All the evidence would seem to indicate NPS is too volaIle to be used as central ROI for customer experience ‐ links to retention, growth and profit are more appropriate and compelling.

Human behaviour dictates that if something is difficult, people will find an alternative route or a workaround.

NPS reflects the symptom – effort closer to the cause. The consensus of opinion is that effort should be part of the customer experience narrative rather than adopted as a measure.

  • Temperature gauge is a good analogy, problem when too hot or cold
  • Use with a mix  of complementary measures
  • Better at judging effect on specific touch points
  • Users are conscious that it is historical by nature, for example it would be difficult for NPS to guide Blockbuster to an online model
  • Do we always want recommendation? Some customers cost us money or are prone to take advantage
  • Attention is needed on emphasis of survey, many Costa customers use the service as somewhere to meet, the coffee may be irrelevant
  • Factor normalisation (see below)


  • Human relationships cannot continuously improve
  • National cultures will guide a set of normal emotional responses
  • UK culture may well provide a negative NPS as normal response
  • We should expect a customer relationship to do the same
  • Over time a positive and negative reaction will normalise and be reflected in the score
  • NormalisaIon is often seen as fluctuaIon in the NPS score
  • Is the new norm the key indicator?  If so, how can we measure it?


Continually rising scores

Continuous improvement in NPS can be achieved from:

  • A low starting point
  • Scores being manipulated by rewarding customers for high score
  • Customers only being surveyed at wow or good complaint resolution point

None of these scenarios are an indication that a new and improved normal reaction has been established.

NPS is better if…

  • NPS starting point is low, is less effecIve at guiding change with higher scores
  • Survey is close to customer interaction
  • A survey is also sent after customer emotion has been rationalised
  • Used in conjunction with partners and others in customer federation (reference reflections on federation 2009)
  • Perhaps best as a guide to the effecIveness of current actions and as a guide for front line staff
  • Scale is good and bad: More replies = less direct linkage;  Less replies = more easily linked
  • Not as effecItie in guiding strategy as driven by a view of historical data
  • It is recognised that NPS will – and should always be  -­  a moving figure, a good indicator of movement
  • Each product and service will have a natural ceiling