That was the title of the 2003 HBR post by Fred Reichhosted that presented the Net Promoter Score as a means to measure customer loyalty.

You are watching: The one number you need to grow

It’s a strong claim that a solitary attitudinal item deserve to portfinish agency success. And solid claims require strong evidence (or at least corroborating evidence).

In an previously post, I examined the original evidence put forth by Reichorganized and tried to find any other published evidence and also debated the findings at the event, How Harmful is the Net Promoter Score?

To develop the validity and make the case that the NPS predicts expansion, Fred Reichorganized reported that the NPS was the ideal or second-ideal predictor of expansion in 11 of 14 markets (p. 28).

The information he provided in the appendix of his 2006 book The Ultimate Question to assistance the partnership shows information from 35 companies in six industries (computer systems, life insurance, Oriental auto insurance, U.S. airlines, Internet Service Providers, and UK supermarkets). His 2003 HBR post consisted of 5 even more companies and also one extra market (rental cars) for a full of 40 suppliers and 7 sectors.

Close examicountry of the information reveals that Reichorganized used historical, not future development. He confirmed the three-year average expansion prices (1999–2002) associated through the two-year average Net Promoter Scores (2001–2002). In other words, the NPS associated through past development rates (as opposed to future expansion rates). This does develop validity (a type of conpresent validity) but not predictive validity.

To assess the predictive capacity of the NPS, I looked at the UNITED STATE airline industry in 2013 and also uncovered a strong correlation in between future growth and NPS (but only after accounting for a significant merger in the industry).

The publimelted literary works on the topic in the last 15 years isn’t terribly beneficial either. I found eight various other researches that examined the NPS’s predictive capacity (Figure 1). I was, yet, a little bit disappointed in the quality of many of the studies provided the ubiquity of the Net Promoter Score.

As Figure 1 reflects, three of the eight studies uncovered medium to strong correlations but supplied historical or existing revenue (not future). Of the five remaining researches that supplied future metrics, two were authored by a contender of Satmetrix (a feasible competitive bias) and also one was from a book through connections to Satmetix and not peer reviewed (with an agenda to promote the NPS).


Figure 1: Overview of papers researching the NPS and also development (many offered historical revenue or had methodological flaws—prefer not actually making use of the 11-suggest LTR item).

Surprisingly, 2 of the 3 research studies that looked at future metrics didn’t usage the 11-suggest Likelihood to Recommend question (Keiningham et al., 2007b; Morgan and also Rego, 2006). One examine that offered a 10-allude variation that uncovered no correlation with service development also uncovered no correlation with any type of metrics at the firm level for 3 Norwegian markets it examined (Keiningham et al., 2007a), which was an inexplicable finding provided all other studies found some correlation via metrics.

Only the study by de Haan et al. (2015) actually supplied the 11-allude Likelihood to Recommfinish item and also found the Net Promoter Score did have actually a tiny correlation with future intent (collected in a longitudinal study). It wasn’t the ideal predictor, however it did correlate with future metrics (which was comparable to the finding from the research by Keiningham et al., 2007b making use of a 5-point LTR).

I think tright here are at leastern 2 reasons for the dearth of publimelted data studying the NPS and also growth:

It’s difficult: Predicting revenue at the customer or company level calls for data from 2 points in time. Longitudinal data takes time to collect (by interpretation years in this case). It’s likewise tough to associate attitudinal data to financial performance. Companies have little bit reason to reveal their own data and also third-party firms have trouble gaining access.

Predicting Future Growth with the Initial Data

A few records I cited over mentioned the difficulty through Reichhosted utilizing historic revenue to show future development yet none I found actually looked to check out whether the publiburned NPS information predicted future development for the exact same industries. Keiningham (2007a) did use some of Reichheld’s data to display that the Amerihave the right to Consumer Satisfactivity Index was an equal or much better predictor of historic revenue, but didn’t look at future development.

So, I revisited the very information supplied to develop the NPS validity—the 1999–2002 Net Promoter Score data Reichorganized published in his 2006 book appendix and also 2003 HBR short article.

With the assist of research aides, I dug through old yearly reports, push releases, posts, and also the Net Archive to match the financial metrics collected more than 15 years earlier. It wasn’t straightforward, as many carriers combined or went out of company, and totality markets morphed (AOL anyone?). We had to piece together numbers from many type of various sources and make some presumptions (listed below).

After a number of weeks of digging we had actually excellent outcomes and also were able to discover information for the exact same six markets supplied in the 2006 book plus the one market included in the HBR short article for the years 2002–2006. Table 1 shows the sector, the metric we used, the year the NPS data was reported in Reichheld’s book, the current/historic years Reichheld provided, and also then the years we uncovered information for to predict future development.

IndustryMetricNPS DataReichorganized YearsOur Future Years
U.S. PC marketCOMPUTER Shipments2001-20011999-20022002-2005
U.S. Life Insurance marketLife premiums2001-20021999-20032002-2005
UNITED STATE airlines marketSales2001-20021999-20022002-2005
UNITED STATE Web Service ProvidersSales20021999-20022002-2005
U.S. car rental marketRevenue20021999-20022002-2005
UK supermarketsSales20031999-20032003-2006
Korean auto insuranceSales20032001-20032003-2006

Table 1: Industries used to create the predictive capability of the Net Promoter Score from The Ultimate Question and also the 2003 HBR article.


We supplied two future growth periods to assess the predictive validity of the NPS. The first are the two years immediately following the NPS information (and graphed below). For the UNITED STATE sectors this was 2002–2003; for the global industries this was 2003–2004 (which matches the years of NPS data Reichorganized used). The second consists of a longer duration of three to four years of expansion (2002–2005 for UNITED STATE markets and 2003–2006 for international). We computed Pearson correlations for each industry, then averaged the correlationships making use of the Fisher Z transdevelopment to account for the non-normality in correlationships. Finally, we converted the correlationships to R2 values to match the fit statistic reported in The Ultimate Inquiry.

Reichheld notes that they discovered the log of the readjust in NPS would boost the explanatory power (R2) of NPS but they reported only raw NPS numbers in the appendix. With just one year of NPS data, we didn’t have actually transforms in the NPS so we replicated the approach in the appendix using just the data from the single Net Promoter Scores.

Table 2 shows the results for Reichheld’s originally reported R2 values using existing or historic revenue and also our R2 worths for the succeeding 2 and also 4 years.

A bit to my surprise (provided the many type of vocal doubters and absence of publiburned data), we found evidence that the Net Promoter Score predicted development in both the subsequent two- and also four-year periods. On average we discovered the Net Promoter Scores reported by Reichorganized explained 38% of the changes in expansion for the salso sectors examined for the prompt two years (low of 8% to a high of 76%). The explanatory power lessened some when the future period raised (which is not too surpclimbing provided what deserve to change in 4 years). For the four-year period, the average explanatory power of the NPS is still 30% (low of 4% to a high of 79%).

To put these R2 worths into perspective, the SAT have the right to explain (predict) roughly 25% of first year college grades, which indicates these R2 values are impressively big.

Reichheld Historical R^Sq2-Year Future Growth R^Sq4-Year Future Growth R^Sq
UNITED STATE PC market68%27%75%
U.S. Insurance market86%39%4%
U.S. airlines market68%8%22%
UNITED STATE Net Service Providers93%20%2%
UNITED STATE car rentals 28%8%8%
UK supermarkets84%76%79%
Korean auto insurance68%48%12%
Avg R2(Fisher Transformed)76%38%30%

Table 2: R2 values of salso sectors from Reichheld’s NPS information compared to historically reported revenue and two-year and four-year development rates by market. The Fisher R to Z transdevelopment was offered to average the correlationships before converting to R2 averperiods. *Reichheld reported an R2 of 68% for Korean auto but our replication from the scatterplots produced a value of ~30%. See other notes below by market.

Below we have actually re-developed the bubble scatterplots from Reichhosted and compared that via our two-year future information. We approximated the regression lines, R2 values and also bubble dimension making use of a similar technique as explained in Keiningham et al 2007a.

PC Shipments

Historical R2 = 74%Future (2 Years): R2 = 27%

Note: Compaq was purchased by Dell so is not consisted of in future years. IBM marketed its COMPUTER industry to Lenovo in 2005 so calculation just contains development prices in between 2002–2004 rather of 2002–2005. Gateway combined with eMachines in 2004; development prices are likewise only 2002–2004 and only incorporate Gateway numbers.

US Life Insurance

Historical R2 = 86%Future (2 Years): R2 = 39%

Note: For Prudential we offered expansion prices in British pounds, but bubble dimension on the chart is established by converted number of life premiums in UNITED STATE dollars.

US Airlines

Historical R2 = 66%Future (2 Years): R2 = 8%

Note: TWA quit operations in 2001 and wasn’t included in calculation for future years. America West Airlines four-year expansion duration is between 2002–2004 as they combined with US Airmethods Group in 2005.

Net Service Providers (ISPs)

Historical R2 = 89%Future (2 Years): R2 = 20%

UK Grocery Stores

Historical R2 = 81%Future (2 Years): R2 = 76%

Note: For ASDA we offered growth rates in USD, yet the bubble dimension on the chart is figured out by converted variety of sales in British pounds.

Korean Auto Insurance

Historical R2 = 68%/30%*Future (2 Years): R2 = 48%

Note: Reichheld reports an R2 of 68% however we calculated a much reduced R2 of 30% from the very same data.

U.S. Rental Cars

Historical R2 = 28%Future (2 Years): R2 = 17%

Note: In 2003 Vanguard Group purchased National and Alamo brands and didn’t separate the revenue so they are excluded in the future analysis.


A re-examination of the original NPS data using future (quite than historic revenue growth) found:

The NPS explains prompt firm development in schosen industries. On average we discovered NPS information deserve to define 38% of the varicapacity in company growth metrics in seven sectors at the company/firm level. This is much less than fifty percent the explanatory power of historic expansion reported by Reichhosted (76%) yet still represents a considerable amount relative to various other behavioral science actions. While not as impressive, it still suggests the NPS is a leading indicator of future development prices, at leastern in some schosen sectors for some time periods at the firm level.

The NPS is still predictive of even more remote growth. The explanatory power of the NPS still continued to be at a solid 30% for a four-year future development period. This suggests that established company plans and growth fads have the right to remain in impact for years (yet not always) and also the NPS might still portend the even more far-off future (aobtain in these schosen sectors and years).

Industry changes are hard to predict via few information points. Companies merge, sectors morph, and unsupposed changes deserve to occur that influence a company’s expansion and also subsequently the predictive capacity of any measure, consisting of the NPS. This was checked out in the auto rental industry (National merged) and the PC sector (IBM sold to Lenovo) and also the airline industry (TWA was obtained after bankruptcy ). When an market has actually few information points (e.g. ISPs through just three), only the strongest relationships are detectible and little transforms in one year deserve to completely remove any type of evidence for a partnership in between NPS and also growth.

Prediction is imprecise. The NPS might be a victim of its own success through its hype leading many kind of to dismiss out on it unmuch less it’s a perfect predictor of development. (After all the headline suggested it’s the ONE number you should grow!) Making predictions is hard and also imspecific yet this analysis says the NPS does have actually reasonable predictive capacity, at leastern as high as other high-stakes procedures favor college entrance exams. It’s unmost likely always the premium measure in eexceptionally industry, provided our earlier analyses on satisfactivity yet this data again suggests it may be an enough proxy measure of future development for many markets.

Tright here is a feasible selection bias. We restricted our evaluation to the markets, carriers, and also metrics reported by Reichheld. It’s likely that these are the best illustrations of the NPS’s predictive (or post-dictive) capability and also may not be representative of all sectors. Reichorganized himself reported that the NPS wasn’t constantly the best predictor of growth (just in 11/14 industries). A future analysis will certainly look at a more comprehensive range of the salso sectors displayed right here and also examicountries at the customer level.

See more: Pin On It Works Before And After Stretch Marks, It Works Stretch Mark Cream


Below are the sources where we discovered expansion prices to match those reported in Reichhosted so you can examine our work-related and presumptions (let us recognize if you see a discrepancy).

US COMPUTER industry (All Firms)

US Life insurance market

New York Life Report 2004 State Farm State Farm Announces Financial Results For 2004

US Airlines