Editor’s note: “My boss hits me over the head each time he thinks I’m spending too much time on the phone dealing with a customer’s issue.” Ouch! That was the way a caller put it while I was doing a radio interview some time ago.

I was being interviewed to describe my espoused principles and practices on service to the customer, and while I was touched by the man’s distress, I’m afraid to say, I had seen this movie before. Clearly, the caller, a seasoned service worker of many years, couldn’t see eye-to-eye with his boss on a metric for gauging the effectiveness of a response to a customer’s trouble call. Insofar as I could gather, the service worker was motivated to answer all of a customer’s questions before ending a call. The boss might have had the same motivation except that it seemed to have been filtered by some unspecified productivity or financial consideration – how much time was too much time on the phone? What constituted a successful completion of a call?” –  which had not been communicated to or embraced by the service worker.

In behavioral economics, this is a classic principal-agent problem. Principal-agent problems exist in business, politics, and in life generally anytime one party is empowered to act in another party’s behalf. In our context, the boss – the principal – and the service worker – the agent – had failed to align their goals and expectations of service performance to the customer. The crux of the principal-agent problem is a failure of the principal to explicitly stipulate incentives – tangible as well as intangible – that channel the agent’s efforts and with which the agent is in full accord. Left unresolved, misaligned goals lead to conflicts of interest where each party behaves in its own self-interest. The upshot of this behavior, if not stemmed, can become internally dysfunctional to the organization, and easily telegraphed to the marketplace. Customers are very adept at discerning ambiguous service commitments and when they do, they do the only logical thing in their power: they vote with their feet.


My suggestion to the caller was to sit down with his boss and see if they couldn’t agree on a standard of performance. In effect, I was suggesting that principal and agent codify or “contract” the measure(s) of service to the customer that they both could agree on. Concrete measures of performance are indispensable: they allow the principal to monitor the effectiveness of the agent’s work and, just as importantly, it allows the agent to police his own efforts. This is crucial if the goals and expectations of the players involved are to be harmonized. These contract measures can be made all the more effective if the agent’s compensation is linked to his agreed-upon performance.

Equally crucial, the boss needs to see that the benefit of having a satisfied customer at the end of a call far outweighs the cost of the agent’s labors – the so-called agency costs. Browbeating or otherwise punishing the agent has the potential to increase the principal’s agency cost if he is left with no alternative but to remove the “offending” worker.  Conversely, the benefit of terminating a call before a customer has had all of his questions answered pales in comparison to the potential damage that a dissatisfied customer can have on the business. In effect, a service interaction is complete only when the customer says it is and not before. And, certainly not after some arbitrary number of ticks on the clock have expired.

Fundamentally, principals must recognize the wisdom of devolving power to the front-line service worker: the agent most entrusted with delivering service to the customer. Further, the principal needs to be egoless about this reality and about the fact that the service worker knows far more about his subject matter – customer service issues – than he does. A less than confident and trustful principal will fail in his relationship with the agent no matter how extensive the “contractual” obligations agreed to by the parties. Similarly, a less than fully competent service worker will likely fail in his ability to hold up his end of the bargain and find himself subject to remedial actions by the principal.


Complicating the principal-agent problem in servicing the customer is that customer expectations are fluid. They constitute an ever-changing standard of performance rendering customer satisfaction totally ephemeral. For that reason alone, it is imperative to constantly and continuously reach out to the customer and gauge his evaluation of the level of service he is receiving. It goes without saying, for example, that a customer can have all of his questions answered by the service worker on the phone and still walk away a dissatisfied customer.  

Principal-agent problems cannot be left unresolved. They can only fester and do nothing but damage even the most apparently solid customer relationships. Striking a principal-agent agreement mutually acceptable to both service worker and boss is essential to delivering the best service the organization can possibly muster. But, recognizing that excellence in service pivots on harmonizing principal-agent dynamics is a most important first step in achieving that goal.


The so-called sharing economy in which online platforms network those who drive cars (Uber), rent apartments (Airbnb), clean laundry (Washio), deliver groceries (Instacart), sell goods (eBay), or render handyman services (TaskRabbit) with consumers seeking those services sheds a different light on what we understand of principal-agent dynamics because of the muddled mess that a network of multiple stakeholders presents. Distinguishing principals from agents in these companies is a challenge as there is no common employer and no shared mission.

Some preliminaries, however, are in order to better come to terms with this issue. The “sharing” economy is not about sharing at all. It is, more correctly, about selling or renting assets or about selling labor. The “gig” economy, which we discuss in a separate essay, Service to the Customer in the Gig Economy is a more fitting term for on-line platforms which sell the services of laborers – blue or white collar. Companies such as Uber or Instacart clearly fit that category. On-line platforms that either sell or rent assets are capital-based or asset-based platforms and are therefore unlike their gig brethren. Airbnb and eBay are notable examples of capital-based platforms.

Most companies understand the benefit of enforcing operating and service consistency in order to build and protect their brands. The business models of on-line platform companies, however, are predicated on the ability to empower their service workers – drivers, landlords, or carpenters – to operate almost totally independently of the brand they are supposed to represent. Consider that Airbnb has four million listings hosted by nearly 700,000 landlords around the world. What are the chances that the networking giant and its landlords are on the same page when it comes to serving the customer?  Uber, for its part, has over two million drivers.

Consumers seeking low cost and digital convenience from on-line platforms need to understand that seeking redress for service, liability or other issues is an area of compromise that is part of the bargain of doing business with these companies. If an Airbnb tenant finds – as has been reported in the press – an apartment bedroom with a concealed webcam, an apartment almost totally devoid of furniture, with dirty linen, holes in ceilings, holes in floors, inoperable appliances, and swarms of insects where does he turn? That’s not the worst of it. Murder and rape have also been reported at Airbnb facilities. In New York, a career criminal was found to have been in possession of duplicate apartment keys which allowed him to rape the unwary tenant. If a Uber customer finds that the driver is inebriated, does not know his way around town, maneuvers unsafely or drives a junker, where does he turn? Or what if a Uber Eats driver has “outsourced” his route to someone with a criminal record, never mind a person without working papers?

A proactive leadership group would seek to grapple with these principal-agent issues by providing enhanced vetting, monitoring and grading of service workers, and by offering more frequent education and training programs. These initiatives would clearly lead to an uptick in operating cost and with millions of service workers in these networks the cost impact would be significant and threaten the very business model on-line platforms are trying to perpetuate. Service to the customer, however, might just be the beneficiary

Management Advisor


Leave a Reply