Social Rocket
The US publisher The Verge wants to let only people who have registered for free onto part of its readership on the website, starting in October 2026. Digiday reported this on 8 October. The test comes at a time when traditional search referrals are becoming less reliable. The idea behind it is not primarily to sell subscriptions straight away: a direct relationship with recurring readers could be more valuable than anonymous one-off visits.
Registration is not the same as a paywall
With a registration wall, a free user account is enough. A paywall, by contrast, requires payment. Both create barriers but pursue different goals. According to Esther Cohen, The Verge's head of audience and subscriptions, the publisher wants to examine whether users arriving via high-reach but so far little subscription-prone channels can be bound more strongly. For the test group, a hard barrier is planned: no account, no access.
The publisher is considering several variants, such as a share of visitors or selected referral channels. Which variant will prevail is not yet known. Reliable results from the test are also not yet available.
The real KPI: value per reader rather than page views alone
Registered users can be won over for newsletters, personalised offers and later a subscription. Beyond that, publishers get a more consistent basis for their own audience segments. This can be more attractive to advertisers than reach alone.
Digiday quotes an anonymous publisher manager whose ad rates for registered users were at times up to eight times higher. This is a single statement from an industry forum, not an industry-wide average, and above all not proof that registration alone causes this difference. Registered readers may already consume more intensively and regularly.
Other experiences argue against blanket promises of success: a food publisher told Digiday that its registration numbers were disappointing, while a paid offer it introduced later performed better. Still other publishers see no subscription prices that would economically offset the loss of reach.
A meaningful evaluation would therefore need to consider registration rate, drop-off rate, recurring usage, consent rate, ad revenue per user and total revenue together – and ideally compare test and control groups.
Why AI search changes the equation
The Reuters Institute surveyed 2026 media executives for its outlook: on average, they expect 43 percent less traffic from search engines over the next three years. This is an expectation of the respondents, not an already measured industry trend. AI answers, platform usage and changing search behaviour increase the pressure to build their own points of contact.
Publishers are already reorganising their reach work: Srocket reported on new audience structures at USA Today and the New York Times. Others buy back lost reach through paid search. Registration is a third strategy – but only for people who can still reach the website at all.
Three technical and economic limits
First: Every barrier can deter readers. Even a free login costs time and trust. A test has to account for the overall yield, not just the supposedly higher value of those who sign up.
Second: An email address does not automatically mean permission for personalised advertising. For German publishers, consents, purposes and data processing must be reviewed separately. Direct contact is no licence for unlimited tracking.
Third: A registration wall needs a clean search engine implementation. Google describes, for protected content, among other things, marking it up with `isAccessibleForFree` and `hasPart`. This does not automatically make protected content easier to find or freely accessible.
For German media and brand publishers
The question is not "login yes or no", but at what point additional data justifies the friction. For a newsletter publisher, an optional form may be the better starting point. Our analysis of personalisable newsletters and their actual conversion shows why open and click rates alone are not enough.
AI visibility will continue to be measured differently from a page visit. The new SPUR standard for AI content telemetry addresses exactly this gap. Registration data does not solve it, but it can make the value of the remaining direct relationship easier to trace.