Optimal Sampling of News Organizations’ Paid Content

Since the New York Times introduced its new paywall in 2011 (its first attempt, TimesSelect, ended in 2007), a number of other news sites have followed suit. One of two approaches is generally chosen: Making all content subscription-based or allowing a limited amount of free content before a subscription is required. The second approach, known as a “metered paywall,” is the strategy the Times implemented.

A 2011 paper from Columbia Business School and the University of Zurich, “Optimal Sampling of Paid Content,” explores two questions about the metered-paywall model: How many free articles a publication should offer; and, what the cost of content behind the paywall should be. The researchers did this by constructing a thought experiment that looked at consumer expectations of quality, demand, subscription price and projected ad revenues.

Key findings conclude:

• Free samples have both positive and negative effects on revenue. They can improve a customer’s opinion of the quality of a publication, which may increase demand, but they also cannibalize existing revenue.
• If customers believe a publication is of high quality, free samples won’t make them more likely to buy the product. Instead, all they do is cannibalize existing revenues. Consequently, if online advertising rates are high, firms should make all content free. If rates are low, the company should put all of content behind a paywall.
• When customers believe a publication is of low quality, free samples may improve revenues. Consequently, if online advertising rates are high, firms should offer a high number of free samples at a low subscription price. If advertising rates are low, the company should offer a low number of free samples at a low subscription price.

While the paper is a logical exercise rather than a data-driven study, the researchers believe that their model has been confirmed by anecdotal evidence. “Our predictions are consistent with casual observations from the media industry,” they state. “Once online advertising rates are high enough, our model predicts that the publisher should offer its entire content for free. This might explain why the New York Times abandoned its online subscription program, TimesSelect, in 2007 when online advertising revenues were soaring. Likewise, our model predicts that the publisher should introduce digital subscriptions and offer a limited number of free samples when online advertising revenues decrease. This might be an explanation of why the New York Times re-introduced digital subscriptions in 2011.”

Tags: consumer affairs, news, technology

By Alex Remington March 13, 2012