Content Recommendation widgets - Editorial teams hate it, news audiences hate it, even the 'business' teams at digital news organizations preempt arguments about it by focusing on revenue streams and settling them a spot on the news story page, where unimportant information will be buried.
Recommended stories are not for everyone. So, does it make sense for your news website? Here are 3 questions you should consider, before you make up your mind.
- What kind of stories do you publish?
Websites that publish stories aimed at information or news typically have a diverse range of topics and various genres of stories such as opinion, analysis, reporting etc. In contrast, publications that focus on entertaining readers (not the entertainment industry, but the expected reaction in audiences) might have largely unconnected stories such as pictures of cats, funny videos and gags.
The nature of stories you publish, will influence the likelihood of your audiences to be interested in other stories as well as the variety of stories that you can offer to them as recommendations.
As a rule of thumb, recommendation widgets that exclusively use only your own content will work better for information oriented publications, while entertaining ones might benefit from serving up a wider variety of stories from around the web, unless of course you post a significant number of new stories every day.
So your choice of content recommendation engine should certainly be informed by the type of stories you create.
- How does your digital publication monetize traffic, or what is your business model?
In a previous post, we did a primer on Content recommendations for digital publications where we covered the reasons you might want to add a suggested content section to your website. Third party content recommendation systems which use stories from across several publishers work in a fashion similar to an ad network, therefore providing you with a revenue stream.
You might already be reliant on advertising revenue and/or subscriptions. In addition, you might have other monetization options that you already use, but you will need to decide on whether the incremental revenue from adding a (sometimes tacky) widget with external content and headlines of debatable quality is worth the degradation in user experience for your audience.
At NativeAI, we would always recommend that you rely on your own library of stories for your recommendations and monetize the improved audience retention that you get, due to higher CTR and re-circulation on your websites. This would work regardless of your monetization model - whether it is advertising or subscriptions.
- Which content recommendation engine do you use currently, or can integrate with minimal effort?
Experiments and A/B tests are crucial to achieve growth in digital campaigns. You need to be able to establish a baseline for metrics such as Click Through Rate (CTR), recirculation, story reads per session and average revenue per session reliably and quickly before you try different content recommendation engines.
If you already use a platform for story recommendations, start by learning more about how the engine works - the inputs it considers before making a suggestion, and compare the quality of the output with what an editor would recommend. Also consider making changes to the design of the widget, size of the images, position of the widget on the page etc.
If you do not currently use a recommendation engine, consider NativeAI's recommendations platform that uses natural language processing and advanced artificial intelligence to identify the best stories that each reader may be interested in, based on user attributes as well as their reading history.
Choose a recommendation engine, that gives you a wide array of templates or themes to pick from, or a high level of customization with easy CSS modifications. Always keep an eye on your publisher analytics reports to compare the performance before and after making these changes.
Recommended stories are a powerful way to attract and keep your audiences engaged on your website. You should consider using such tools to delight your readers, while prioritizing user experience, relevance of suggestions and the traffic numbers that demonstrate the value of a recommendation service.