A streets department manager is feeling pretty good as he reviews his latest community satisfaction survey: 85% of the respondents rate the experience with his department as good or better.

Meanwhile, though, six people are queued into his department's call center, and four have been on hold for more than three minutes. Eight were misdirected out of the call center, and three hung up in frustration over the telephone menu. Two residents are having an awful time finding a parking space while two others are lost in the building due to incorrect signage. Three residents can't find the answers they want on the city's Web site, and two have been waiting three days for a simple answer to an e-mailed question.

Because the survey doesn't provide details on why 15% of respondents did not give a good or better rating, the manager is oblivious to these customer-service hiccups.

“We've used surveys for a number of years to gather data regarding county services,” explains Brian Cox, director of egovernment/customer service for Mecklenburg County, N.C. “However, survey data generally lacks the depth of information needed for improving customer service processes.”

So, to pinpoint practices in need of fine-tuning, Mecklenburg County hires mystery shoppers.

Mystery shoppers are trained to pose as customers and evaluate a service anonymously. Most times they're real customers, in that they are from the community and make a purchase. But their purpose is to appraise a list of service standards and provide analytical assessments.

Though used in the private sector for years, the public sector has only begun using the service. Singapore uses it to assess a wide variety of government services. In Britain, mystery customer audits are used to evaluate social services. And here in the United States, more cities are hiring private consultants to shop their services.

Goals vary depending on an organization's needs, but local governments typically seek an overview of service consistency and effectiveness. The results identify opportunities for improving the ease and effectiveness of communication with constituents.

This could mean evaluating how easy it is to navigate a department's telephone system. It could mean measuring the attitudes, skills, and knowledge of staff encountered in visits to government offices, or how long it takes to find the correct answer on a Web site. It can even look into how long it takes employees to answer e-mailed questions, and how well the responses address the issues.

Mystery shopping is valuable as a research tool because it's real time, not retrospective; detailed, not surface-level. “Survey data reveals where a problem exists, while mystery shopping provides insight into the cause of the problem,” says Cox.