Q: Why ask three questions about satisfaction and not just one?
The CFI approach of asking three questions is based on customer behavior research from leading universities, perhaps the best known of which is the Kano Model.  Since customer perceptions are both conscious and subconscious, measurement must close the gap between the voice-of-the-customer and the mind-of-the-customer.

By asking three questions – overall satisfaction, meeting expectations and being ideal – CFI builds a composite measure which offers superior reliability (freedom from measurement error) and precision of scores compared to a single-item overall measurement. Precision and reliability are essential to effective action planning.

Q: How do we make results more actionable?
It is clear that many organizations struggle to make their measurement programs actionable. The fundamental ingredient to effective action planning is predicting the impact of individual drivers of satisfaction on the desired outcomes of satisfaction.

Accuracy, precision and reliability are critically important but not sufficient. It is also critical that you measure the right things (garbage in, garbage out). In measurement systems, this starts with designing a survey that effectively identifies and correctly measures the high-impact drivers of satisfaction. This is why up-front qualitative analysis and specialized expertise in measurement design are so critical to delivering meaningful, comprehensive and reliable results.  When properly performed, qualitative research helps assure that survey questions are developed that relate to business processes.  Once the survey results are received, it is then clear where action is needed.

Q: How do you relate survey measures to business processes?
A successful linkage of survey-based measures to business processes or other hard metrics requires three things:

1. Precise, actionable, survey information
2. Available, managerially useful, operational metrics
3. Proper 'units of analysis,' i.e., data points that can be assigned both a survey score and an operational metric

Items (1) and (2) imply that the more sensitive the survey instrument, and the more useful the metrics, the better the linkages will be. But absent a way the metrics and survey measures can be tied together, a linkage is impossible.

Only rarely can a particular operational metric be linked to a particular survey respondent. More typical and more valuable is a link of aggregated survey responses with operational metrics. Sometimes that unit of aggregation is at a store or branch level -- say, linking changes in store revenues to changes in customer satisfaction with the store. Other times that unit of aggregation is time -- say, linking monthly changes in queue times at a call center with monthly customer scores of call center performance. Geography or customer type are some other potential levels of aggregation. CFI Group then uses a variety of statistical techniques to estimate the linkage, ranging all the way from simple correlation to advanced non-linear modeling.