5. I’m the Expert: Why Connections Are More Important Than Test Scores

When I was in the fifth grade, I got blisters on my thumbs. The cause wasn’t an overly exhausting rock-climbing expedition or working long hours at the wood shop. No, my blisters were caused by a far grander, nobler endeavor: I was playing video games.

Over the course of the year, I conservatively estimate that I played Nintendo for about 2,000 hours in total, or about 5 hours a day. During this time, my games of choice were Final Fantasy and Contra. It was while playing the latter that I developed the aforementioned blisters.

Contra is an action game where you control a commando fighting alien hordes. As you progress through a series of levels, you must dodge enemy fire and dispatch said enemies with your own attacks. If you’re hit once by any enemy bullet you instantly die, and the fast-paced nature of the game means that to succeed you need to quickly change directions while simultaneously mashing the “fire” button.

After approximately 1,000 hours of playing Contra, I had become an expert at this game. I knew exactly how to react to my enemies, and I had learned the strategies for vanquishing the various bosses that appear throughout the game.

I had become the prototypical expert. I had mastered every aspect of a particular (if narrow) field, and I could perform at levels well beyond those of novices and intermediate players.

Throughout our lives we go through similar phases in many different areas, for most of us outside the world of gaming. We go to college and become experts in a field of study, we work for a company and become an expert in a technical area, we get married and become an expert on a particular relationship, and so on. This learning and evolving of our understanding is central to the human experience.

The way organizations have nurtured and helped us refine our expertise, however, has changed drastically over time, mostly due to the ever-increasing complexity of the type of work that we do. As a concrete example, compare a computer programmer posting from 1958¹ and one from 2012:

Although the job description itself looks humorous and quaint today, there are a few things to pay attention to. One is that almost no pre-qualifications are needed for a job that can be incredibly complex. Applicants essentially needed to take a few classes in programming and be interested in the field to qualify. Another is that the specifications even for a senior-level position don’t require experience working in this field.

Now look at a recent job listing for an entry-level programming position:

From a skills perspective, the recent job posting is incredibly complex. The applicant needs to have trained as an undergraduate for four years to learn this field and have specific experience developing programs using the languages the company uses, which speaks to the larger trend of the increasing need for training and expertise in today’s world.

Imagine a programmer from the 1970s going to apply for a position today. He would be quickly shown the door. On the other hand, a programmer from today could easily go back and win a position a few decades ago. Today’s novices are yesterday’s experts.

This increasing reliance on expertise has caused a shift in the way that organizations and potential hires view their relationship. In exchange for an employee’s devotion to a single organization, a company was obligated to train him and nurture his skills until he became an expert in his field. In the past, companies looked to acquire the best people who could learn the skills they needed to succeed, whereas today employers look for people who already have all the skills.

To a large extent, this trend is being driven by societal forces. As Chapter 3 covers, there are many good reasons to broaden one’s experience and dip into many different fields. This type of experience enables workers to think creatively about problems and bring diverse perspectives into organizations. Unfortunately, it often means that companies won’t worry too much about rewarding experts for the work they do by informally training other people. After all, if employees are just going to leave in a few years, why reward someone who’s training them?

There are, however, some big exceptions to this rule.

As of 2008, one out of every 27 CEOs of publicly traded companies had at some point worked in one of two companies: General Electric and IBM.² General Electric (GE) can count CEOs of Fortune 500 powerhouses such as Boeing, Pfizer, and Home Depot among its alumni. The basis for this success can be found in GE’s dedication to nurturing its employees.

When GE identifies someone as a high-potential employee, they start rotating them through jobs in unrelated fields to build up their overall knowledge of the company. Current GE CEO Jeffrey Immelt, for example, started in the appliances division, moved to plastics, and finally ended up in medical systems. Even though training and integrating a new employee in these divisions takes time, GE strongly believes that this rotation is a key driver of their success.

Beyond changing team assignments, GE invests heavily in developing curricula and courses for its 150,000 employees. GE spends about $1 billion (with a “b”) annually on corporate learning initiatives, which involves about 9,000 people a year physically traveling to GE “schools” for training, while about 60,000 people a year go through formal channels to enhance their skills.

As an organization, GE puts its money where its mouth is in regards to learning. It has numerous formal training programs that connect employees with the experts they need to succeed in their careers.

Interestingly, in mid-2012 this philosophy began to shift.³ As expertise becomes more and more important, GE realized that competing at the highest levels with employees who had very little experience in a single business unit wasn’t possible. Although top leaders must still be exposed to different business areas, the expectations are definitely lower than they were just a few years ago.

Although GE by any measure has been wildly successful over the last few decades, the need to constantly innovate at increasingly faster intervals has led executives to strongly focus on expertise. By keeping people in a particular division for longer periods of time, employees are encouraged to develop practical work experience and deep connections that will lead to even greater learning.

The problem for GE as well as other companies is that expertise development is typically attacked at a formal level. Training programs work great, but the company must be able to identify and keep experts long enough to make the training pay off. For practical experience to matter, a company must make sure that domain experts are actually teaching their less-experienced counterparts. The following sections investigate not only how these exchanges can be measured via the Sociometric Badges mentioned in Chapter 1, but also how you can reward employees for them.

• Where many people worked on individual tasks that didn’t require them to talk to each other

• Where most of the work people did was on computers and with people in other places

• A lot like the one where my team from MIT took the Sociometric Badges

The organization we studied sells multimillion dollar servers through a distributed group of salespeople spread across the country. These salespeople talk with clients to figure out what they need their servers to do and then submit these specifications to a computer system. This system automatically assigns this task to one of the organization’s engineers in its Midwest office, who actually lays out the physical hardware and tries to anticipate other requirements for the system. When he’s done with the configuration, the engineer sends it back to the salesperson for negotiation with the client.

An engineer is assigned these tasks on a first-come, first-served basis. After he finishes a task, he goes to the back of the virtual line and waits for his next turn. These tasks are all logged by the computer system, which knows exactly when he started working on a task, when he completed it, whether he made any errors, and even the difficulty level. These tasks take anywhere from five minutes to eight hours to complete, and although difficult tasks often take longer than easy tasks, that’s not always the case. To give employees extra motivation, their bonuses are determined entirely by how many of these tasks they can complete each day. From the company’s perspective, any time not spent working on a task is time wasted.

From this process one would assume that people don’t talk to each other. Why would they? Every second spent talking to a coworker is dollars ticking off an employee’s bonus check. There was a realization at this company, however, that something was amiss. Even though this division had only existed for a few months, some employees with high formal skill levels were struggling while employees with less experience on paper were the highest performers.

The IT firm wanted to understand these dynamics better, because clearly skill and task data alone weren’t able to predict performance. The social element was missing. That’s when our team from MIT went in with the Sociometric Badge.

In particular, the study focused on two types of behavior: individual behavior and social behavior. Individual behavior includes things such as movement levels, how workers tend to speak to other people, and how much time they spend at their desk. Social behavior, on the other hand, looked at who they spoke with and how well connected they were within the overall social network. Again, from the company’s perspective, individual behavior and skills should be all that mattered. Any time spent talking to other people was essentially time wasted.

Imagine the company’s surprise, then, when the strongest predictor of performance turned out to be whom an employee talked to. Specifically, how cohesive an employee’s social network was had a major impact on productivity. In this case, the more the people an employee talks to talk with each other, the more productive the employee was. Also, it wasn’t a small effect. If an employee spent 10% more time talking with her core group of contacts, she made the company roughly an extra $100,000 a month.

Those general results in hand, the next step was to understand what was driving differences in performance on specific tasks. Data from the company listed exactly who was working on a task and when they were doing it. This let us zoom in on the period of time that people were working on tasks and investigate whom they were talking to and how their behavior correlated with their performance.

Just looking at these patterns was extremely revealing. Plotting a social network diagram to look at who was speaking to whom when they were working on tasks revealed a bright center to this network. All communication paths eventually led to one of four individuals. Even more interestingly, the more central the person an employee spoke with, the more quickly she ended up completing her task. In fact, interacting with one of the four most central employees could cut task completion time by 66%.

These employees all had similar educational backgrounds and different levels of prior experience. They all had the exact same job title. From a formal perspective, they shouldn’t have had any more or less of an effect than anyone else, but the badge data had revealed the informal experts.

These experts were spending a large portion of their time talking to other people. Perhaps it’s not a surprise, then, that an expert’s individual performance was only middle of the road. They were simply spending an inordinate amount of their time helping their coworkers.

Although for the organization this isn’t necessarily a bad thing, it causes two problems. From the experts’ point of view, they don’t feel appreciated. Their guidance is helping to raise the performance of the entire division, but while their coworkers get rewarded, experts see no impact on their salary. They don’t even get any formal recognition of their contributions from the company. It would be one thing if they got promoted due to their efforts, but their job title remained the same as everyone else.

The problem from the company’s side is that without knowing it, management could easily cripple the performance of the entire division. Looking at the performance numbers, upper managers could have easily reassigned one of these experts to another role, which would eliminate the performance boost he or she gave to coworkers. The company just didn’t know that these people were serving a critical function. Similarly, these experts could finally say enough is enough and leave for another company where their work would be appreciated. Both possibilities would be catastrophic for this division.

That’s not to say that experts should be rewarded with monetary incentives. The difficulty is that without the badge data, knowing the exact effect an expert has on other people is extremely difficult. Although ideally companies could reward that impact directly, there are some other ways to encourage the general behavior.

One particular change we recommended for the IT firm was in their bonus structure. At the time of this study, this department’s rewards were 100% individual. Even if everyone else in the division did poorly, it had no effect on an employee’s salary or bonus. The beauty of these experts, however, is that they raise the overall performance level of the entire division. So it makes sense that their overall performance should be an integral part of any bonus or salary discussion.

Rather than make salary and bonus entirely dependent on an individual’s performance, we suggested making a large component of the bonus dependent on group performance. This could take the form of an overall performance target for the division or simply a graduated bonus that increases along with overall performance. Although this system would help experts receive more compensation, it also invites free riding. After all, experts end up getting the same bonuses as non-experts. Someone could easily coast along, not help anyone else out, and still get rewarded for the efforts of others.

In organizations this situation is, to a certain extent, unavoidable. Even without a bonus, an employee can choose to go out of her way to help coworkers or take an individual approach and focus on her own work. The hope is that by formally recognizing the value of expert contributions, an employee’s focus, group versus individual, will come into play when being considered for promotions or work assignment.

As this study demonstrates, informal instances of advice giving and coworker learning can have massive effects on performance. Unfortunately, most companies focus exclusively on promoting learning at a formal level. Companies such as GE invest millions of dollars in massive training facilities and course programs to help their employees develop. These programs are certainly valuable and end up building the base skill level of the workforce, but they are still just a base.

In the IT firm study, all the employees had the same base, but a variety of experiences nurtured the expertise of a few individuals. By promoting and valuing that expertise, companies can get a level of education in their workforce that is nearly impossible to obtain in the classroom. Problems in these domains are so intricate and complex that one person can’t exhaustively cover them all, but employees probably know someone who can help solve them.

This insight offers an important view into how to grow experts in an organization. The results from the IT firm study show that the most valuable experts aren’t just knowledgeable; they’re able to share that knowledge with their coworkers. The highest performers in the IT firm were probably experts themselves, but because they didn’t share their knowledge with anyone, they had a very small effect on the overall performance of the division.

Successful companies supplement their training programs and educational initiatives with efforts to encourage knowledge sharing. These programs cultivate base expertise levels and financially encourage employees to share knowledge. However, going a step beyond that and creating formal and informal frameworks would make it easy for people to discover experts. Many companies go the formal route, listing expertise on intranet profile pages or the like, but in nearly all companies these pages are completely ignored or forgotten. Also, an employee who can find an expert through these systems might not be comfortable contacting a stranger.

Publicly identifying and praising experts in meetings is a better tactic. Companies could identify who the highest individual performers are in different areas and encourage people to ask them questions, because if their individual performance is high, they likely have valuable information to share. More informally, during meetings people could discuss who they went to for advice over the past week and how it helped them in their work. You might hear a lot of the same names getting called out, but injecting new people into the mix can help expand the pool of experts.

This practice is a major cultural shift. In many organizations, an employee who publicly praises a coworker for helping him do the job might be seen as weak. In fact, this is actually a sign of great strength. An employee can only do so much work alone, but people who are adept at finding and popularizing experts, and sources of knowledge in general, are essential for the overall performance of a group. In many ways, these people are meta-experts: experts at finding experts.

These meta-experts have the wherewithal to continuously discover new information and the ability to disseminate it to others. Experts have similar skills, but they also can’t be afraid to advertise their experience. Compared to discussing advice one has received, this might seem like grandstanding. After all, an expert is announcing how much she knows and telling coworkers that they should look to her example for help in their work. Organizations need to shed that impression and instead realize that people need to communicate their skills. That is what ultimately drives performance in companies. It doesn’t mean advertising one’s self in a way that demeans colleagues, but looking for appropriate ways to share knowledge.

When we’re particularly adept at certain problems, we must realize our power to help out coworkers. The potential for a positive impact is simply too big to ignore. On the other hand, acknowledging experts while also actively working to discover other experts is important. These experts are the tendons that span hundreds or thousands of individuals, ensuring that employees can continue running together as a whole. A company without those tendons, no matter how much it has built up its muscle, won’t even be able to limp along.