Understanding how customers’ perceptions and behaviors have evolved over time is part of developing a complete picture of customer insights. In its most simplistic form, this can translate to trending customer performance. This simplistic notion is somewhat complicated by the fact that survey research is ever-evolving - mode can change, question text can change, scales can change, etc. Can customer performance be trended when methodological changes are implemented? The simple answer to this question is, “yes.”
Ask questions first…
When considering methodological changes, ask many questions before moving ahead...
1) How does the proposed change relate to the overall business objectives/customer strategies?
2) What other decisions are made using this information (e.g., process improvements, goal setting, etc.)?
3) Is compensation tied to any of the impacted measures?
4) What are the advantages to making the change (e.g., enhanced methodology, more relevant metric, etc.)?
5) What are the disadvantages (e.g., stakeholder buy-in, reliance on original metrics, etc.)?
6) Do the advantages of changing the methodology outweigh the disadvantages?
There are also more tactical questions to ask…
1) How is information reported (e.g., in total, by customer segments, etc.)?
2) What is the expected sample size?
3) What specific measures/scales are involved in the proposed change?
Careful consideration of these questions is critical before moving ahead. If the answers to the above questions lead to proceeding with the change, we arrive back at the original question -can customer performance be trended with methodological changes...
Implementing a solution…
One analytic method that provides a bridge when making a scale or text based methodological change is a regression-based conversion. This solution is simple to implement; the complication is in the design. We will talk more about the solution and the design next week…Amy Heleine
Director, Marketing Sciences