Editor's note: Illona Guzman is a marketing research analyst at Phoenix based WestGroup Marketing Research.

A shopper study, sometimes referred to as a "mystery shop," is a marketing research technique of examining service quality within a service industry. Professional "shoppers" visit or call various customer service locations and evaluate the service they receive by filling out a questionnaire after each visit. Information obtained from shoppers can be objective (e.g., How long did it take before you were served?), or subjective (e.g., How would you rate the employees on friendliness?). Usually, a combination of both is used. Results from all visits are tallied and data processed, so that a score on each measure is achieved. Scores can then be evaluated in total (i.e., among all locations shopped), by region, or by individual location.

Shopper studies pinpoint specific ways for service industry businesses to improve their performance and differentiate themselves within their industries. The study accomplishes this by providing a current, objective, factual report on specific service levels, as opposed to customer satisfaction surveys, which can rely on generalized measures from a broad perspective. The study is used mainly for training or comparison purposes. The main objective is to identify needed changes so that quality of service improves, thus improving customer satisfaction.

Shopper studies have made extensive use of in-person visit to banks and various retail establishments. In-person visits can capture the entire experience. However, when a prospective customer is looking for a new bank, or an auto parts store at which to buy an accessory, or even a restaurant at which to dine that evening, the decision is often made from the telephone conversation. Telephone shopper studies, therefore, provide information about this important aspect of service.

The experience of the telephone conversation is short and highly impressionable. Once the call is completed, the burden of action is on the prospective customer; it is up to the voice on the other end of the phone to entice him in. Consider this excerpt from a banking scenario:

A young lady's voice answers the phone with what must have been the bank name (at least it sounded like the name in the phone book), her name (it was either Sherry, Sharon, or Shelly), and the most mechanical "can I help you" you've ever heard. You tell her you're new in town and that you're looking for a bank.

The first thing she asks is, "Do you have $1,000?" If you did, she says, then you would be eligible for their Super Packaged Account that gives you " a whole bunch of free services. If you ever got $5,000 saved," she says, "you would then qualify for the SuperDuper Packaged Account that gives you even more." If you wanted, you could come in and talk to a PB (or was it BP or ET? You guess that must be some sort of new accounts person.)

This scenario is an example of lost potential business. In a mere 90 seconds, the banker demonstrated tactlessness, an unprofessional attitude, and lack of concern.

The telephone shopper study is simple and provides a relatively low cost method for evaluating service over the telephone. Any service industry can benefit from this type of study. Here is how it works:

Professional shoppers call your place of business with designated scenarios. After each call has ended, the shopper fills out a questionnaire about specific aspects of the service they received. Scoring is calculated for all measures after all responses are tallied.

Numeric measures can be collected, such as number of times the telephone rings, amount of time on hold, or number of times transferred before being helped. Other objective measures can also be obtained. These bits of data result from the yes/no types of questions, such as whether or not the employee sounded friendly, whether the employee spoke clearly, or whether he mentioned a specific promotion. Subjective information can also be collected. This is obtained in the form of ratings. Instead of asking whether or not the employee was friendly, if he cross-sold products, or if he enticed you to come in, the shopper is asked to rate how well the employee did these things. By asking the shopper to evaluate depth on a one to five scale, more detail can be achieved. The scores from each measure can be compared with competitors or among different locations. Summary scores can also be calculated by weighting different measures accordingly.

The telephone shopper study can be used in conjunction with the personal shopper study to get a broader view of service quality, or as a separate study. Both can be used as inherent parts of quality control program. They are great for employee training, as well as for competitive or branch-wide comparisons. Used as a tracking study, shopper studies tell the story of change over time. They can be used to reward specific locations for improvement since the last set of shops. Shopper studies can also be used to commend individual employees or to reward specific locations for outstanding achievement on a single set of shops.

The study is as large or as small as you make it and can be tailor-made to specific needs. All locations can be shopped, or a sample of them can be shopped. The number of calls per location can vary, too. Shopper studies usually consist of only a few cases per location. For a large company with many locations, the overall sample is generally large enough to be treated as statistical data, but information on an employee or a location is merely indicative of their performance.

Here is an example of how the telephone shopper study has been used:

A major bank noticed that one of its locations was not getting as much loan business as the other branches in the area. Looking at the data from their telephone shopper study, the answer became evident. Half of all auto loan scenarios and two-thirds of all home equity loan scenarios resulted in transfers to the telephone loan department, an occurrence that was only supposed to happen during the bank's peak activity time. When this matter was investigated, it was determined that three of the employees were uncomfortable giving out loan information and that several others were simply tampering with the system. Employees were re-trained on the loan process and re-educated on the bank's transfer policy. Customers subsequently began to be invited into the branch to personally discuss the loan process and fill out an application.

Shopper studies should not be used in a negative way to discipline poor performers, since each shop provides information about a single instance. Management should be alerted, however, to any blatantly negative experiences, as well as to consistent performance problems with an individual or with a particular location. These instances can then be investigated further. Poor results indicate a need for positive training.