A Report Card for the Net Promoter Score

Should you use the Net Promoter Score? Maybe, maybe not.

We’re not here to debate whether you should use it or not (and you may not have a choice). Instead, we want to use data (rather than opinions) to review and grade 13 claims made about the NPS, some from NPS critics and others from NPS proponents.

At the end, we give a report card on how well these claims stand up against the evidence.

This is a preview. You can read the full article on MeasuringU's Blog.

Summary

Figure 4 shows the final report card for these claims about the NPS.

Our review of these claims and the evidence for and against them somewhat favors the ability of the NPS to predict growth, but it’s not consistently better than more traditional measures of satisfaction. It is also partially predictive of discouragement, but not well enough to substitute for specific discouragement metrics when carefully measuring intention to discourage is important.

On the more positive side, the NPS is a reliable metric and, despite seeming “wacky” to some critics, Reichheld’s box scoring produces a reasonable partitioning of LTR responses into the three NPS categories (detractor, passive, promoter) with evidence that promoters actually promote and detractors actually detract. The single LTR item is sufficient to measure recommendation likelihood, the eleven-point scale works well in practice to identify extreme responders, and future recommendation works a little better than past behavior when predicting future recommendation behavior.

Regarding when the NPS should be used, neither of the extreme claims (always use it or never use it) is warranted. The decision about whether to implement the NPS in a company depends on whether it’s possible for their product or service to reasonably be recommended. Companies that already use the NPS should probably continue to do so, while companies that use a standard satisfaction metric should probably continue to do that.

This is a preview. You can read the full article on MeasuringU's Blog.

MUiQ: Measure Your User Experience

Full Control Over Your Study Design

Sophisticated UX research often requires more than a simple, linear path.

To get valid and reliable data, you need to control for variables like learning effects and order bias. This means having the power to randomize tasks and conditions, and to direct participants down different paths based on their specific feedback.

MUiQ is built to give you this level of control.

The platform is highly flexible, allowing you to implement the study design that best fits your research questions. You can easily set up conditional logic to branch participants to specific questions or tasks based on their previous answers. MUiQ also fully supports both between-subjects and within-subjects study designs, with robust features for randomizing the order of tasks and stimuli to ensure your data is clean and unbiased.

Whatever your study design, MUiQ provides the flexibility you need to execute it with confidence. Learn more about designing complex studies in MUiQ.