Automation in the newsroom – many publishers talk about it, relatively few have any actual experience. Where do you start? How do you generate any real benefits? It may sound like a technical issue, but the tech is actually the most straight-forward part. This blog lays out some of the top success factors for newsrooms deploying automated content and explains how it’s all about improving journalism and growing business.
After four years of getting newsrooms up and running with robot generated content, we have identified a number of best practices from publishers who are now leveraging automation. Note that common to all our cases is that the content comes from structured data sets, which are regularly published. (The data sets we currently draw on include all types of sports, property sales, traffic incidents, weather and local business registrations.)
These are our top tips for how you get the most out of newsroom automation:
Have a distribution strategy. Depending on what data sources are involved, your CMS will potentially receive huge amounts of articles. The volume of content is a key USP of automation, but only if you are strategic in how you use it. There are two main aspects to this. Firstly, the robot articles need to be properly mixed in with reporter generated journalism – don’t flood your site with automated texts. Secondly, and importantly, you’ll only maximise the value of the content if it reaches the right audiences. Distribute the right article to the right user at the right time and through the right channel depending on the use case (e g one of the ones listed below).
Set a clear objective or objectives. By far the important success factor for automation in the newsroom is executing on specific use cases. When deployed strategically, the technology underpins a number of potential business effects, such as generating additional ad inventory, driving engagement and enabling personalisation. Be clear about what the goal of the automation is, e g:
a. Growing ad revenue. By publishing articles on topics not previously covered, e g all divisions of all local sports, you’ll have more inventory to sell to local advertisers. With the extended coverage you also reach specifically interested readers, which means more sellable target audiences and in turn an opportunity to increase CPM levels.
b. Adding value to subscriptions. There are lots of opportunities to develop extra services based on the automated content, such as a follow function for the subscriber’s home team (including in a different country) or favourite player, push notifications about traffic news on a particular route or monthly updates on house sales in a town or village.
c. Driving retention. Swedish publishers have found that robot generated hyper-local sports and property sales coverage increases the relevance they can deliver to subscribers, which in turn drives engagement and thereby retention. That level of granularity and volume of articles would not be possible to produce through reporter resources alone.
d. Establishing a new local site or topic vertical. A new online presence requires a certain amount of content to get off the ground. So if you’re launching a new local site, use automatically generated content as the base (traffic news, sports, property sales, weather etc) and let reporters do the qualified journalism. Or launch a football vertical where you can promise to cover all matches, with robots generating the bulk of the articles and journalists reporting from the top games. In both these use cases, of course, the automated content also generates extra ad inventory.
Get everyone involved. Our robots are an integral part of the newsroom. Below is a perfect illustration of how robot reporting underpins journalists' work – from Swedish publisher Aftonbladet's Stockholm section recently. The breaking robot text (middle) was published automatically/directly with the subsequent reporter story digging deeper (top). Communicate with your newsroom about the benefits of automation and let them feel the positive effects of the freed-up reporter time that it creates.
We’ll cover some of the use cases above in more detail in coming blogs.