Plenty of businesses strive to be innovative by spending millions of dollars launching new products. While new research from Global Innovation 1000 showed that big companies are planning to invest more money in research and development, Adobe figured out a cheaper way: Don't focus on innovations, instead spend money turning your engineers into an army of innovators.

"When we went from putting software on round shiny things in boxes to delivering it all on the cloud, we changed our approach to innovation," Adobe's chief strategist and VP of creativity Mark Randall said on Thursday at the Lean Startup Conference in San Francisco.

According to Randall, Adobe used to invest in about a dozen projects a year and spend nearly $1 million developing and researching each idea. Today, the company funds several hundred new projects for less.

"We wanted to build innnovators, not innovations," he said. "That meant giving our engineers experience in both succeeding and failing as innovators. We weren't failing enough, which meant we weren't being bold enough and [we were] missing opportunities."

That's how Randall and his team came up with the Kickbox, a kit for engineers to pursue a passion project and find real customers for it. Adobe trusted its engineers and all ideas were pre-approved and given a budget of $1,000.

"They were the CEO of their idea," Randall said. "There was no expense reporting required."

The Kickbox includes tasklists to guide the engineers from idea inception (finding something that aligns personal motivation with Adobe's enterprise purpose) to infiltration (pitching the idea and subsequent market research to executives).

One of the projects was a online marketplace that connected Photoshop designers with customers looking for creative services. The project launched as an standalone service without any Adobe branding and through trial and error discovered that customers were willing to pay $3 for quick-turn Photoshop fixes.

Not all Kickbox-led ideas were approved for further funding, but the data they collected from real customers was still valuable research for Adobe.

"The best pitches have come from engineers who have discovered that the data did not support their idea and that we shouldn't fund it," Randall said. "All the executives in the room stood up and applauded because instead of spending millions of dollars developing a bad idea, we spent $1,000 and got important customer feedback."

Adobe's Kickbox incorporates aspects of crowdsourcing to get engineers comfortable with failing fast and weeding out tons of bad ideas to get to the right one. So far the kit has been given to 1,000 engineers at the company, and other startups can follow in Adobe's steps starting in February when the Kickbox is expected to go open source.