Eight times out of 10, companies must focus on reducing friction at the journey point; using the NPS measure is usually not the best way. There are other, better ways to estimate the reduction in friction. The Customer Effort measure, for example, (Net Easy) is better suited for this purpose.
The Relative/Competitive NPS score in publications (e.g., Loyalty Effect, Winning on Purpose), which connects NPS with financial data, is not based on a singular customer journey event. Instead, it's a double-blind study (where the company does not know who is answering the survey, and the respondent does not know which company is asking the question) done by an external agency that measures the NPS score for an entire industry with all the competition.
This is often referred to as the Relationship NPS. Measure this number and track it over time. Again, this is not benchmarked at an individual journey point but assesses the overall customer journey. This is an excellent leading indicator for the CNPS (above). You can use this to link your internal financial data (e.g., do my promoters buy more from me over time?)
At the customer journey event, your 'hero' question should be the effort question mentioned above. The NPS question serves a different purpose. It should primarily be used to pull out specific interactions that generated those ten scores. This is where you will learn what kind of exemplary employee actions creates a promoter. This is equally important learning to take away from the customer's voice. However, you can skip the NPS question if a) you are still trying to figure out your problem areas and b) you want your survey brief.