Recently, a customer asked, “In terms of “best practice”, do you have a view on whether NPS should be calculated based on the date a survey was sent or the date of the response?”
My response was perhaps not as specific as they had hoped : “It depends.”
In many ways, it doesn’t matter which you choose so long as you stay with the same date. The client was concerned about being “most accurate” when calculating the score but this is a fuzzy concept here. “Most consistent” is probably more useful idea in this case. A consistent NPS data collection process is key in obtaining data that you can trust and action. See this blog post for more on this topic: Three Prerequisites to setting Net Promoter targets.
When running a transactional survey approach, another date that could be used is the order date or transaction date. If you use this date, you can potentially tie changes in the customer scores to events in the order or touch-point process. This can be very useful in the root cause analysis process.
Some organisations like to use response date because that means that as of a particular date the reported scores will not change. If you use order date or date sent, the reported NPS for, say, February can change during March if someone fills in a survey in March that was sent in February.
However, because we’re most often talking about email surveys this effect is quite small. Email surveys are normally done within a day or two of being sent so by March 5, nothing will be changing in the February report. That may however be enough of an issue to make you want to change.