The Journal The Authority on Global Business in Japan

Many experts share a common problem: our customers cannot tell, even after purchase, whether we’ve done a good job. Translators, attorneys, doctors, educators, and other knowledge professionals are prone to this problem. Economists say that we’re selling “credence goods.” Credence goods are products and services that cannot be evaluated even after purchase.

Because purchasers cannot directly assess quality and true cost, they must find an indirect method of evaluation. For example, a second translator may be hired to perform a “back translation,” or a second accountant may be hired to audit a first. But not all purchasers want to hire two professionals to do the same job. When cost is an issue, price, certification, or reputation may be used to decide among alternatives. Higher-cost professionals may be assumed to be more trustworthy. Those with certifications may be seen as better than those without. And the same product from a company with a better reputation may be valued more.

Exactly how much does reputation matter? It depends. (As a mathematician, this is always the answer I give when it comes to real-world problems.)

A mathematical model can help us answer this question. Luis Cabral of the Stern School of Business wrote an excellent primer on a game-theoretical model of reputation for experience goods, which can be evaluated after purchase.

For credence goods, a different model is needed. An epidemic model is useful here. Epidemic models are used by epidemiologists to predict the spread of disease in the population, but such models can be applied to other problems. For example, videos that have gone viral are often studied in this way.

Let us consider the SEIRS model. In this case, potential customers begin in a susceptible state (S). They will be open to purchasing your product in this state. When they encounter your product, they have been exposed (E) and might spread your reputation but not purchase. Just as it may take multiple exposures to catch a cold, a potential customer may need multiple exposures before they will seriously consider a purchase.

In epidemic models, the ease with which the infection can spread from one host to another is its infectiousness. When we apply this to businesses, improved reputation will increase the chance that an exposure will lead to “infection” (I): that is, purchase or recommendation to others.

The three key factors in the spread of a disease (or video, news story, etc.) are the network (how likely exposure is), the infectiousness (how likely exposure is to result in spreading), and the removal (R) factor. A highly infectious disease that immediately kills or immunizes those who catch it will not spread far: all potential carriers will be removed from the population.

In the business context, R represents many factors, among which service is under our control as sellers. Because purchasers of credence goods cannot directly assess quality, customer service is critical: a bad experience will both damage overall reputation and immunize others from purchase.

The final “S” in the SEIRS model is (again) susceptibility. Customers who have a bad experience—whether our fault or not—will sometimes give us a second chance.

The SEIRS model also suggests how to improve reputation while building sales. Opportunities that expose us to potential customers in a way that highlights expertise are going to be the most effective. This is because they target both the “E” and “I” steps. Educating potential customers, cooperating in the establishment of industry standards, and being vocal about what a good company/professional looks like in your industry will improve your exposure and your reputation.

As an added benefit, improvements to an industry typically increase the cost of entry, which reduces competition. In selling credence goods, this often results in overall improvement of quality because the competitors that are kept out are those least likely to provide good quality.


Chad Musick is vice-chair, ACCJ–Chubu Business Programs Committee and operations manager of  ThinkSCIENCE K.K.
Because purchasers cannot directly assess quality and true cost, they must find an indirect method of evaluation.