I agree with you – an NPS score that is based on the average of an average of an average is not going to be all that useful.

It sounds like the organization in question is trying to take results from a bottoms-up NPS survey and cram it into a top-down NPS score, which is unlikely to yield good data.

I’d recommend the following post from Fred Reichheld on this topic: http://www.netpromoter.com/netpromoter_community/blogs/fred_reichheld/2006/12/27/the-right-number-of-questions-for-nps-surveys

I am looking for a little help with a NPS calculation query.

I have seen an organisation where they are calculating the NPS score of departments by taking an average of the NPS score of individuals in that section. Then calculating the NPS score of the business as the average of the sections.

Also to calculate a yearly score they will average the monthly scores.

But statistically I didn’t think you could take an average of an average especially with small sample sizes and in this case there are scenarios where only a handful of NPS responses are received for individuals i.e. some team members have over 200+ responses and others only 3 but there scores carry an equal weighting if the average of their scores are taken.

Thanks in advance your thoughts are appreciated