How much did your new neighbour pay? Who bought the house you couldn’t afford? This is of interest to most people and, as we saw in our data, of exceptional interest to MittMedia’s readers.

MittMedia wanted to start producing articles about every property sold, but it wouldn’t make sense for our reporters to do this. So we built The Homeowners Bot. This productive bot is now our most efficient “employee” with more 10,000 articles in four months. It’s also the most read type of article by our paying users, and it has converted hundreds of users into digital subscribers.

If The Homeowners Bot were a human journalist, it would by far be our most effective employee at a fraction of the cost:

• During the first four months we were live with the robot, it has written 10,000 articles.

• The Homeowners Bot is No. 1 when we sort the number of behind-paywall pageviews by byline, outperforming all human journalists.

• More 300 users, and that’s a lot, have been converted into digital subscribers from a Homeowners Bot article.

During fall 2016, MittMedia launched its revised paid content strategy, and it has been a huge success. This was not the first or even the second try to implement a sustainable digital subscription business model. The difference this time was a solid foundation in data and a focus on the content needs of readers.

In our data analysis of what type of content that make our loyal users convert into digital subscribers, real estate always stands out. For decades, Swedish newspapers have published lists of property sales that include sellers, buyers, and prices. The layout of the page looks like something out of an old product catalogue, but despite its unexciting format, this is something our print readers don’t want to miss.

When we started to publish this material online more than 10 years ago, we didn’t really do anything to take advantage of the digital format. Again — long lists with addresses, names, and numbers. It worked fine and had high readership, but it was clear that the potential was greater.

To meet the digital demand, we needed to be innovative. How could we better cater to the user’s individual interest in house sales?

Since more than 100,000 homes are sold within our geographical area every year, it wouldn’t be practical or make economic sense to let reporters write articles of every sale. So The Homeowners Bot was born to generate these articles automatically.

In September 2017, we started publishing articles of every property sale in our regions. Together with United Robots, Scandinavia’s leading robot journalism company, we developed a simple yet efficient automation. United Robots, founded by MittMedia and two other partners, has built a powerful text automation platform that MittMedia’s editors used to develop The Homeowners Bot.

The bot’s most important feature is selecting the right angle. By analysing the property data, provided by Sweden’s Land Registration Authority, the bot finds the most interesting angle and writes a headline.

Knowing how important images are for user engagement, we also developed an image service by using Google Street View and algorithms to cleverly select the best image available.

The Homeowners Bot articles are distributed by all of MittMedia’s local news brands. Articles are published as any others on their Web sites. Since we use an algorithm to optimise the user’s experience, the most interesting articles are displayed first.

Homeowners Bot articles often are among the most read on the site. Whenever there is a new house sale article from a geographical area of your interest, readers get a push notification from MittMedia’s mobile app.

We are transparent about the fact that the article is automated. The byline on each article is “MittMedia’s Text Robot,” but users do not mind. In a survey, two out of three didn’t notice this, and those who did notice didn’t care.

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