The Swiss media group is launching a new experiment with a smart, personalised newsletter called “My NZZ.” Catering to its German-speaking audience, My NZZ automatically curates a list of articles that the reader missed during the week, personalised to each reader’s individual interests.
Why it matters:
Neue Zürcher Zeitung (NZZ) publishes up to 200 pieces of content every day, and with subscribers’ busy lives and the sheer volume of written and visual journalism produced, its readers see only eight of those pieces, on average, on any given day. The My NZZ newsletter aims to create a smarter news experience for those readers, by showing them what they missed on topics that are most relevant to them.
NZZ worked closely with its Data and Product teams to learn what features its users would most like to have, and found that “catch-up” features were rated the most useful; and that readers would like to receive those via email.
Two core factors drive the My NZZ personalised recommendations:
- Editorial Score: NZZ editors are in the driver’s seat, curating the content by position and time on the homepage.
- Personal Score: based on personal interest and behaviour of the reader.
NZZ outlines this algorithm design strategy in more detail here. Creating this personalised experience involved significant technology, particularly with the email inbound system. Another hurdle was the fact that news content expires quickly. NZZ chose to start with a weekly email that goes out each Friday afternoon to readers, provided by a general purpose RESTful API.
- The My NZZ email newsletter displays five articles, catered to the mobile devices that most users read it on.
- NZZ started with the final email design in mind, and then worked backwards.
- The recommender API serves the article ranking for each user, that is then enriched with metadata before being sent.
- Using a cloud-based email provider, NZZ uses a two-feed option — one contains the metadata for each article, and one contains the article ranking for the individual user.
What NZZ Learned:
Enabling personalisation is a complex task, and NZZ is still heavily experimenting with it. Some of the key factors are:
- Having the right MVP and strong collaboration across many disciplines and departments.
- Creating “Trust Features” which will give our subscribers more transparency and more control regarding how our news personalization works.
- Finding clever ways to integrate personalized experiences to the core of our news products — with editorial integrity in mind
- Closely monitoring how the new email product is performing.
Once NZZ has a better handle on how subscribers like My NZZ and the value it brings to them, as well as internal systems performance, it will fine-tune the product. The company has placed a high priority on being led by the data and its readers, to act fast and collaborative.