I’ve
been really impressed with the quality of the comments and the constructive
debate that was sparked by my last post on Fighting
the War for Talent Right. As I went
back and read the post and the smart comments, I realized that I had
unwittingly implied a dangerous half-truth: That doing things in a tight team
is just always better than having them done by a collection of individuals who throw their work into some kind of common “pot,” but don’t interact much.
Sociologists talk about this as the difference between a group with strong
“interdependence” versus one that has little or none at all.

If
you look at research on teams by social psychologists and organizational
researchers it turns out that Freud got it right about 100 years ago: Groups
bring out the best and worst in human beings. One my mentors is J.Richard Hackman, who has spent about 40 years studying
what drives group effectiveness. I
worked on a team effectiveness project that Richard led in graduate school,
that resulted in a book he edited called
Groups That Work, and for a complete and more recent summary of his
perspective see his Leading
Teams.   About 10 years ago, I recall
Richard expressing annoyance because the then best-seller
The Wisdom of Teams was being used as a justification for herding people
into tight-knit teams, on theory that teams outperform individuals.

It
turns out that if you look at team effectiveness research, the lesson is that if you can get the
conditions right, teams will outperform a loss collection of individuals on a wide range of tasks. But
there is also striking evidence that a bad team will bring out the worst in
people as well, and if you can’t get the conditions right it, it might be
better to organize them along the lines of a group dental practice, where they
share the same building, perhaps a common receptionist and a few other
resources, and each do their own thing. A similar lesson emerges from experimental research on the differences
between decisions made in groups versus individuals. A series of studies show
that mean level of decision quality is about the same, but there is much wider
variance in decision quality in groups
compared to individuals. In other words,
groups either to a really good or a really bad job of making decisions – I
believe that Max Bazerman’s book on Managerial
Decision-Making reviews this research.

The
upshot of this research is that if you have well-functioning team – or if you
want to hire one, to return to my last post – then it is worth doing all you
can to keep them together and to building trust and a shared point of view. But
if you have bad group, where peopleagree that few things are worth doing
well, don’t believe in learning, is rife with fear and so on, the best you can
do (if you can’t disband them… sometimes a wise move) might be to reduce the
interdependence between them, to organize their work so they don’t see much of
each other, don’t have to work closely on tasks, and – applying the group
dental practice model – at least aren’t dragged down by each other.

Now, I confess, this is not the optimal
approach, and I love talking about great teams and top performance and
all that, but the fact is, there are times when –- at least for the forseeable future – it is a great
victory if you can move a team from being downright bad to being simply
mediocre. The other implication of this
research is that if you are leading – or part of – and effective “group dental
practice” that taking steps to tighten the links between people carries a
high risk of making things worse rather than better, especially if you don’t have the
resources, time, and power to do it right.

So,
teams aren’t a panacea; they are more like magnifying glasses or multipliers
that bring out the best – and worst – in people.

P.S. If you
want to see a strong argument that
individuals matter more than teams or organizations, see Locke, Edwin, A.,
Diana Tirnauer, Quinetta Roberson, Barry Goldman, Michael E, Latham, and
Elizabeth Weldon (2001). The
importance of the individual in the age of groupism. In Marlene E. Tuner’s
Groups at Work.Mahwaw, NJ:Earlbaum, 501-528. We try to show that this the authors take their argument too far in  Hard Facts, but
this article is charming and well-argued.

Tom
Davenport and Larry Prusak wrote a wonderful book a couple years back that is
called What’s
The Big Idea. One of their most
clever observations was that, if you start digging into the waves of management
fads, you can see that the emphasis tends to shift among three different
themes: efficiency, innovation, and people.  They showed how, even though all three themes
matter to all organizations, at any given time, one and sometimes two of these
themes seems to be in vogue, while the other one or two are in the background.

If
you think back over the past 8 or 9 years, it works pretty well. So, after the
dot.com bust it was all efficiency all the time – layoffs and outsourcing. Then it was innovation and innovation is
still hot (but doesn’t feel quite as hot as it did about 6 months ago), and I
am starting to see signs – look at the cover of the current Economist – that people are starting to move toward center stage.

 If
you read this issue of The Economist,
if you consider standard HR practices for recruiting, hiring, evaluating, and
compensating employees, if you listen to most HR consulting firms, and if
you look at how employee records are organized in enterprise software you will
see that the – usually unspoken but pervasive – assumption is that a focus on
people means a focus in hiring the most talented individuals. Indeed, talent is the word people like to use
talk about good people, a word that conjures up images of superstar actors and
athletes. Certainly, having talented
individuals is important. But focusing on individuals alone – as the HR mindset
seems to do, in an automatic mindless way without ever questioning the
assumption – is a dangerous half-truth. It blinds managers and executives to a
growing body of literature that shows performance is heavily dependent on
having people who are experienced at working together and who work together for
a long time. Just look at that Economist’s
cover, at how implies a search for that lone beautiful pearl, one that is fixed
in its hardness and beauty. This analogy
mirrors the talent mindset and its fixation on individuals.

I
thought of this when Boris
Groysberg sent me a Harvard Business
Review article that he published earlier in the year with Andrew McLean and
Nitin Nohria.  It is called “Are Leaders
Portable,” and it examines what happened to 20 former General Electric
executives who were hired as CEOs between 1989 and 2001. The article focuses on
what they call annualized abnormal returns. In English, this means: How did the
company do in the three years after the new CEO from GE was hired compared to
market as a whole and similar firms?  The
main finding of the article – which I don’t think the article quite comes out
and says – is that these GE CEOs did a lot of damage, perhaps more harm than good. For example, they show that the 11 CEOs who
“matched” their new companies were associated with a 14.1% positive return,
while the 9 who didn’t match were associated with a -39.8% return. I actually doubt that the CEO’s alone had
this much effect, as more controlled research shows that leaders of large firms
usually have much weaker
effects on performance, but the number of firms that apparently weren’t
saved by an executive riding in from GE is instructive.

One
of the most interesting findings – and a challenge to HR’s obsession with practices
aimed at individuals rather than teams or networks – was that GE executives who
brought along 3 or more GE alumni to join their teams had “annualized abnormal
returns” of 15.7% above average; while those that hired one or none from GE had
-16.7%. Groysberg and his colleagues
call this past experience working together “relationship human capital,”
horrible language from economics. Other researchers call it “prior joint
experience,” which isn’t much better. But whatever you call it, while HR practices turn attention to individual
stars, study after study shows when people have experience working together –
and have learned who knows what, how to read those little signals that people
send off, and can communicate ideas quickly and efficiently – their teams and
organizations perform better. The list
of studies is long and uncover consistent findings. My Stanford colleague
Kathleen Eisenhardt found – in a large longitudinal study of new
semiconductor firms – that those firms that were founded by people who had
worked together before were more likely to survive and be financially
successful, and such positive effects got stronger as the firms got older. Exhibit one in this industry is Intel, which
was founded by the “traitorous eight,” a group of executives who fled from
Nobel Prize winner (and racist) William Shockley to start their own firm, which
included Gordon Moore and Andy Grove. Similar findings has been replicated in studies of start-ups funded by
venture capital firms, surgical teams that do coronary bypasses, airplane
cockpit crews, and product development teams. The only study I know that challenges this pattern showed that, after
about 4 or 5 years of working together, the productivity of product development
teams starts to slow if they don’t start moving out some old members and moving
in some new ones.  But productivity gets
higher and higher in these teams every year until about the fifth year.

The
implication of this research is pretty clear and shows the limits of modern HR
practices, assumptions, and even the enterprise software systems that they
use. If you are going to hire some
“talent,” don’t focus on just landing that lone star – focus on hiring as much
of his or her team, or network, as possible. You win the war for talent by bringing aboard talented sets of people, not
talented solo acts. Indeed, if you recall my earlier posting about the hazards
of mergers and acquisitions, I showed how the
typical merger fails, but I used the example of Cisco to show how when the
focus is on acquiring small companies and on going through great effort to
weave the people into the culture – and I would add now, keeping their
relationships in place – it can be quite effective. Some of the most effective companies – Google
appears to be an example – do this informally. People who are hired by Google are often pulled from intertwined
networks and groups that have worked together in the past. In fact, speaking of Google, if you look at
the history of the company, the two founders started working together when they
were doctoral students at Stanford, as did the founders of Yahoo! Paul Allen and Bill Gates met each other –
and started working on computer stuff – in high school, as did Steve Wozniak
and Steve Jobs.

Call
it whatever you want, but as the war for talent seems to be heating up again,
companies that fight it right will spend less time looking for solo stars and
more time looking for dynamic duos, teams, and networks of people that have
worked together in the past and want to work together more in the future.  And perhaps it is time for modern HR practices
to catch-up with the evidence.