Social Rocket
Publishers are increasingly relying on personalised newsletters, because search traffic is weakening and direct reader relationships are becoming more important. Yet new experience reports from the Digiday Publishing Summit show: personalisation does not automatically increase clicks or revenue. Some tests show measurable uplifts, while in others performance stays the same – and mainly editorial work disappears.
The important catch: the examples come from a closed publisher round held under Chatham House rules. The companies are not named, so the figures are not general benchmarks. For marketing teams they are still interesting, because they show which question is often asked wrongly when it comes to personalisation.
20 percent more clicks – but not everywhere
According to Digiday, one publisher reported early tests with individually personalised local daily briefings. For some registered users, content was selected automatically based on their previous reading behaviour. In one early test, the click-through rate rose by 20 percent, according to the company.
Another publisher automated a newsletter that had previously been curated by editors, after a staff change. Content was assembled using an identity graph and individual reading behaviour. The result was much less spectacular: open and click rates remained practically unchanged.
The company nevertheless counted this as a success. The same referral performance was achieved with less editorial effort.
This creates two different business cases: personalisation can generate additional usage – or produce the same usage more cheaply.
Personalisation is not automatically differentiation
The second case also shows the strategic problem. If an automated newsletter performs just as well as a curated one, this does not necessarily mean the personalised newsletter has become more valuable to readers.
At the summit, one publisher voiced the concern that personalised newsletters could increasingly become a basic feature rather than a unique selling point.
The counter-example among professional audiences is particularly interesting. There, one publisher reported that readers actually did not want a strongly individualised selection. They wanted to see the same topics as colleagues and competitors, so as not to miss anything relevant.
For B2B brands this is an important point: a shared information baseline can itself be part of the product value. Too much personalisation can then even work against the reason people use it.
The bigger lever may lie in the funnel
The most exciting thesis from the publisher round therefore concerns not the subject line or the order of headlines, but access and conversion.
Some publishers are experimenting with treating newsletters differently depending on funnel status. Registered users could first receive more freely accessible links and later be confronted more strongly with paywalled content. Engagement data would then determine not only which article is shown, but also when a user is presented with a paid offer.
For marketing teams this is far more relevant than pure content sequencing. Personalisation thus becomes a question of lifecycle, access and monetisation.
Precisely because search traffic is becoming less reliable, publishers are rebuilding their direct audience systems. We have already looked at how editorial teams are preparing for this shift in our piece on post-Google audience teams and AI search.
What marketers can learn from the tests
Personalisation needs a control group. An automated feed, newsletter or content hub should not only be measured against its own figures from the previous year, but against a cleanly defined non-personalised variant.
Teams should look at at least four levels separately:
- Engagement: open rate, click-through rate and active use
- Efficiency: editorial effort and production costs
- Conversion: registration, lead, subscription or purchase
- Retention: repeat use, churn and long-term user value
A higher CTR is only truly relevant if it fits the business model. Conversely, an unchanged CTR can be economically attractive if production becomes significantly cheaper.
This also applies to other direct distribution channels. With LinkedIn newsletters, for example, it becomes clear that not only the current issue counts, but also the long-term discoverability of older content.
Direct reach is becoming more valuable
The background to the personalisation debate is structural: publishers are losing reliability from external traffic sources. Google search is changing through AI Overviews, referral traffic is becoming more volatile, and many media companies are investing more heavily in registered users, newsletters and their own apps.
That some publishers are now even buying back search traffic reflects the trend towards rising paid search spending. More on this in our analysis of publishers in the zero-click environment.
For marketing teams this means newsletter personalisation should not be viewed as an isolated feature. It is part of a larger question about how first-party relationships, data and direct distribution are used economically.
Personalisation is a test, not an end in itself
The publisher experiences so far do not deliver a simple "personalisation works" or "personalisation isn't worth it".
Rather, they show three possible outcomes: more engagement, the same performance at lower cost, or in the worst case a more interchangeable product.
The most sensible strategy is therefore not to assume personalisation is a mark of quality. It must be tested against editorial curation, segmentation and simple standard delivery – using the metric that actually earns money or builds loyalty.