Industry
Contents
Business and AI, with a European lens
Growth Signal

The Guardian hid the skip button and sign-ins jumped 80%

The Guardian hid its skip button and pushed more readers into sign-in.

Prefer this source
Guardian app onboarding experiment showing a sign-in flow with the skip option hidden
The Guardian's onboarding test raised sign-ins by changing the visibility of the skip option.
Experiment summary

Experiment at a glance

Test
Agency A/B test
Reported result
Registrations and sign ins up 80%, agency reported
Takeaway
Hiding the skip option lifted sign ins, but the case cannot say if those accounts stuck.
  • The Guardian app’s first screen offered “Sign in” and an easier “Start reading…” button, and readers took the easy route.
  • The treatment removed that button and several panels, leaving one “Continue” button that led to a sign-in screen explaining account benefits.
  • Conversion, the agency behind the work, reports an 80% increase in registrations and sign-ins.
  • The case study discloses no baseline, sample size, duration or significance, and it combines new registrations with returning sign-ins.
  • Nothing public shows whether those accounts improved advertising performance or retention.

The Guardian made registration harder to skip. Its app test reveals how a small design change shifted reader behaviour.

The first screen in The Guardian app gave readers two obvious choices. They could sign in, or tap the more inviting “Start reading…” button and go straight to the journalism.

For a reader who had just opened a news app, the second option did exactly what it promised. For the publisher, it created a problem. The Guardian wanted more registered and signed-in readers, yet its onboarding made the route around registration easier to understand than the reason to complete it.

The first experiment therefore tested a narrowly defined change. The team did not begin by adding an incentive, shortening a form or rewriting every benefit. It changed which choice the opening screen made prominent.

The easy exit was winning in the Guardian app onboarding experiment

The experiment appears in a case study published by Conversion, the agency that worked with The Guardian’s product and data science teams.

Conversion says The Guardian wanted more known visitors, readers who were registered and signed in. Repeat readers could visit several times a day across different devices. Without a login, the publisher could not reliably connect those sessions or build the same view of the reader.

The commercial argument was that better customer understanding could support advertising revenue.

Before changing the flow, the team observed usability sessions in The Guardian’s research labs. According to Conversion, users did not understand the benefit of creating an account or signing in. The wording also created an imbalance. “Sign in to the Guardian” sounded like a task for someone with an account. “Start reading…” sounded like the thing every new reader had opened the app to do.

The control therefore mixed two problems. It did not explain the value of registration, and it placed a direct route to the content beside the registration route. A low sign-in rate could have reflected either problem.

One screen changed the choice

The treatment removed several blue panels and the “Start reading…” call to action. Readers saw a single “Continue” button, which took them to the existing sign-in screen. There they could use Facebook or Google and see reasons to sign in or create an account.

Readers could still continue without signing in. The option remained available on the next screen, but it was less prominent. The test changed the sequence and visibility of the choices rather than imposing mandatory registration.

The treatment needs to be read in full. The treatment did more than remove a button. It sent more readers to a screen that explained account benefits and offered social sign-in. The reported result therefore belongs to the complete change in the path. The source does not isolate how much came from reduced prominence, the extra explanation, the social options or simple exposure to another screen.

Conversion describes the comparison as an A/B test and reports an 80 percent increase in registrations and sign-ins. It does not disclose the baseline conversion rate, sample size, duration, traffic allocation, platform mix, confidence interval or statistical significance. It also combines new registrations with returning-user sign-ins, which may represent different behaviours.

The agency says this was a proof-of-concept first experiment. It later worked with The Guardian on broader registration, engagement and contribution projects, but those later outcomes should not be attributed to this onboarding test.

The test measured a choice, not customer value

An 80 percent lift sounds decisive until the missing denominator enters the picture. A rise from 1 percent to 1.8 percent and a rise from 20 percent to 36 percent would carry different product and commercial implications. The case study does not tell us which kind of change occurred.

Registration can also move while reader value stays flat. Some people may create accounts and never return. Others may feel pushed through an extra step, reach the content later or leave the app. A publisher assessing this treatment would need to pair the registration metric with completion time, onboarding abandonment, first-session reading, return visits and account use across devices.

The clearest evidence is narrower. When The Guardian made the skip route less prominent and moved the explanation of account benefits into the main path, more people registered or signed in. The public evidence does not show whether those accounts produced better advertising performance or stronger retention.

Test the path before rebuilding the pitch

The experiment is transferable to businesses with an optional setup step that becomes more valuable after a customer has experienced the product. News apps want logins, marketplaces want saved preferences, retailers want accounts and software products want users to connect data. In each case, the easiest route through onboarding can compete with the action the business wants customers to take.

A lean version would begin with the current onboarding flow as the control. The treatment would keep the optional route available but move it behind one screen that clearly explains the value of setup. Product access, account requirements and the form itself should remain unchanged so the test stays close to a question about sequence and prominence.

The primary metric should match the valuable action, such as verified account creation or completed sign-in per eligible new user. The decision threshold should require a pre-agreed improvement with acceptable uncertainty while onboarding completion, time to first value, early retention, support contacts and deletion rates remain within guardrails.

If the treatment wins at registration but harms first-session engagement, the next test should separate visibility from persuasion. One version could change only the prominence of the skip option. Another could preserve the original choices while adding a concise account benefit. The comparison would show whether customers needed a clearer reason or simply followed the path placed in front of them.

The Guardian’s test began with a small piece of interface copy, but the useful question sits deeper in the flow. When customers bypass an optional step, are they rejecting its value, or has the product made bypassing it the clearest decision?

Two other tests in this series show the same instinct at work on a product surface rather than a registration screen. Bunzl removed its most-clicked homepage section after data showed the clicks were not the help they appeared to be, and Sport Chek watched a winning shipping message turn into a loser once the underlying offer changed. In each case, a single conversion metric said less than it first appeared to.

Experiment detailWhat the reporting shows
ChangeMade the skip option harder to spot on the first screen.
Reported result+80% registrations & sign-ins
EvidenceAgency case study, no baseline/sample size disclosed
LimitsSign-ups rose sharply right after the change went live.
Experiment summary

FAQ

What did The Guardian change in its app onboarding?

It removed the “Start reading…” button and several blue panels, leaving a single “Continue” button that led to the existing sign-in screen. Readers could still continue without signing in, but that option was less prominent.

What result was reported?

Conversion reports an 80% increase in registrations and sign-ins from an A/B test. It does not disclose the baseline rate, sample size, duration, traffic allocation or statistical significance.

Does the test show the new accounts were valuable?

No, the experiment measured registrations and sign-ins only. It cannot show whether the accounts improved advertising revenue, loyalty or retention.

How could another business test the same idea?

Keep the current flow as control, move the optional route behind one screen that explains the value of setup, and track onboarding completion, time to first value and early retention as guardrails.


Source note The experiment mechanics and reported 80 percent increase come from Conversion, the agency that ran the work and published the case study. The source was checked on September 21, 2026. It does not disclose the test date, baseline, sample size, duration, allocation or statistical uncertainty. No independent replication or downstream revenue result is provided.

Spotted an error? Contact the desk. Our corrections policy explains how we handle corrections and updates.

Edited by Yajush Gupta

Industry Contents reporting draws on direct interviews, company statements and public records. Figures appear as published by the sources named. Nothing here constitutes legal or investment advice.

Written by

Growth Signal is Industry Contents’ editorial series examining published growth experiments, commercial outcomes and the limits of the available evidence. A Growth Signal Desk byline identifies a publication workflow, not an individual reporter. Source-reported results are distinguished from independently verified findings. Editorial responsibility rests with founding editor Yajush Gupta.

Keep reading

Get TalkHead

Every Friday, TalkHead checks what your preferred AI tools cost, how businesses test their claims and what buyers should check before committing.

Unsubscribe anytime. Your email stays confidential.