Even on the web, sometimes actions really do speak louder than words.

The technology site Ars Technica has a tech-savvy group of readers, of which about 40 percent have installed ad-blocking software in their web browsers. That’s a plugin that allows you to avoid seeing most ads on a site. The financial consequence for Ars is “devastating”, editor-in-chief Ken Fisher explained in a post. Ars sells ads based on impressions, not clickthroughs — which means it takes a big financial hit because of browsing habits of its users.

On Friday evening, Ars tried an experiment: Readers running ad blockers got a blank page instead of the story they intended to read. The move was a technical success, but caused an uproar (and confusion) among users. In hindsight, Fisher told me, the site’s experiment in retribution was the “wrong approach,” causing confusion among many readers.

“What we weren’t expecting is so many people were blocking ads and didn’t even know it,” he said. “It left a lot of people very confused. They started digging around, wasting an hour trying to fix their broken computer.” There was nothing on the site to explain to readers why content had been blocked.

But the experiment still generated positive returns for the site’s bottom line. Fisher wrote a lengthy post on Ars (similar to many the site has run before) about its goals and why ad blocking was a big problem for the site:

My argument is simple: blocking ads can be devastating to the sites you love. I am not making an argument that blocking ads is a form of stealing, or is immoral, or unethical, or makes someone the son of the devil. It can result in people losing their jobs, it can result in less content on any given site, and it definitely can affect the quality of content. It can also put sites into a real advertising death spin.

And since Saturday, Fisher has received about 1,200 emails from users saying they had whitelisted the site — meaning they had told their ad-blocking software it was okay to show Ars’ ads. Based on Ars data from IP addresses, 25,000 users whitelisted the site in a 24-hour period — evidence that the goodwill the site has built up with its audience could be converted into user acts of generosity.

Another 200 users signed up for Ars’ premium accounts, which run $50 a year or $30 for six months. A subscription gets users access to an ad-free version of the site, full-text RSS feeds, printable PDFs of posts, and closed community sections of the site. (But Fisher notes that many subscribers just feel a sense of obligation, not a desire for premium features. “We get many people who subscribe just because they love us. They just want us to survive.”)

I asked if the $50-per-year subscription makes up, financially, for the loss of ad revenue on the ad-free version of the site. It depends on the user, Fisher said. For anyone who visits the site more than its user-average 89 visits per month, probably not. But he doesn’t think of the equation in those terms. Fisher views the subscription fees as covering the cost of specialized content that only the most dedicated user would want, like the online community sections. Ads alone wouldn’t generate the revenue to cover that. An advertising strategy that assumes a broad audience can cover the more general-interest content that audience wants. Having a multi-pronged revenue approach allows the site to provide different kinds of content for different audiences.

Fisher said he’s also had good experiences using a sponsorship model to support specialized content, including in-depth coverage that attracts a highly engaged, technical audience, but not huge pageviews. For instance, IBM sponsored a recent series on the future of collaboration. The writers didn’t know IBM was the backer, and IBM was told only the broad topic for the stories. Topic-specific sponsorship “delivers more value than display advertising, in my opinion,” he said. “It’s much more targeted. It takes the best of contextual advertising.”

But Ars’ bottom line still relies heavily on traditional display advertising. Its particular audience likely has a worse ad-block problem than other sites. But the benefits Fisher found from communicating directly with readers — making the ask along with a gentle but clear nudge — can apply to any site.

“It affects so many sites,” he told me. “And just getting the message out there makes a difference.”