Notes on Exploration and Exploitation

Questions & notes ch 6 - Warglien

Exploration and exploitation

Exploration and exploitation are two important strategies for an efficient company. These are different tasks, which require diverse structures and cultures. A successful company is the one who is able to manage an effective trade-off among the two.

Exploration

On one hand, exploration is a strategy mainly pursued by start-ups, it involves radical innovation and high levels of risks. It requires the ability to think “outside the box” and see opportunities in broader domains and over the current limits. Explorative strategy requires an organic structure, where information is widely shared and there are high levels of communication, thanks to cross-functional teams.

Exploitation

On the other hand, exploitation is about exploiting current organizational resources to create continuous value. Therefore, it is not only a matter of static efficiency but it involves great levels of incremental innovation, where an existing output sees a step-by-step improvement to meet the more demanding consumer needs. Exploitative strategy is common in more mechanistic structures, where the design facilitates greater focus and execution. Incentives are strong based on quantitative measures, since behavior and tasks can properly be described in advance.

However, exploration is harshly measurable because of the great uncertainty companies face. Thus the appropriate behavior cannot be specified in advance, consequently, the tasks are not highly measurable, which lead performance pays to be subjected to real randomness. As a matter of fact, exploration determines a motivational problem, which the company overcomes through two kinds of incentives that encourage people to be risk-taking. The first are stock options – options to purchase a company stock at an agreed-upon price within a certain period of time – which allows a direct reward to the employee as soon as his/her ideas take off in the market and generate returns to the firm. The second is a more intrinsic motivational reward, which increases the value of the worker in the internal labor market.

Nevertheless, both exploitation and exploration are extremely important to a company and they must be managed simultaneously in order to be innovative while continuing on delivering good products or services to the customers. As a consequence, these opposite tasks determine the multitasking problem.

Managing exploitation and exploration

How can an organization develop a design to manage both exploitation and exploration? Often many firms solve the problem with the division of jobs, separating explorative from exploitative units. This eliminates the innovation problem but it creates a coordination problem among complementary units. Moreover, internal competition for resources arises. Indeed, subunits are incentivized to increase their effort, pursuing only their self-interest without taking into account the impact their actions have on other divisions. In conclusion, it is impossible to eliminate the multitasking problem, someone in the company has to bear its cost.

So, the responsibility to manage multitasking activities is shifted to the top management team. Nevertheless, the latter doesn’t have the first-hand knowledge to meet the goals of the required tasks, so it results inefficient to delegate this responsibility to top executives. Decision making has to be based on second-hand knowledge due to high information asymmetries.

Therefore, companies try to separate completely the two different tasks but, in order to avoid the “creosote plant” effect. So the explorative units are separated completely via “skunk works” model. The term refers to working on top-secret projects, therefore the teams are isolated completely by the organization.

Nevertheless, the best way to solve the multitasking problem is to employ the use of the “high commitment human resource management,” which is about creating a culture and a related design that emphasizes the values such as collaboration, cooperation, and mutual respect. The key elements of the latter are guaranteed employment, self-managed teams, hard work, premium compensations, extensive socialization and constant training for employees, and transparency of information. The integrating mechanisms are mutual adjustment and social pressure.

Define the problem of observability and how it relates to moral hazard

At first, we have to explain the problem of motivation in order to understand the observability issue. People are selfish, and they tend to pursue their own interests without taking into account the impact that their actions have over the society, in this case, the firm. This problem arises because the employee doesn’t produce something that directly he would consume. Moreover, he doesn’t directly receive all the benefits borne from his labor, because these accrue to the firm and the agent gains only a small fraction. Therefore, he becomes reluctant to work as hard as if he would receive the full fruits of his labor. So, the motivational problem arises, the principal has to find a way to incentivize employees to work on her behalf rather than their own self-interest. Incentives will be based on the performance measures of the agent actions, and they will determine the strength of effort required. The stronger the incentives, the harder people will work, mathematically we can describe this relationship as a monotonic function. The issue, now, is to determine the performance measures and the problem of observability arises. Ideally, the principal and the agent would sign a contract where the appropriate behavior is specified in advance so that she could buy the exact amount of effort and give the perfect incentives. This contract, in order to be valid, has to be enforceable by a third party but it is almost impossible to monitor directly the actions of the agent so the principal has to accept to monitor performance in an imperfect and unclear way. The most she can do is to control the final results of the agent’s actions and to provide an incentive scheme, a contract or compensation that rewards or punishes the performance of employees’ actions. Due to this impossibility to directly monitor actions, the agent has the opportunity to act opportunistically since the performance is a function of his effort, but also of other external variables, called noise, that blur the effective measurement of the final results. For example, an employee might give very little effort to his work, but thanks to pure luck his performance results very efficient. Hence, the problem of observability leads to the moral hazard.