If you want real innovation, forget about continuous improvement, says Verizon CIO Shaygan Kheradpir.\n\nAt Verizon, Kheradpir primes the innovation pump by looking for places where IT can have a breakthrough impact on the business. He then challenges his IT group to come up with new ideas in those areas, which in turn set off rounds of discussion and brainstorming. When a potentially big idea emerges, Kheradpir takes it to a line-of-business head or to his boss, Lawrence Babbio, vice chairman and president of Verizon's Telecom Group. If they agree that the idea is worth checking out, Kheradpir takes a venture capital approach by giving a team 30 days to come up with a prototype, putting aside any constraints that might get in the way, such as scalability, usability and functionality issues.\n\n"Show me what this could look like," Kheradpir tells the team. "Bring it up on one leg, hobbling."\n\nIf that version shows promise, the team gets another 30 days to expand the prototype, with a few constraints added to the mix.\n\nProjects move forward in 30-day cycles. If they survive, the cycles expand to 60, then 90 days as more constraints are added and the projects move into operational readiness tests.\n\n"Even though we're looking for breakthrough results, we make sure [teams] execute incrementally so we can course-correct," says Kheradpir.\n\nAlthough VC firms ditch ideas that don't pan out after two or three 30-day cycles, at Verizon, they're not automatically killed. "We look at it as a learning experience," says Kheradpir. "We backtrack and try again because we still need breakthrough results." Generally a dozen new ideas are in the innovation pipeline at any time; Kheradpir estimates that more than half make it all the way through to execution.