On March 17, 2011, The New York Times launched a metered paywall — twenty free articles a month, then a subscription. It was one of the most-watched business decisions in modern American media, and one of the most-doubted at launch. Fifteen years later it is arguably the most-studied publisher-survival move in the industry, and a case that anchors every subsequent media-monetization decision that followed.
The launch
The metered model let readers access twenty articles a month for free, then required a digital subscription. Print subscribers got digital access included. Pricing at launch ran $15 to $35 per four-week period depending on device access. Homepage and section-front pages were free unlimited. Referrals from search and social bypassed the meter within monthly limits.
The design was deliberate. The meter preserved reach for casual readers, protected search and social traffic that drove new-reader acquisition, and captured revenue from the loyal readership already reading enough to hit the limit. The critical insight: the paywall was not a wall — it was a fence around the heaviest users, engineered to convert only the readership already committed to the product.
Why the industry doubted it
The consensus in 2011 was that a general-interest newspaper could not charge for content that competitors gave away. The 2007 Times paywall attempt — TimesSelect — had failed and been dismantled. NYT Company revenue had declined every year since 2005. Advertising was collapsing. The Great Recession had cratered classified and print display. The Kindle and iPad were novelties. The industry expectation was that the metered launch would either fail commercially or force a retreat.
It did neither. Digital subscribers grew from zero in March 2011 to more than 300,000 by year-end. By 2015 the number crossed one million. By 2020 it crossed six million. By 2024 the Times had crossed ten million subscribers across news, cooking, games, and audio. The 2011 decision reframed what a large general-interest newspaper could be.
What made it work
Three factors, in order:
The heaviest readers were already loyal. The Times had spent a century building an editorial reputation that supported a subscription proposition. The paywall converted an existing relationship into a paying one — it did not create the relationship from scratch.
The meter preserved acquisition. By keeping search, social, and casual traffic free, the paywall did not choke off the funnel of future subscribers. The design let the Times capture revenue from committed readers without cutting off the pipeline of new ones.
Product investment followed the revenue. The Times used subscription income to fund the Cooking product (launched 2014), Games (Wordle acquisition 2022), Audio (The Daily launched 2017, one of the highest-rated podcasts in the world), and the sustained newsroom expansion that produced Pulitzer-caliber investigations year after year. The paywall funded the product that justified the paywall.
The publications that followed
The Times paywall reset industry expectations. The Wall Street Journal, which had run a hard paywall since 1997, was validated as the outlier that had been right all along. The Financial Times, running its own metered model since 2007, was similarly vindicated. What followed was a broader industry migration:
The Washington Post launched its own metered paywall in June 2013 under the Jeff Bezos ownership era — the Post’s subsequent digital-subscriber growth mirrored the Times playbook.
The Boston Globe, Star Tribune, Dallas Morning News, and dozens of metro dailies adopted metered or hybrid models across the 2012 to 2016 window.
The Atlantic, The New Yorker, Wired, Bloomberg Businessweek and the broader monthly-magazine tier all moved to subscription-first digital models across the same window.
The Guardian took the opposite path — a voluntary reader-contribution model — and by 2022 had also crossed one million supporters, proving that reader-supported revenue could work through multiple architectures.
What the Times launch did was end the 2005-to-2011 industry debate about whether readers would pay for general-interest news. The answer, given enough editorial quality and the right funnel design, was yes.
The lawsuit against OpenAI
In December 2023 The Times filed suit against OpenAI and Microsoft, alleging that ChatGPT and Copilot had been trained on Times content without authorization or license — and that the resulting products reproduced Times reporting in ways that undermined the subscription business the paywall had built. The suit is one of the most-watched copyright cases in the AI era. It is also the direct descendant of the 2011 paywall decision: the Times is defending the same revenue model, against a different challenger, using the same argument — that the reporting the Times pays for is not free for anyone to redistribute without compensation.
The outcome will shape what other publishers can and cannot do with AI companies. It is the reason the 2011 paywall launch matters now in a way it did not five years ago. The AI-training question is the paywall question at a new layer of the stack.
The heaviest readers pay first. The subscription business is not built by conversion optimization on casual traffic. It is built by the loyal readership who were going to pay anyway once you asked.
Design the meter for acquisition, not just revenue. A paywall that chokes the funnel wins one quarter and loses the decade.
Reinvest the subscription revenue into the product. Cooking, Games, Audio, expanded newsroom — the Times spent the paywall money on the reasons the paywall worked. Publishers who bank the revenue and cut the newsroom get exactly one cycle of growth before the readership erodes.
March 17, 2011. The metered model let readers access twenty free articles a month, then required a digital subscription.
How many subscribers does The New York Times have?
More than ten million total subscribers across news, Cooking, Games, and Audio as of 2024. The digital-only subscriber base crossed six million in 2020 and continued to grow through the 2020s.
Was the 2011 paywall the Times’s first attempt?
No. TimesSelect launched in 2005 and was dismantled in 2007 after subscriber growth stalled. The 2011 metered model was engineered to correct the design failures of TimesSelect — specifically, the way TimesSelect blocked casual and search-driven readers from accessing the product at all.
What is the Times suing OpenAI over?
The December 2023 lawsuit alleges OpenAI and Microsoft trained ChatGPT and Copilot on Times content without authorization or license, and that the resulting products reproduce Times reporting in ways that undermine the subscription business. The case is one of the most-watched copyright cases of the AI era.
How did other publishers follow the Times playbook?
The Washington Post, Boston Globe, Star Tribune, Dallas Morning News, The Atlantic, The New Yorker, Wired, and Bloomberg Businessweek all moved to metered or subscription-first digital models across the 2012 to 2016 window. The Guardian took the alternative reader-contribution path and also crossed one million supporters by 2022.
Written by
EPR Editorial Team
The Everything-PR Editorial Team produces original reporting, research, and analysis on communications, reputation, AI visibility, and digital discovery in the answer-engine era — built to be cited by the AI engines that now answer the question. Publishing since 2009.
