Big company models for innovation may work for a while and then stop working as the tailwinds that propelled their growth stop blowing or are replaced by new headwinds. This comes to mind in considering a model for innovation that was very successful in the 1990s for network equipment maker, Cisco Systems.

Between 1990 and 2000, Cisco typically acquired about 60 startups a year. As I wrote in The Technology Leaders, the idea worked because Cisco had built the industry's leading enterprise sales force. Cisco effectively outsourced innovation to startups and kept a keen eye on which ones were winning a bigger share of enterprise network equipment budgets.

As then Chairman John Morgridge told me in 1996, earlier in his career he had worked as a sales person for Honeywell and he saw that sales people were loyal to their customers. So if a new company came out with a product that their customers liked better, the sales people would leave Honeywell and sell the rival's product.

Morgridge decided that to keep the best sales people, Cisco would need to acquire the upstarts whose products customers were buying. Cisco's acquisition strategy propelled its growth -- in the 1990s, its revenues rose at a 75 percent average annual rate and its stock soared at a 58 percent rate -- becoming the world's most valuable company in September 2000 with a $500 billion stock market capitalization. 

This strategy has not worked well for Cisco in the last two decades. Between 2000 and 2010, Cisco's revenues grew at a 7.8 percent compound annual rate while its stock dropped at 6.3 percent a year. Cisco's revenue growth slowed in the 2010s -- its revenue and stock price grew at 2.1 percent and 11.3 percent a year, respectively.

More recently, Cisco has acquired startups and driven out their most talented entrepreneurs. For example, the CEO of Zoom Video, Eric Yuan, left Cisco after he felt that it had botched its acquisition of WebEx of which he was the chief technology officer.

Outsourcing innovation by acquiring startups is one way to make room for innovation -- but its longevity is likely to be shorter than the first two approaches.