There's already an excellent answer on calculating MoEs and confidence interval to compare NPS results from two samples.
However I would like to find a p-value (or at least a z-score) as a result of the comparison. Is it possible at all? My stats knowledge is rather limited.
