Re-Evaluating The Net Promoter Score

December 2007from being uniformly loyal, some were far more
Fred Reichheld, an American loyalty guru, hasprone to being tempted by competitor offers
argued that the days of customer satisfactionthan others. The key difference was the
surveys are over (Reichheld, 2003). Instead wecustomer's level of involvement with cell phones.
should only be measuring customer loyalty.I found that amongst the Promoters who were
According to Reichheld all that is necessary is apsychologically involved with cell phones,
single question, which he labels 'The Ultimatecompetitors offering 20% discounts would create
Question'.a smaller increase in the switch rate than amongst
The Ultimate Question is simply "How likely arethose Promoters who were not involved.
you to recommend company X to friends andIntriguingly those who were involved, although
colleagues?" Respondents are asked to rate thebeing an asset in that they were more resistant,
likelihood of recommending on a scale from 0,spent less on their cell phones every month than
meaning extremely unlikely, to 10 meaning verythose who were involved.
likely. Promoters are those who give a rating of 9In essence, 'Promoters' are not equally loyal.
or 10. Promoters are important not only becauseRelationship With Growth Not Clear
they are more likely to promote your brand toOne of the foundational claims which undergirds
friends and family, they also tend to spend morethe value of the Net Promoter Score, is the claim
than non-Promoters (Reichheld, 2003).that the Net Promoter Score relates to corporate
According to Reichheld you need to take it a stepgrowth. However this is questionable. In a study
further if you want to relate the rating to futureof 80 companies over a 7 year period, the
corporate growth, by calculating a Net Promoterresearchers Morgan & Rego (2006) found
Score. The Net Promoter Score is simply thethat the Net Promoter Score was not predictive
percentage of respondents who give your brandof company growth rates. In fact, customer
a 9 or 10 rating on a 0 to 10 recommendationsatisfaction outperformed the Net Promoter
scale, minus the percentage who give you a ratingScore as a predictor.
of 0 to 6 on the scale (Reichheld, 2003).Weighing-up the Positives and Negatives of the
PopularityNet Promoter Score
On the back of Reichheld's finding that the NetReichheld has brought up some valid points about
Promoter Score correlates with future corporatesatisfaction and loyalty measurement. Simplicity is
growth across a number of sectors (Satmetrix,not only appealing to high-level decision makers, it
2004), the measure has gained broad supportalso appeals to customers who don't want to
from executives across a variety of industries.answer long questionnaires. The measure also
Some, such as GE's CEO Jef Immelt, have takenlends itself to standardization, to the point where
the unusual step of publicly praising the approachit may be a useful number to include on the
(General Electric, 2005).balance sheet.
Besides the claimed relationship with corporateThat is one side of the story. The other is that
growth and customer behaviour, an additionalthe 'Ultimate Question' is far from being the most
attraction is simplicity, something very appealingaccurate measure of loyalty, and so will more
to time pressured executives.often mislead than help.
Along the way Reichheld has popularized variousIt can also prove frustrating when it comes time
ideas in tandem with the loyalty measure - suchto understand how to resolve a loyalty problem.
as his controversial claim that retaining 5% morePerhaps the highest-level decision makers are not
customers equals 100% more profits. Ofoverly concerned about tactical level problem
particular concern to those who have investedsolving, however someone in the company will
heavily in customer satisfaction measurement, ishave to be; and a single question won't give the
the claim that customer satisfaction measurementanswers needed.
is redundant.Another issue is the delay before loyalty changes.
Reichheld pitches satisfaction measurementIf executives wait until loyalty has been affected,
against loyalty measurement and proclaims theit is sometimes too late. Satisfaction with service
death of satisfaction measurement. He claims thatquality will often drop long before it affects
customer satisfaction measures are overlyloyalty.
complicated, a heresy to some in the boardroom,So Where to Now?
and that they also fail to explain corporate growthReichheld has confused satisfaction with loyalty.
differences in a sector.Satisfaction and loyalty are two different issues.
He also laments the long windedness of theWhile you would naturally expect loyalty
questionnaires typically employed to measuremeasures to relate more closely to customer
customer satisfaction.behaviour and future growth than satisfaction and
The Ultimate Question?service quality perceptions, claiming that you
There is nothing terribly wrong with the Ultimateshould drop these measures in favour of loyalty
Question - other than that it is far from beingmeasurement makes no sense at all.
'ultimate'. Reichheld claims that of the 20 differentCustomer satisfaction with the service quality
measures he tested, it was the best predictor ofdelivered, is only one amongst a number of
customer behavior and future corporate growthloyalty drivers, but that doesn't mean it should be
(Reichheld, 2003).ignored.
Not the Best Predictor of Customer RetentionMeasuring loyalty alone, may tell you how many
You don't need a degree in statistics to figure outloyal customers there are, but it doesn't tell you
that there is something intuitively wrong with thehow to fix the problem if there is one.
claim, that whether a customer says he willIt is still necessary to measure customer
recommend your company or not is the bestsatisfaction, along with service quality perceptions
predictor of whether he will stay with yourin order to flag issues and tackle the root causes
company.of loyalty failures. Throwing out satisfaction in
Differences in the tendency to recommend don'tfavor of loyalty is like deciding that you only need
automatically mean there will be differences into report net profits on an income statement and
loyalty behavior - whether retention or increasedcan dispense with all the rest of it.
spend. Results from an analysis I conducted usingLastly, looking beyond the hype, it is clear that
data collected by Ask Afrika, of a number ofwillingness to recommend is a sub-optimal loyalty
different markets ranging from short termmeasure. There are far more accurate loyalty
insurance to cars made this clear. Using anmeasures in existence which have benefited from
approach which has been established as a reliableyears of development by professional marketing
predictor of the percentage of customersresearchers. They offer better accuracy in
switching away from a company, I compared thepredicting loyalty behaviors, and are typically
results to the Net Promoter Score (this approachcoupled with diagnostics which aid in gaining an
is discussed in a related article "The One Numberunderstanding of what creates loyalty.
You Need to Measure Loyalty").Copyright reserved (2007). The author gives
Clearly the Net Promoter Score does not relatepermission for the article to be re-published,
to the percentage of customers switching awayhowever the article may not be altered or
from each institution.shortened and must be re-published in its entirety.
A survey of 8000 customers in the banking, retailReferences:
and ISP industries, conducted by Keiningham et al.General Electric, (2005), "GE annual outlook
(2007) lends support to this. Keiningham et al.meeting - final", Fair Disclosure Wire (Quarterly
(2007) asked customers the likelihood ofEarnings Reports), Dec 13.
recommending question along with a rival set ofKeiningham, T.L., Cooil, B., Aksoy, L., Andreassen,
questions, and then followed up with questionsT.W., Weiner, J.W., (2007)"The value of different
about retention a year later. The results showedcustomer satisfaction and loyalty metrics in
that the Ultimate Question barely improves on apredicting customer retention, recommendation,
conventional satisfaction measure as predictor ofand share-of-wallet", Managing Service Quality Vol.
retention, and was outperformed by a simple17 No. 4, 2007 pp. 361-384.
repeat purchase intention question.Morgan, N. & Rego, L. (2006) "The value of
Doesn't Identify Those Who Are Truly Loyaldifferent customer satisfaction and loyalty metrics
Reichheld (2003) states that some clients buy outin predicting business performance", Marketing
of habit, which means that retentionScience, September/ October.
measurement (i.e. did they stay or did they go) isReichheld, FF., (2003), The one number you need
not a good measure of true loyalty. He suggeststo grow, Harvard Business Review, December,
that the Net Promoter approach is able to identifyVol. 81, Issue 12.
true loyalty. This suggests that Net PromotersSatmetrix, (2004), The power behind a single
should be more resistant to competing offersnumber: Growing your business with Net
than those who just buy out of 'habit'.Promoter, Satmetrix Systems.
However, using cell phone data collected by Ask*Note: Net Promoter is a registered trademark of
Afrika, I found hidden differences amongstSatmetrix Systems, Inc., Bain & Company
consumers who were identified as 'Promoters'. Farand Fred Reichheld.