What are the common notions about "lost knowledge?" We normally think that employees retiring or moving to other firms are "lost" and that they walk away with intellectual capital that is gone forever. But a recent CIO Magazine piece from Knowledge@Wharton appears to contradict that notion. The Social Network Benefit:
But a new study suggests that losing an employee, at least in a high-tech field, is not necessarily as bad as it seems. "Firms can wind up learning when employees leave their firm, which is contrary to the conventional wisdom—that firms learn by hiring away employees," says Wharton management professor Lori Rosenkopf. She and Wharton doctoral student Rafael Corredoira present their conclusions in a paper titled "Learning from Those Who Left: The Reverse Transfer of Knowledge through Mobility Ties."
[via Bruce Hoppe via Odin Zackman]
I would find it difficult to imagine someone who moved from Spacely Sprockets to Cogswell Cogs sharing knowledge with their old colleagues - and that's not quite what was discovered in this study. The researchers focused on the "reverse transfer of knowledge" through social ties that survived across companies in the semiconductor industry. They emphasize that this isn't corporate spying, but the flow of ideas across social networks, like meeting one another at industry events. This is social capital vs. intellectual capital in the discussion of human capital.
Thinking about it, isn't this why similar companies happen to be located in the same geographic areas? You want to start your tech firm in Silicon Valley or Seattle or Boston because that is where the talent is. But there are also deep networks of people who have tried, succeeded, failed with many others. These industry networks provide access to knowledge on many aspects of the business from the technical specifics to operations to legal to finance...
Deeper into the article, it talks about this insight as well. It turns out that people who move to firms in different geographic regions tend to offer more opportunities for reverse knowledge exchange than do those who move to new companies nearby. The researchers suggest that the local networks are more redundant amongst companies, whereas companies that are geographically removed have fewer ties.
These research is based on study of patents in the semiconductor industry. Reverse knowledge flow is suggested when patents cite the work of the new firm after an employee has moved.
Bruce Hoppe's comment is that there is also an important learning opportunity that happens at a time of transition / loss. Interesting idea, as the action taken during the transition is likely to have more immediate impact, while the network effect that this study suggests is likely to be longer-term.