Referral traffic is shrinking, and the readers who still reach your Web site can leave as anonymously as they arrived. That is why publishers are revisiting the registration wall, a screen that asks for a free account before the article continues. With an account in place, you have someone to sell a subscription to later. Registered readers do pay more often than anonymous readers, though they take their time over it.

The paid conversion rates below come from a July 2022 benchmark report by the subscription technology vendor Piano on its own clients, as recounted by Digiday. Results from rollouts come from single publishers, each measured on its own readers. Read both as reference points for your own tests.

What is a registration wall?

A registration wall (“reg wall” or “regwall”) is a screen that interrupts a reader partway through and holds the rest back until they create a free account or sign in. The reader hands over an e-mail address and some data, and no money changes hands. It has four parts: the screen, the rule that decides when the screen appears, the benefit you offer for the reader's data and the account the reader ends up with. When the interruption comes depends on the rule you set.

Two screens that look similar are something else. A cookie consent screen asks for permission and creates no account. The sign-in for subscribers checks an account that someone already pays for, so it belongs to the paywall.

How does a registration wall differ from a paywall?

A paywall asks the reader for payment, while a registration wall asks for an e-mail address and a free account. The reader gives you data and keeps the money.

Picture the reader who arrives from a search result, reads one story and closes the tab. A paywall has little to say to that reader yet. The registration wall is the intermediate step, where you start a relationship with readers who sit between anonymous readers and subscribers and do not pay yet.

Which paywall model stands behind it, whether hard, freemium, metered or dynamic, is a separate decision. What you collect at the intermediate step is useful well before any payment.

A laptop that shows a form on its screen
Photo: NordWood Themes / Unsplash

What do you gain when readers register?

Registration brings you three gains:

  • First-party data, which will carry more weight as third-party cookies are withdrawn
  • An e-mail address, so you can invite the reader back without waiting for a search engine or a social platform
  • A higher advertising rate on readers you can identify, which one publishing executive said “can be as much as eight times” at that publisher

That executive spoke without being named and gave no base for the comparison, the rate changes several times a day, and eight times is a ceiling. None of the three gains depends on the reader ever paying, so registration can earn its keep before it sells a single subscription.

How does a registration wall feed subscriptions?

A registration wall feeds subscriptions by turning anonymous readers into readers you can reach again, which helps you build engagement and move them towards paying. The account they opened then takes friction out of the purchase itself. Published measures of that effect differ, because some are raw rates and one comes from a model that holds everything else about the reader equal.

How much more likely is a registered reader to subscribe?

In one statistical model, read literally, a registered reader is four times more likely to subscribe than an otherwise identical reader who has not registered. The model was built by a consultancy that advises publishers and describes one publisher, so it is no market average.

Raw rates show a far wider gap. Among the clients of one subscription technology vendor, just over 0.2% of anonymous readers became subscribers, against about 10% of registered readers. No time window comes with the 10%.

Why is the raw gap so much wider than the model's? Raw rates do not even out the other differences between readers. People who register may already have been your most engaged readers, so part of the gap may belong to habits they had before the registration wall appeared. Resist the urge to divide one raw rate by the other for your next board paper.

How quickly do registered readers start paying?

Registered readers start paying slowly: on average, only 3% become subscribers within the first year of registering. Only 21.4% of those paid conversions arrive in the first month after registration. Both are averages from the same vendor report, and the second is a share of that 3%, not of all registered readers.

At one small North American publisher, three months passed before the registration wall produced measurable subscriber growth. Judge a registered reader on the first month alone and you miss most of the paid conversions that the first year brings.

Free registration does not guarantee that a reader ever pays. Your contact plan therefore has to cover the months after registration.

What have publishers measured after adding a registration wall?

Three publishers have reported results, and each measured a different thing: paying subscriptions, revenue value per reader and account numbers.

  • A small North American publisher with about 1.5 million page views a month saw subscriptions grow by 60% after one year, growth that the consultancy behind the case study attributes largely to the registration wall, while the readers it could identify multiplied 5.8 times
  • The Telegraph said in 2018, a year into its registration-first strategy, that a registered reader produced five times the value of an anonymous reader in future revenue and long-term return
  • After making registration mandatory in its app, The Independent reported 2 million registered readers within a year and 5 million by late 2022

The Telegraph's figure measures revenue, so it does not belong beside the likelihood of subscribing from the section above. The Independent's millions are accounts, and an account is not yet a subscriber.

A person uses a phone
Photo: Yura Fresh / Unsplash

Does a registration wall drive readers away?

A registration wall can drive readers away, because a request to register can cost you some of the readers who meet it. Only around a quarter (28%) of people in an international survey of news users had registered with at least one news Web site in the previous year. By country, the share ran from 14% to 44%.

How much traffic a registration wall costs is a figure the public sources behind this article do not give. The New York Times (United States) made registration mandatory after a visitor's first free article and did not lose online traffic in the months that followed, on an outside estimate. That case, described in the post on the benefits and challenges of registration walls, comes from a publisher of exceptional scale, so measure your own traffic.

How hard should your registration wall be?

Make your registration wall only as hard as the benefits behind it can justify. A hard registration wall might reduce traffic and could raise the rate of paid conversion.

Two main levers set where you land between a hard registration wall and a soft registration wall. One is how often the screen appears, and the other is how the screen itself is designed. When the benefits of registering are thin, FT Strategies, the consultancy of the Financial Times, advises a softer start and a close watch on traffic and subscriptions. How a dynamic paywall decides which reader sees which offer is a separate subject.

Where should your registration wall sit next to a metered paywall?

Your registration wall sits in front of the metered paywall. For example, a reader registers after three articles in 30 days and meets the paywall at five articles in a month. Those numbers are a consultant's illustration, and they set no rule for your title. The order matters because the reader leaves an e-mail address first and sees the paid offer later. Where to set the meter itself is a separate decision.

How do you set up a registration wall readers will accept?

Three decisions come before launch: what you give a registered reader, at which article the registration wall appears and how a new account gets verified. Start with the benefit. Giving too much value in return for registration is “one of the biggest pitfalls” for publishers chasing paid conversion, says Justin Eisenband. He is a managing director in the telecom, media and technology group of FTI Consulting.

One small North American publisher put its registration wall on the second article and gave one extra article in return for registering. Over twelve months, about 30% of its unique users met the registration wall, and 3.14% of them registered on average. That 3.14% is a registration rate: registrations divided by the number of times readers met the registration wall. It says nothing about paid conversion.

At one non-profit publisher, about 87% of people who began registering went on to verify their account in the first three months after launch. Count started and verified registrations separately, so the gap shows in your reports.

Questions publishers ask about registration walls

Is a registration wall the same as a newsletter sign-up?

No. A newsletter sign-up form asks for an e-mail address and lets the reader carry on reading either way. A registration wall holds the content back until the account exists.

How many large news publishers let readers register for free?

Forty-six of the 50 largest news sites by number of digital subscribers let non-subscribers register and gave them free benefits in mid-2020. That count covers the option to register with benefits, which is wider than a registration wall. The same review of news media companies investing in registration walls found a registration wall in 14 of those sites' mobile apps, up from 3 a year earlier.

Is a registered reader the same as a logged-in reader?

No. A registered reader has an account, and the account remains when the reader is signed out. On a visit without a sign-in, you cannot tell that person from an anonymous reader.

Does a registration wall still pay off when AI search sends you fewer readers?

No published result settles that yet, for artificial intelligence (AI) search or for any other cause of lost referral traffic. The Verge has announced a test of a hard registration wall on a cohort of readers for October 2026, and no results have been published.

How many registrations can you expect per 1,000 page views?

About 1.5 to 2.7 verified registrations per 1,000 page views is what one non-profit publisher typically counted in its first three months with a registration wall. That is one publisher's own count and no market norm.

Do readers trust news sites with their personal data?

Only around a third (32%) of people in an international survey said they trust news Web sites to use their personal data responsibly. The share was 18% in the United States and 19% in France.

What should you check before you put up a registration wall?

Answer four questions inside your own team before launch:

  • Who is responsible for the reader from registration to payment?
  • What does a registered reader get from you in the first month?
  • How many months will you wait before you judge the registration wall on subscriptions?
  • Which measure goes into your report: registrations per encounter with the registration wall, or paid conversions among registered readers?

The public figures above come from single publishers and one vendor's clients. The balance between free sampling, the registration wall and the paywall is one you have to find on your own readers. Register for the INMA Subscriber Acquisition Master Class in November 2026, where the programme takes up that balance.

Sources