Any ad revenue declines that
result from fewer page views are likely to pale in comparison to the revenue
gains from new subscribers. Their bet is that creating a controlled targeted
sampling experience will improve subscription acquisition conversion rates. My
experience from conducting similar experiments at the Financial Times is that
it will.
http://www.niemanlab.org/2017/12/the-new-york-times-has-halved-its-free-monthly-articles-to-5-its-most-significant-paywall-change-since-2012/