How Mediahuis used data to boost retention and how you can too

As I have said, we’re in the middle of writing a major report on retention at Pugpig, where I’m the consulting services director. We identified retention as one of the major themes for the year.

The first piece looks at how Belgian-based European publisher Mediahuis runs its data operations. I am a big fan of A/B testing, but Jessica Bulthé, Data Science Business Partner with the group, outlined a few things to consider. In addition to the operational overhead of managing multiple, serial A/B tests, there is also the case that tests can, if not designed correctly, optimise for your existing audience rather than audiences that will drive future growth.

This piece also speaks to the increasing sophistication that publishers are using in their data approaches with propensity and next-best-action modeling. It is a good overview of some of the practices that are being used to balance acquisition and retention.

This conversation touches on a lot of issues including balancing the use of data and creative skills by the editorial staff as well as the competition for product talent. These challenges will be familiar to many, but it’s useful to hear them from a leader at The Economist.

Economic uncertainty is causing some level of anxiety. I’m hearing from publishers that advertisers are trying to delay bigger bets until they have more clarity. But at some point, publishers will have to move. The question for publishers is trying to divine which way that might be.

In this piece, there is a dicussion about “news recommender personae”, which understands that people change as they move through life. Personae need to be adaptable as people move through different life stages and how that will affect their relationships with publishers.

One thing that we’re hearing from publishers as we research attention is about their attempts to maintain a relationship with a subscriber, even after they have lapsed in case their life brings them back.

Of course, the other major shift in personalisation is the application of AI, which promises the ability to crunch a lot of data. It will take a new level of understanding for publishers to be able to understand if the models are delivering.

The New York Times’ sports site is laying off about 4% of its staff. The pivot appears to be about focusing more on league-level coverage and cutting back on coverage of some teams that didn’t attract a wide-enough audience.

When I become a story. My friend Adam Tinworth writes up my recent issues wtih Beehiiv. I think I have tracked it down to having the editor open in multiple tabs on my laptop and iPad. It might also be related to Beehiiv logging me out but the editor seeming to remain active. Either way, it’s not brilliant UI, but I’ll adjust.

Regulation of AI will happen sooner rather than later with other technologies, such as social media. Schmidt’s call is relatively nuanced, as he calls for regulation of things that pose a clear threat such as AI-generated misinformation but a more calibrated approach to issues that are still evolving.

I have to say that this is the first thing that Musk has done since he bought Twitter that is even remotely constructive, at least from a business perspective.