Europe, an English language digital publisher based in Sweden which serves the international expat community. Since its founding in 2004, The Local has been rolled out to 9 countries in Europe and, when we spoke to him on stage at this year’s Digital Media Europe conference in Copenhagen they had, that very same day, rolled out their membership initiative for The Local France.
Morning, Paul! We’re really interested to hear about what you’ve been doing at The Local, particularly about how you operate in this era of easier access to data and analytical insights. What’s the data culture like in newsrooms, in your experience?
I think across the industry it’s changed – and it’s changing. When I started out data culture was pretty much non-existent beyond following headline metrics like unique visitors and pageviews, but now there’s cross-industry consensus that you ignore data at your peril.
At The Local, we are digital natives which gave us a headstart on traditional media when we launched 14 years ago.
So you’ve always been a digital-only endeavour?
That’s right. Because of this, we have been able to focus fully on digital metrics without worrying about the transition from print and were able to grow quickly in Sweden before expanding to eight other countries.
We now have ten journalists reaching five million users a month. That basic data point is something all our journalists are aware of, and we always examine data for each individual site in our editorial meetings and strive to learn from what the numbers tell us.
The biggest problem for us has been extracting data that tells us more about user behaviour on a given article beyond just pageviews. We want to be able to see at a glance what articles are achieving in terms of engagement, read time and loyalty – essentially anything that tells us how much readers value the content we are producing beyond the first click.
So you’re trying to find more actionable insights than so-called vanity metrics?
Yes – although I think the industry at large is coming to this realisation, so I don’t think we’re unique in that. Especially now we’re seeing more of a pivot to reader revenues, I think publishers need to understand how their content is performing much more deeply – it’s vital to know if and how readers are prepared to invest more in your publication.
What strategies had you been trying at The Local?
Over the past couple of years we’ve been undergoing a cultural shift towards greater depth. We’ve emphasised to our journalists that we’re focusing on quality rather than quantity.
In practice this means staying true to our core values. Our audience is primarily made up of pioneering foreigners working and living in a new country. Our main duty as a publication is to serve them by producing content that helps them with the practicalities and deepens their understanding of their new environment. With this in mind we now ignore a lot of stories that we know from experience would give us clicks, but don’t fulfil that central function.
We had anecdotal evidence to suggest that readers appreciated this approach and we saw some metrics to confirm this but we felt we needed more data both to confirm our hypotheses and to help guide editorial decision-making.
Speaking of which, congratulations on the launch of the membership program at The Local France! It must be exciting to be rolling that out. In moving towards a reader revenue model (at least in part), have you been able to draw any lessons from other publishers who are already reporting success with this approach?
Yes we have. We’re learning a lot from looking at the Nordics, which is the market we know best.
To take some examples, the Norwegian media house Schibsted is using digital behaviour data and predictive scoring to segment its subscriber newsletters, resulting in higher opening rates, click rates and ultimately better retention of subscribers.
In Sweden, Dagens Nyheter is successfully using a research engine to identify why subscribers aren’t renewing. This has helped them bring their churn rate way down.
And again in Norway, the country’s biggest newspaper Verdens Gang harnesses user-generated data on major stories to enable readers to help each other, forging stronger connections with the brand in the process.
We’re deriving a lot of inspiration from all of these examples.
Why is monitoring user behaviour so important during this transitional period for you?
As I mentioned earlier we needed analytics that confirmed or disproved our editorial hypotheses and helped inform our decision-making process. The insights we get from Content Insights have helped us enormously in this regard. By and large we can see that the more in-depth content we’re producing is working. Conversely we can also see which articles are not scoring well on metrics like engagement and loyalty.
It’s also easy to look at things like read depth. We can see for example that 25% of readers have reached the end of an article offering tips about buying an apartment in Sweden. That’s high, and tells us that the article was valuable to a lot of readers.
We can also look at how much time readers have spent on an article relative to other articles in the same section.
Premium articles by their nature are going to get fewer views so it’s essential for us to be able to extract meaningful analytics about how readers are interacting with articles behind the paywall. Understanding what loyal readers want is crucial as we roll out our membership model in other markets.
On The Local Germany, this story of the depression that can follow from culture shock in a new country got more people to subscribe than any other:https://t.co/d7W16yzOj0
— James Savage (@SavLocal) April 24, 2018
Has anything surprised you about the way your content is performing since you’ve started to look at analytics more holistically?
I’ve been surprised to see just how well our more in-depth content is performing. It was a risky strategy for us to devote more resources to content that might not necessarily travel well internationally – especially as we operate with a lean editorial staff and I guess we expected that loyal readers would appreciate the change, but I’ve been amazed to see just how much more time readers are spending on these articles.
Can you give an example of this?
Absolutely. Sweden has an election coming up this year and we had an article a few months ago profiling the party leaders. It performed reasonably well in terms of traffic but when we looked at the more holistic analytics on Content Insights we saw it was one of the best performing articles of the last few months. On closer inspection it turned out that attention time on this article had been 11 times higher than the section average. That kind of insight is invaluable and it’s something we might have missed before.
Has that made you reevaluate your editorial strategy?
It’s good to know that our editorial instincts have been vindicated and it’s also useful to see that this kind of coverage is well received. If a primer about a national election works in Sweden, there’s a good chance it will be of interest to our readers in France or Germany when that season comes around. The more we know about our content, the more we are able to learn about our readers and how we can deliver the best content to them.
Best of luck with the continued rollout, Paul!
Many thanks to Paul for his time and insights. If you’d like more information about how the Content Insights tool could help your newsroom, or to arrange a demo you can email David or Mario in our sales team.