In fall of 2010, New York-based advertising startup Appssavvy was soaring. It was three years old, with 75 employees and newly reached profitability. Its roster of clients was impressive, and it would routinely sell million-dollar one-day ad integrations to big companies, including Frito-Lay and McDonald's.

This was around the time when the Zynga game Farmville was played on a daily basis by more people than live in France. And other Zynga games were wildly popular as well; in fact Zynga had become by far and in large the most popular platform for the innovative, in-app advertising Appssavvy sold. 

But it was this imbalance that might have foretold the drastic change Appssavvy underwent over the next year, by way of a dramatic, drawn-out pivot at the height of the company's success.

Appssavvy co-founder Chris Cunningham said he started hearing murmurs of skepticism from the company's board members, who warned about devoting the resources of all of the company's salespeople to selling ads in just one platform. 

"When you have 20 people out there in the market representing those games and the brand, you become synonymous with them," Cunningham says. "You become Zynga."

Of course, this is more than a year before Zynga started teetering as it broke away from Facebook, culminating in a dramatic second-quarter-of-2012 collapse. But Cunningham had his doubts that working with the former social-gaming giant would allow his company to grow. That's because while deals were sizeable, repeat business for large brands was hard to come by.

"The client would say, 'wow, that was incredible!,' and we'd win an award. And we'd be like, 'when are we going to do this again?'" Cunningham explains. "And the agency would be like, 'we're not! But you killed it. Thank you.' We'd exhausted their entire 'innovation' budget for the year."

In the advertising world, you're only as good as your next hit. And this left Appssavvy in the morale-crushing place of constantly having to start over. "You're a hamster on a wheel," Cunningham says.

It also meant his company wasn't scalable. So, he did something that shocked his staff: He decided the company would need to build its own advertising platform. That would require cutting off the relationship with Zynga. He and Michael Burke, his co-founder, gathered the staff in early 2011 and explained the symbiotic relationship with Zynga was over. His employees weren't exactly delighted.

"I remember talking with Michael and saying, 'I think the meeting went well! That felt good to communicate,'" Cunningham says. "But the sales staff probably thought I was out of my mind--the biggest psycho ever. Pretty much the entire year it was a slow bleed of staff. There were too many question marks."

Madison Avenue watched on as Appssavvy had a public divorce with Zynga, shortly followed by a public divorce with much of its sales staff. Cunningham says it felt like he'd brought the company to the top of a mountain, and now it was back down at the bottom starting over.

"We had made a decision and laid down our tracks and we had to stay focused," he says. "And we were, but inside I was actually shaking." 

But Cunningham and Burke weren't flying blind going into the pivot: They had an idea for the future of the company even before cutting it off with Zynga--and they continued ad sales to keep revenue rolling in for months while Appssavvy started working on hiring a team of engineers and developing its own intellectual property--the ad platform. That's not to say it was easy.

"One was I had to remind myself that what we were embarking upon was more than what most founders would dream of: Developing technology out of New York City to try to change advertising," he says. "Day to day what we were doing was super challenging."

Two years later, Appssavvy the company functions very differently. Its staff is mostly engineers, and they've built a technology for serving up big, bright, full-screen ads on smartphones. Mostly, the ads can appear within games, during what the company has dubbed a "natural break," say, when a player wins a round. The ad might show context for how it relates to the game by saying something like, "congrats" right on the full-screen ad.

Through the pivot, though, some things haven't changed. Appssavvy's principles--down to its tag-line--are the same as they ever were. "Rethinking the delivery and reception of advertising focused on real-time behavior" is still the mission. And, through all that, the company has retained many of its same clients. 

Appssavvy is back up to employing about 20 people, and also re-reached profitability this month, according to Cunningham.

"The good news is we are heading in the right direction now," Cunningham says. "We just had to go to hell and back to prove it."

Cunningham says he learned a tremendous amount--through trial and error--over the past three years. Here are four things he thinks any founder should know before changing a company's direction.

Your old game plan won't work.

Cunningham says the day after his big meeting announcing the company would focus on building a new platform, "We were a new startup on that day, but we were behaving like, and convincing ourselves, we were the same company." He says he should have reassessed the company's revenue projections for the next year, rather than assuming the same roadmap would work. "We had been so successful in the past, we were confident things should translate." Instead, he says, founders need to consider that when they pivot, they are back to Day One. He says: "Adjust and learn the same way you did the first day you started your company." 

Your staff won't stay. 

Figure out who you really need as soon as possible, Cunningham says. After many of the company's sales staff left in months subsequent to the pivot meeting, two rounds of layoffs followed. But Cunningham says Appssavvy needed engineers to build its next iteration: "The kinds of people required for that, it took us too long to identify," he says. 

It will cost more than you think.

Remember, Appssavvy was at its zenith when it decided to pivot. And it got through that transition with just $13 million in venture capital investment. But more would have been helpful, Cunningham notes in retrospect: "We should have raised an enormous amount of capital to support the transition." He suggests other companies making a pivot seek funding before making the leap, and tied specifically to that leap.

You need to communicate clearly--even if you're unsure.

"I should have been more clear about the fact that I didn't know what [the change] meant for us, and that I couldn't guarantee what it meant for anyone individually," Cunningham says. Unclear? Sure. At least it's transparent.