How this local newspaper lets robots write their real estate articles 

In 2024, 12.418 property sales were registered in Oslo, with an average price of 600.000 EUR. Since this is such an important topic for Oslo residents, it is natural for VårtOslo to closely cover the real estate field.

Vegard Velle, Editor-in-Chief at the norwegian local newspaper Vårt Oslo, says they now achieve this through automated articles created with the help of artificial intelligence.

- In Oslo, great interest is attached to the place where you want to live. Many people attend property viewings, and many dream of finding a permanent place to call home. What sets Oslo apart from other places in the country is its size. Since so many people move here each year and many switch homes, there is also a high number of property sales. 

United Robots Collaboration

Vårt Oslo have automated their real estate section by using a service delivered by United Robots. They help editorial teams free up reporter time, expand coverage and grow business. 

United Robots' Chief Product Officer, Henning Johannesson, talks about the datasets they use for this kind of service:

-  We work with data from Norkart (real estate), Proff (annual reports), and the Norwegian Meteorological Institute (weather warnings). Our services are rule-based and rely entirely on available data. Unlike GPT-based solutions, our texts adhere strictly to the truth. The articles are tailored to our needs and automatically published, with Google Street View images used as illustrations, he says.

Velle further describes the goal they have with their real estate articles and how they work now:

-  We show what properties in the city are sold for, who is selling and buying and how much interest there is in buying a home in the local area in question. We also receive alerts about particularly noteworthy property sales. 

Vårt Oslo uses this automation to draw pictures of the properties that are recently sold from Google Image directly to the articles, as well as they automatically drag in the map with the address of the property from Norkart.  What is additionally automated is a summary of the articles, done with the help of Labrador AI. The articles are short and concise, focusing on giving the reader precisely the information needed, the price, the number of the properties sold in the area and the list with the top 5 most expensive properties in the area.

Collaboration with Labrador CMS

Both Vårt Oslo and United Robots commented on the partnership with Labrador CMS' in delivering of this project. Velle says: 

- To get this service up and running, we received valuable assistance from both United Robots and Labrador CMS, who had to make some adjustments to be able to publish these articles. Funding for the development project was provided by Mediatilsynet. 

On the other side, Johannesson comments on the whole project, describing it as successful: 

- We are proud and excited to collaborate with Labrador CMS. The news site Vårt Oslo is the first to launch our automated content directly via Labrador CMS' solution. With the integration now in place, we can offer all Labrador customers automated texts on real estate sales, annual reports, and weather warnings. Our content supports editorial teams in achieving digital success. With the integration now in place, the content appears in the CMS just as if it were created by journalists. 

Velle concludes by mentioning their overall experience with the service: 

- We are now very pleased with how the service operates. We keep a close watch, but the articles require little additional work from us. We see that many of our readers engage with these articles to learn more about the sale of their own or their neighbor's house, and we have succeeded in creating a more comprehensive local newspaper.