MittMedia, Sweden’s biggest local media group, is working with increasingly limited editorial resources. The news media company recently spotted and seized an opportunity in what some call “robot journalism.”

It started with a sports bot. MittMedia identified a way to cover all local sports and all leagues all the time, at almost zero incremental cost — and soon led to automatically generated articles about real estate, too.

Now MittMedia drives engagement with loyal users, as well as subscription sales, through its automated editorial content. This also frees up newsroom resources so reporters can focus on producing quality local journalism.

MittMedia is one of the co-founders of United Robots, a company that automatically turns big data sets into editorial content using Natural Language Generation. Since the launch in 2016, several other Swedish local media groups use the service to publish sports text as well as news on local business registrations.

MittMedia’s editorial sports team’s strategy was simple: With the sports bot producing all the pre- and post-match stories, reporters could spend their time on more in-depth journalism, more video content — freeing up resources to do more live streaming for lower league football.

This in turn has meant that rather than robot journalism reducing the number of reporters, MittMedia actually has been able to hire additional sports reporters — because 40% of paid subscription conversions now happen through sports content.

The results have been remarkable. It turns out sports is the highest converting content of all on MittMedia’s sites. From September 2017 to September 2018, the sports bot generated 41,000 articles and 1.2 million logged-in pageviews (2.5 million pageviews in total) for all of MittMedia’s news sites.

After the success of the sports bot, MittMedia started looking at other robot journalism opportunities. Lists of property sales from the Land Registry generated a lot of interest in print and online, and the company began testing generating articles based on the data.

In September 2017, MittMedia started publishing automated articles, including images, about homes sold in all local areas. A year later, more than 25,000 of these articles have been published, generating 2.5 million logged-in pageviews. The new property content engages existing subscribers and helps drive conversions — more than 700 subscriptions were started from the 25,000 articles.

According to Li L’Estrade, head of content development at MittMedia, readers don’t really distinguish the automated content from articles written by reporters, even though the company is transparent about what is what.

“A really good robot text can have a bigger impact and be more read than a really good news article — but only if it’s a topic readers really care about,” L’Estrade said. “On the whole, it is the volume of articles that is the robot’s big advantage. Each article reaches a smaller group of readers on average, but in total we get an exchange on par with anything written by our most read reporters.”

Another key advantage is the freeing up of editorial time.

“The robot produces more routine type of material, allowing our reporters to focus on getting out, meeting people, and producing really quality and unique local journalism. It’s a win-win-situation for both readers and reporters,” L’Estrade said, adding that the huge volume of articles produced by the robot does not generate value per se. Staff must develop functionality and products that provide value to users. “The point of the automated articles is that they are pretty niche and are super interesting to a small number of people. That makes them perfect for a personalised news feed or a niche product. It’s key to nail the context.”

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .