Net Promoter Score (NPS)

WHAT IS NPS?

NPS is a  widely used measure of firm performance, often regarded as the most important measure to include in any customer survey. Specifically, NPS measures the concept of customer loyalty, which is likely to impact firm performance via word-of-mouth. Harvard scholar Frederick F. Reichheld depicted this relationship with the following figures.

The above figures show the positive correlation between firm growth and firm NPS. In other words, firms with greater net promoter scores (X-axis) experience higher rates of growth (Y-axis).

HOW IS NPS MEASURED?

NPS asks hotel guests “How likely are you to recommend [Hotel A] to a friend or colleague?”. Responses to this question are measured on an  11-point likert scale, ranging from 10 to 0, where 10 is labelled “Extremely Likely” and 0 “Extremely Unlikely”.

HOW IS NPS CALCULATED?

Guests that answer the above question are categorized based on their responses. Those who respond 10 or 9 are labelled “Promoter”, 8 or 7 “Neutral or Passive”, and 6 or less “Detractor. The score, NPS, is the percentage of “Promoters” minus the percentage of “Detractors”.

HOW IS NPS USED?

Due to the observed positive relationship between NPS and company financial performance, NPS is often used as an indicator of overall firm performance. Some firms take this a step further by using NPS as a predictor of future financial performance. At Guestfolio, we recommend hoteliers use NPS as a benchmark of firm performance. Specifically, as an:

  • Internal benchmark for individual hotels
  • External benchmark for comparison with sister hotels (comparable hotels within your hotel group)
  • External benchmark for comparison with peer hotels (comparable hotels that are not owned by your hotel group)

As an internal benchmark, scores can be compared over time, to indicate how an individual hotel is performing compared to itself. If extraneous variables are controlled, managers can see how their operational changes are impacting guests.

As an external benchmark, scores between hotels can be compared at one point in time and over time. This provides an indication of relative performance among sister and/or peer hotels, in other words, compared to the market.

ONCE WE HAVE THE DATA, WHAT DO WE DO WITH NPS?

Analysis of NPS provides valuable insights that can be actioned by management. Here are a few examples:

  1. Comparing within groups of hotels allows lower performing hotels to learn from higher performing sister hotels.
  2. Drilling-down into scores by demographic allows managers to better understand and tailor communications and operations to suit the needs of certain guest segments.
  3. If the hotel guest survey has been designed well, relating scores to other question responses uncovers root causes that can be communicated back to employees to improve the guest experience.
  4. Similarly, the relationships between hotel department performance and NPS can be examined to understand which guest touch points have the greatest impact on the guest experience. This sort of analysis provides management with a basis for resource allocation and ROI prediction.

Notes:

Some common questions might be around the scale direction and the labels at each end of the scale. Firstly, in some cases there is left-bias when measuring social science concepts with scales. However, this is either very minor or not statistically significant. NPS scales exist in both directions. We use from highest to lowest because it’s consistent with the rest of our surveys and has a negligible impact. Further, the way NPS is calculated (top 2 box), scale direction is even less likely to have any statistically significant impact. 

Second, labels at each end are commonly either 10 = “extremely likely” and 0 = “extremely unlikely”, or 10 = “extremely likely” and 0 = “not at all likely”. Currently, no research testing any difference has been found. This is likely because the potential impact on respondents is just too negligible. We use “extremely likely” and “extremely unlikely” because it is simple and consistent for the respondent.