If there’s one lesson I hear from founders over and over and over again, it’s on the importance of focus. Jessie Middleton, the founder running three startups at once, eventually had to scale back one and kill another in order to focus on the most promising, Backstory. Music startup Songza didn’t gain much traction until its founders peeled away functions to reveal the simplest, best version of their product.

Four-time entrepreneur Yaron Galai learned that lesson too, but with one small difference: he had millions of dollars worth of exits in the process. He sold Quigo to AOL in 2007 for $363 million; he sold Ad4ever to Atlas and aQuantive in 2004; he sold NetWorks Web Design before that.

Finally, with his fourth startup, Outbrain, he’s building the company he envisioned all along. But he only realized that fact recently, when one of Outbrain’s early seed investors found an email from 1995 that contained a business plan he had written. Looking at it today, Galai noticed that with some small changes, the business he envisioned almost 20 years ago was basically the same as Outbrain. Which is to say, he had wanted to build an engine to recommend digital content–news stories, mostly–to readers while helping media outlets spread their content.

The businesses he ended up building didn’t ultimately go in that direction because he got distracted, he says. Investors suggested he charge more to publishers, or charge more to brands, or focus on advertisements. And that’s how he ended up with his biggest exit, contextual ad targeting network Quigo, which sold to AOL. (He’s in the strange position of having sold a company to AOL and later divested one from it: Last year Outbrain bought the technology of a company called Surphace from AOL. Galai said Tim Armstrong went into the meeting thinking Galai was pitching him to buy Outbrain and walked out with a deal to spin off a piece of his own company.)

The lesson learned is on the importance of saying no.* Outbrain’s goal is a pure and simple one: provide a service that helps the reader and supports the content producers. He took a hard line against any suggestions to charge more money, or offer more ad-like products. The stance was clear: He wanted to actually eat his own cooking. Each of the previous companies he’d founded ended up doing banner ads, something he and most web users do everything in their power to avoid. Most advertisers admit their products stink. Why else would .01% success rate be the industry average?

“Everyone seems obsessed with squeezing the poor ad lemon for more and more revenues,” he says. “That’s why we get publisher pages that are so full with what I call highly optimized crap.”

So Outbrain launched to provide an actually useful service. The New York-based company powers those content recommendation engines you see at the bottom of web pages, mostly on blogs and news sites. Sometimes the suggested links are for outside sites, which are paid placements by the media companies and marked as such. In less common cases they’re purchased and promoted by brands. Several hundred media outlets including Atlantic, CNN, Fox, MSNBC, Us Weekly, and USA Today use Outbrain as well as 100,000 blogs. You’ wouldn’t know it, though, because its white labeled. The company serves around six billion pageviews a month and has experienced triple digit revenue growth since it started making money in 2010. It’s raised $65 million in venture capital and staffed up to 100 people.

Even with a staff that big, Galai still has the last word on anything that could damage trust with the readers. He throws out around half of the potential revenue coming in from brands because he doesn’t want to push spammy content, he says. He’s obsessed with trust–readers need to trust that the links his engines are suggesting are to legit content, even if it’s being promoted by brands instead of media outlets. If they lose that trust, his clickthrough rates will be no better than an ad unit. And if he learned anything from his first three companies, it’s to know when to say no.

*One group he doesn’t say no to is the Israeli Navy: As a naval officer, he still returns to service for one month each year, even when he’s negotiating deal terms.