.
Quality is more difficult to measure with services than with manufacturing due to services" characteristics of being intangible, inseparable, and by the fact that they vary with where, when, how, and by whom they are provided. These factors create many challenges in deciding what data is of significant value, how to measure it, and in defining what is indeed a problem with the service being provided. .
With services, it may become difficult to account for all sources of variation, as some significant sources may be obscured by ineffective process measures, ineffective recording of information, and because data values may have multiple sources and causes of variation. Using simplified, broad measurement units or rounding-off may result in data that falls outside of control limits even when the process is in control. The easily quantifiable, easily calibrated, discrete measures in the manufacturing environment are not so distinguished in the service environment. .
Human factors inherent in service industry data create many challenges in trying to use methods of Statistical Process Control. There are significant chances of motivational use of data (e.g. during performance appraisals) which could lead to dysfunctional measurement. Simply setting targets such as "answer the phone in x seconds" or "take more calls per customer service representative per day" may not lead to improvements in customer service quality at all. Employees may learn to "cheat", do anything they have to make their numbers and avoid being paid attention to. People tend to do what you count, not necessarily what counts. This makes it much more difficult than with the typically unbiased data monitored in a manufacturing process. .
Effective implementation of SPC can result in increased productivity, more consistent and reliable services, and improved customer relations. The key is to ensure that its application remains relevant and useful, and that it is used to monitor the process, not the employees or the product.