A reader with fifteen minutes on a tram and a reader preparing a meeting are not failing your homepage. They are asking for different shapes of the same reporting.
The Reuters Institute’s January predictions report gave this a name, liquid content: stories that adapt to context instead of sitting as one finished object. FT Strategies’ note on the idea makes the production point. Publishers currently mould a story, dry it, and ship one form. Liquid content keeps the underlying material workable, so the same reporting can become an audio briefing, a short selection, or a sourced answer without starting from zero each time.
That is a change in how news is consumed, and it is also a change in what converts. People do not abandon a newspaper only because the journalism is weak. They abandon it when the format does not fit the moment, and the moment is now often a phone, a commute, or a question. Audio for the ride in. A cut of the day rather than the whole site. A query across the archive when the job is to walk into a room prepared. Les Echos was already testing pieces of this last autumn. The pattern is not a chatbot pasted onto the homepage. It is a small number of formats, each with a job.
Two constraints keep this from becoming a sloppy feed.
First, the journalism stays the source. Reformatting is not a licence to add facts the reporter did not establish. If the audio, the summary, and the article disagree, the reader has been given a reason to leave.
Second, personalisation has to leave a shared core. A newsroom that shows every subscriber a different front page stops being a publication and becomes a set of private channels. The useful version is a common selection, plus a layer that meets the person: interests on the app, a format switch, a professional tool for the people who use you to work.
The economics are the quiet part. One reported story that can be entered in three honest ways will be finished more often than three separate productions of uneven quality. Unread work is usually a format problem before it is a demand problem.