4 a.m. is the most productive hour of the day, suggests a Wall Street Journal story from this past August. It’s a light piece brimming with provocation (“4 AM???? Areya kidding me? I didn’t even know there was a 4 AM” writes one commenter), and one that travels well on social media.

“That’s a story, historically, where most sites with paywalls might go, ‘Oh, this is lighter content. We’re not going to charge for it because no one will pay for it,'” Katie Vanneck-Smith, Dow Jones chief customer officer, told me. But the Journal shook up its paywall structure over the summer so that its stories were no longer impenetrable, but accessible through links shared on social by Journal staffers and already paying subscribers. It began offering 24-hour guest passes to give non-subscribing readers more of a choice in sampling Journal stories, without having to pay for a full subscription right off the bat. Under its bendier paywall structure, the 4 a.m. productivity story was a “best-selling story,” according to Vanneck-Smith, referring to readers who clicked on the story and purchased a Journal subscription directly from an offer that popped up there.

More than 50 percent of the Journal’s subscription sales now come from individual articles like the 4 a.m. story, a “considerably higher percentage” than before the Journal changed up its paywall strategy, Vanneck-Smith said, adding that “you’d expect 20 or 30 percent of your sales coming in that way, in a paywall model.” Readers who entered their email for a guest pass were six times more likely to convert to a regular subscription. (Vanneck-Smith wouldn’t share exactly how many readers had shared their emails for pass access.)

“When you think about a subscription membership website, often you’re selling subscriptions from the shop or the sales page where you’ve got your compare and contrast of different options,” she said. “Now our strategy is that any story that’s been shared by a member or journalist — either by email or social channels — is your invitation to come in. And 50 percent of all sales are now directly a result of something that’s been shared, which is an interesting change for us.”

In August, the Journal hit 948,000 digital-only subscribers. Now, it’s surpassed a million (though actual figures for the quarter ending in December won’t be released until sometime in late January). The increase isn’t just thanks to a post-election wave of interest in and support for quality journalism, according to Vanneck-Smith: “Like everybody else, we’re seeing that growth accelerate, but we were seeing a great uplift in sales pre-election, too.” (I kept referring to “paying readers,” and Vanneck-Smith emphasized that they are “members.”)

The Journal is continuing granular analysis of exactly what makes more casual readers want to subscribe, with around a few dozen A/B-testing type experiments going on at any one time.

“Our previous one-size-fits-all experience is now quite different, depending on what content you’re coming in from, depending on what we know about you as a reader from previous behavior, whether you’re coming in from social,” Vanneck-Smith said. “You might get a different experience if you’ve read five or more stories, versus someone who hasn’t. There’s a different experience if you’re converting with a video, versus if you’re coming from search. There’s a lot of optimization happening.”

“We’ve got an ever-growing rule book of smart little things we know, such as how many stories do people read before they subscribe — I’m not going to give all of that away! — building a fuller picture,” she added.

Throughout the fall, a cloud of bad news has surrounded the Journal, amid its WSJ2020 strategy for newsroom modernization for the next three years. Advertising revenue fell 21 percent in the quarter ending September 30, according to parent company News Corp., which posted a $15 million loss that quarter. It cut entire print sections, pushed buyouts, and laid off staffers.

“I won’t pretend that this didn’t become more pressing with the decline in print advertising. We’ve seen that very sharp decline this year,” Editor-in-chief Gerard Baker told Nieman Lab in an interview last month. But, he emphasized, “we are focused really hard on circulation growth. That’s not to say that we don’t want advertising — we want advertising. Again, for the long-term future of the paper, we’ve staked our strategy for the paper — by the paper I mean the paper in the broadest sense of the term, The Wall Street Journal, including digital — on growth in digital circulation.”

I asked Vanneck-Smith whether she remained as optimistic about the Journal’s future as she was when we spoke back in August.

“If I was not at Dow Jones, I would be more worried,” she admitted. “We’ve got a more diversified set of revenues — not only do we have circulation from the Journal and Barron’s, but we also have a subscription B2B business, selling data and information to corporations. Having that triple play of revenue streams means we’re more cushioned to weather some of the challenges of the industry. I think what you’re seeing is a transition, after 12 months of which we come out stronger and as, if not more, profitable, and increasingly digital, with a strong, diverse portfolio of revenue streams.” (The Journal does not break down what percentage of revenue comes from B2B sales.)

“On an industry level, I’m pretty confident that a very clear proposition — be that in editorial positioning, be that in quality journalism, be that to a specific niche — if you have a very clear proposition, and you have a very strong relationship with your customer base, those types of publications will survive,” she said.