The next segment we want to cover is interdependence. This is another challenge that you face as you do your talent analytics. So much of our work depends on the impact of other people. We rarely get to evaluate an employee in isolation. Rarely is our work done in isolation. We are humbly dependent on other people, so this poses a big challenge for talent analytics where mostly you're trying to infer the quality of an individual despite the fact that he or she works among a team. I had a conversation recently with an executive who works with Google and we were talking about exactly this challenge, and she got really mad she says it makes me so mad when people over emphasize individual contributions. Practically everything we do is with other people. If so and so left or so and so left I'd quit. I couldn't do my job. She was really worked up. I was struck by how quickly she got worked up. And this is a person who not only performs very well, but her job is to help people make these evaluations better. And she appreciates, as well as anybody I know, how difficult it is to parse the individual contribution, from the greater team impact. So interdependence is an important thing to keep your eye on as you try to make sense of the numbers. So, a great study by Boris Groysberg at Harvard from a few years ago demonstrates this. He looked at investment analysts, in particular rock-star investment analysts and their performance as they moved from one firm to another. So this is something that happens in that industry. An individual will rise and gain a reputation as having a particularly good bead on an industry or a set of companies and build up some equity, and then be poached by another organization, and they move across organizations. The nice thing about this setting is that we can track analysts forecasts against what actually happens and we can do this over years. And so we can evaluate how they perform in one organization and how they subsequently perform in other organizations. So Groysberg goes in, saves a view years of this,1988 through 1996 he looks at over a 1,000 star analysts and their performance at 78 different investment banks, as they rose up in one, and then moved or didn't move to another. He defines star as any analyst who was ranked by the magazine Institutional Investor. This magazine puts out annual surveys on the best in the industry, best analyst in the industry. And everyone in the industry takes these ratings very seriously and those who do well in these ratings are recognized as star analysts and often those are exactly the ones who get the offers to move to new firms. So Groysberg goes in and finds out what happens when these star analysts move firms? This is what he found. He finds that performance plummets by an average of 46% in the first year and even over five years it's about 20% below where they started. So on average, these guys not only dip dramatically in the first year, which you might expect a little bit of an adjustment cost, they never get back to their level that they were at when they were at the original firm. So what happens here? Does the star lose intelligence? Does the star forget the lessons of their experience overnight, and never get them back? It's unlikely, very unlikely that this is what's going on. What goes on is, they were part of a team before. They had researchers surrounding them, all the institutional resources that were available to them and they understood how to tap into, are gone now. And it's really tough to replicate that. Moreover, when we assess the ability and the performance of that individual analyst, we think of it as being the analyst. We don't consider the team they're surrounded by. We don't consider the institution they're embedded in. We think it's them, we pay them a lot to come work for our firm, and then we find out they were only one part of a broader machine. The impact, unfortunately, is broader than just the income of that particular group. The impact affects the group that brings them in. So from a head of research at one of these firms, the research head says, I painfully learned that hiring a star analyst resembles an organ transplant. First, the new body can reject the prized organ that operated so well inside another body. On some occasions, the new organ hurts healthy parts of the body by demanding a disproportionate blood supply other parts of the body start to resent it, ache, and demand attention, or threaten to stop working. You should think about it very carefully before you do a transplant to a healthy body. You could get lucky, but success is rare. So this is an example from the financial services industry, important example I believe. Stakes are high, performance is quite observable. It's kind of a best case for getting this right. And they don't seem to be getting it right. We observe it in other countries. Lord knows we observe it in professional sports. But this is systematic of the general challenge of interdependence. When you're trying to analyze how good somebody is, how successful they've been, you've got to somehow parse the individual performance from the team and it's difficult. The main thing is to be careful, to not over attribute. Individual performance is something that is fundamentally at the group level. So this is, one prescription would be, as much as possible, make your performance evaluations at the group level. And,convention is to make these at the individual level, but as we gain greater and greater appreciation for how interdependent performance is the performance evaluation should migrate more to the group level. We also might think about you're not real sure how someone's doing if it's so interdependent until you've seen them perform with multiple teams, you need to see them in multiple environments before you can reliably say they are the reason these teams are succeeding. And then finally, increasingly we get some performance measures that analytics is helping with this. That assess an individual's impact on a team. So there have been some studies recently, for example, in network analytics, where you can find, who team members go to for support, who team members go to for advice. So even if they don't show up as stars to those who are evaluating them, if you look at the actual behavior of the team members you can suss out who's making the biggest contribution. Techniques like that are gonna help us get a better and better lens on where that individual contribution is, parsing it from the interdependence.
