How to get more from your score
Share
Adding extra layers to the net promoter score technique of assessing levels of customer satisfaction is enabling marketers to take their insight into the boardroom.
Since its inception by business strategist Frederick Reichheld in 2003, the net promoter score (NPS) method of assessing a brand’s customer satisfaction levels has been widely adopted by marketers.
Reichheld claims the system attributes a brand’s predicted revenue growth to the responses given to just one question - how likely out of 11 would you be to recommend this company to another person?
Those who score a brand between 1 and 6 are classed as “detractors”, while those who score a brand 9 and 10 are “promoters” and those who score a brand 7 and 8 are seen as “passive”.
While Reichheld’s theory has its critics, there is no doubt NPS has proved useful for marketers. Mark Blayney Stuart, head of research at the Chartered Institute of Marketing (CIM), believes the use of NPS continues to increase because the system is so straightforward and easy to use.
He says: “So much information is available these days that one of the problems can be how you make enough sense of that data to do something with it. NPS can give a clear picture.”
Good and bad feedback moves exceptionally swiftly. You need a mechanism to ask customers why they have scored the way they have
Phil Evans, Carole Nash Bike Insurance
However, Blayney Stuart acknowledges that NPS has flaws and says marketers should treat it as a springboard to prompt more in-depth research using other methodologies.
He argues: “One inherent risk of NPS is that it turns things into a black and white picture. If you are only looking at ’promoters’ and ’detractors’ you’re only seeing a broad-brush picture and stripping out lots of qualitative information.
“The key is to use NPS as a prompt for obtaining more qualitative views to find out why people are recommending you or not. That can start to give you the insights for the direction you might want to take in the future.”
Viewpoint
Sholto Douglas-Home
Group marketing director
Hays Recruitment
When I joined Hays three years ago, I found the net promoter score (NPS) concept to be a one-dimensional tool.
However, about a year ago we introduced some extra dimensions. We allow respondents to give free text feedback. With these additions NPS becomes incredibly valuable as a richer tool to measure service and satisfaction.
It’s fantastic to have promoters, but we want to understand just how satisfied our customers really are. We have eight stages of the recruitment process from the initial encounter through to job placement so we ask candidates and clients to rate us out of 11 for those key stages to help us build a picture of genuine satisfaction.
The managing director or regional director reviews every free text comment. All comments from those who score us 9 or 10 are passed on to the individual consultant so they can see the benefits of their professionalism. We also draw out the positive comments and anonymously showcase them on our intranet. This is a way of demonstrating how much good service is appreciated.
The way NPS has been measured has been consistent over the past five years and the operational side of the business has really bought into it. I’m on the management board so every quarter I do a presentation that is shared with company directors. By making NPS broader, the sales and management teams here see the benefits of this three-dimensional approach.
Marketers who use NPS must then take the initiative to bolster their quantitative results with qualitative methodologies to delve deeper into why a respondent has given a certain score, as well as make more use of respondents who are identified as promoters.
One of the latest brands to join the NPS party is insurance and investment firm Standard Life, which is running a pilot among its network of intermediary financial advisers. Standard Life head of group insight Carol McCreadie explains that finding out who is favourable towards the firm, and why, is essential to the business.
She says: “Recommendation is really important within the financial services sector, particularly with financial advisers, who have a lot of clients coming through their doors and are looking for advice.
“We want to understand how often they are choosing Standard Life over a competitor and why. We will look at individual accounts to find out how much business clients are actually placing with us, and monitor this as we work with them to see how our score can increase and how it impacts on our business.”
Like many brands, Standard Life will continue conducting its brand tracker activity alongside NPS, which will be reviewed by senior managers.
Sholto Douglas-Home, group head of marketing at Hays Recruitment, explains that NPS gives marketers a numerical, results-based language that they can use to speak to their company board (see Viewpoint, above).
The key is to use NPS as a prompt for obtaining more qualitative views
Mark Blayney Stuart, CIM
While some brands are still experimenting with the scoring system, others are incorporating it with a social media strategy. Aviva, for example, uses software from provider ServiceTick that prompts respondents who score the brand a 9 or 10 to post a recommendation to their Facebook or Twitter pages.
Aviva online customer experience manager Liz Habgood explains that the survey automatically generates some text for the relevant respondents to edit if they wish, and post with one click to a social network site. She says: “It can be risky if it looks contrived, but we have tried to word it in a way that sounds fairly natural. People can add their own comments to reduce the chance of it looking too forced.”
Aviva is using social media in anticipation that these positive sentiments could be streamed in search results in the future, she explains.
Social media isn’t currently used in web search rankings, but people do use social media to gauge opinions of brands. And it could be that Google will eventually use social media feeds in search.”
While incorporating social media adds value to Aviva’s use of NPS, Habgood acknowledges that NPS is not all-encompassing. To get around this, the company also uses a brand tracker, focus groups and analyses the transcripts of its online customer service web chats.
Specialist motorcycle insurance broker Carole Nash also uses ServiceTick technology, but has expanded the use of NPS by allowing customers to add verbal comments after completing a telephone survey.
Q+A: Brand in the spotlight
Philips
Suhail Khan
Head of customer experience and market driven information
Marketing Week (MW): How long have you been using net promoter scores (NPS) and how has it impacted on your business?
Suhail Khan (SK): We started using NPS about four years ago. We wanted to increase our market share and understand where we stood compared with our competition. We also wanted to obtain insights in terms of how we go to market with our products. Our pilot led to some great insight, so we have now activated it through a large portion of our company. We try to get a proxy of how our business is behaving in all markets and how much the promoters are working for us in our markets.
MW: How can you act on the insight delivered by your NPS programme?
SK: NPS is a fantastic way to get a quick insight to what’s happening in a market at a given time, for a given product in a given sector. That’s step one. Step two is root cause analysis - that is to understand what causes a particular response.
NPS is also a great indicator of market behaviour. When NPS goes up we can anticipate that revenues will follow. When NPS goes down, we can warn the rest of the enterprise to take action because of this feedback. And our management has formulated NPS-based deliverables as far as growth and providing the right customer experience.
MW: How are you using the views of your “promoters”?
SK: We launched the Promoter Reference Programme last year to capture the voice of our customers through video, audio and case studies. We feed that back into our sales organisation so we can use the voice of the customer in our pitches. We have stopped talking about NPS as a score because if you think of it as a number that’s all it will be without an understanding of what it means.
Passionate enthusiasts
Phil Evans, head of operations at Carole Nash, explains that bikers are enthusiasts so it’s important to keep them onside. He says: “There are just over 1 million bikers in the UK, and good and bad feedback moves exceptionally swiftly among groups of people who are exceptionally passionate about what they do. You need a mechanism to ask customers why they have scored the way they have.”
Carole Nash’s NPS activity also incorporates internal advocacy, through a monthly survey asking employees how likely they would be to recommend Carole Nash as an employer to friends and family.
“A high internal NPS will actually deliver on and attract the external NPS,” claims Evans. “If you have people who are advocates of the organisation, that will come across on a customer service call.”
Harnessing the word of mouth power of promoters is high on the agenda of consumer electronics giant Philips. Suhail Khan, head of customer experience and market driven innovation, explains how the launch of a Promoter Reference Programme last year allows Philips to pinpoint people who score the brand 9 or 10. Philips not only uses their positive comments but also investigates the motivations behind their score and how it extends to their real-life behaviour.
Blayney Stuart at the CIM says such activity illustrates how NPS can be the starting point for brands to engage further with their advocates. He is clear that NPS is a tool that can help marketers develop a broad understanding of how the company is rated. But he warns that marketers must assess what role NPS has to play in understanding their brand’s performance, and if it really is the best measure for what they want to achieve.
Criticism of NPS
The American Customer Satisfaction Index (ACSI) was developed by Claes Fornell in 1992 and has been used to predict sales and profit but in 2003 Frederick Reichheld claimed that NPS could be used as a single measure of word of mouth to provide a better prediction of brand performance.
However, Kingston University professor of strategy, marketing and entrepreneurship Robert East argues that NPS does not measure negative word of mouth effectively because ex-customers and non-customers are never sampled, and these are the people who would express negative sentiments about brands.
East, whose paper on the topic will be published in a forthcoming issue of The International Market Research Journal, adds that the ACSI is also “insensitive to dissatisfaction”.
He says measuring the effect of word of mouth through its volume and impact on purchase probability expressed by users of the category would be more effective rather than just users of a specific brand.
Meanwhile, senior vice-president of Ipsos Loyalty Tim Keiningham claims that NPS has been “vastly oversold to companies”. In 2007, Keiningham and a team of academics took NPS measures from 15,000 consumers and 21 firms, and compared them with actual revenue growth. They concluded there was no real link between the two. Keiningham says business leaders should not adopt NPS if their only reason is that they have been sold on the superiority of the metric.
Other criticisms of NPS include that it uses an over-simplified scale and doesn’t take into account cultural relativity. For example, NPS dictates that people who score a brand 9 or 10 are “promoters” but in some countries those who score 7 or 8 might also be very satisfied.
Readers' comments (7)
Anonymous | Thu, 17 Feb 2011 2:11 pm
The calculation of NPS is one of the main problems with it. By subtracting one group from another, the error in the measure is actually multiplied. On a related note, this measure offers less statistical power than a properly-constructed index like the ACSI. Anyone who has taken a graduate course in research methods should know enough about building reliable and sensitive measures to see right though NPS as an oversimplification of a complex concept that is also fraught with error.
Unsuitable or offensive? Report this comment
Adam Whittaker | Thu, 17 Feb 2011 2:18 pm
In theory NPS is great. In practice (at least in my experience) it can be misleading because different customers react differently and provide the answer they want you to hear and not necessarily their true answer. Maybe by giving a low score the company will give them something to make them happier. Or maybe by giving a high score the same will happen. Human's are devious little blighters!!
Unsuitable or offensive? Report this comment
Adam Dorrell | Thu, 17 Feb 2011 2:58 pm
Having implemented dozens of Net Promoter Score surveys for major brands, I can share a few insights for marketers.
1. Measuring continuously rather than using discrete annual surveys is more successful for NPS. Tracking the NPS of 100s of touchpoints or transactions a day is a really good way of monitoring process
2. It's less useful as a market research tool. It won't replace a brand survey, and I suggest you don't mix up any marketing questions however powerful the temptation. Instead, pre-segment your list rigorously - then ask one or two questions. Then split results by segment for your marketing
needs.
3. The output is most useful for "Fire-Fighting" and "Fire Prevention". If a customer asks for more help on the survey, think of it as a fire alarm. Count the number of fires, and you have an idea of where to prioritize your fire prevention efforts. Typically this will be outside marketing, so setting up team that can work with NPS is vital. This is hard. Do not underestimate the effort to get people in your company to change the way they work. However, this is the real benefit of NPS.
4. Finally, NPS is becoming a widely accepted metric, and one that is easy to implement and explain at all levels of the organisation. There are not many standard metrics in the world of marketing! Criticism is of course valid, but we've not found a good substitute yet.
Adam Dorrell, CustomerGauge
Unsuitable or offensive? Report this comment
Jan Van den Bergh | Thu, 17 Feb 2011 7:47 pm
1. NPS is indeed a springboard for more in depth research. Nobody has been telling it was useless to add that qualitative phase. The Holaba-platform in China which uses a variant of the NPS-measurement gives all brands the opportunity to ask whatever question to whatever registered user.
2. NPS indeed focuses on actual users since it started as a loyalty-measurement-tool. Holaba China broadens the scope and measures ex- and non- or not yet-consumers. They are all part of the word of mouth pool in which all opinions fight to win and not only those of the users of a specific brand.
3. Measuring purchase probability is an interesting tool: you measure the result of the ongoing word of mouth war. However by asking Reichheld’s One Question in the Holaba-context we measure the recommendation power of brands and consumers (we literally identify the recommenders) and give brands the possibility to gather further insights and instantly communicate with all segments.
4. Of course those who give a 7 or 8 might very satisfied, but only if they start to voice their opinion, they become a powerful force. The cultural relativity is no issue since you compare brands within 1 sector. In some sectors a high score is needed to be the winner. In other sector a moderate score can make a winner.
5. NPS doesn’t disagree with the claim that “recommending a brand” is the result of complex process. But then again you can be satisfied, you can be loyal but only when you voice your opinion, your voice counts. That’s what the One Questions tries to grasp. With a simple question that unveils the sum of multiple complex decisions.
Jan Van den Bergh
Chairman
www.Holaba.com.cn
Unsuitable or offensive? Report this comment
Alain Thys | Sun, 20 Feb 2011 2:41 pm
In my opinion, the biggest argument for Net Promoter thinking lies outside the realm of scoring and statistical accuracy.
Every time I've seen a CEO, board or senior executive team look at the numbers, and expecially the customer verbatims explaining them, the room lights up.
If you combine these insights with some hard financials on the value of promoters vs. detractors, executives finally see a business case for customer happiness, which prompts them to ACTION.
I have yet to see this type of executive reaction to "more accurate" methods (which are indeed more accurate).
But until I do, I'm a pragmatist. If the 80% solution gets the executive team to listen to and act on the customer's voice, I prefer it over the 100% solution which makes their eyes glaze over and lulls them into complacency.
Once you get the business in motion, you can start refining the observations and introduce more accurate, multivariate, techniques.
Which is why, after 5 years of customer-centricity programmes in Europe, Asia and North-America, I'm still banking on NPS.
Alain Thys
Managing Partner
Futurelab
www.futurelab.net
Unsuitable or offensive? Report this comment
Michael Lowenwstein, Ph.D., CMC; Executive Vice Pr | Sun, 20 Feb 2011 4:38 pm
Objective critics of the one-number score, hoping to guide users, detractors, and those considering the technique, have made statements (in a LinkedIn group dialogue) like:
- “ …it is simply an indicator. It must be accompanied with a lot of additional information to really be helpful. You need to know why people will or will not recommend”
- “…the approach is incomplete at best and potentially misleading at worst. Rather than simply accept the (technique) as a panacea for customer retention and business growth, executives must critically and objectively evaluate the merits and risks associated with this measure.”
- “…it is a red herring without backup questioning. What good is having a score if you can’t explain what the score means?”
- “As a measure, it is cycloptic as it does not account for those who say they will recommend but will cancel the relationship when faced with a better, cheaper offer.”
- “It doesn’t work because it obsesses on the dynamics between two different groups while ignoring their velocities over time. I have seen a company lose 40% of sales in the same time period that the (one-number recommendation score) increased by 25%. Quite simply, the company lost a lot more detractors and neutrals than promoters, causing the equation to slope upward as sales caved.”
So, marketers really have to understand what brands and products customers would consider within a category (known as the evoked, or consideration, set); what they're currently using; what their level of favorability is; and what their likelihood (and evidence) of saying positive and negative things about the product is. And if they then elect to build recommendation into that kind of construct, they will have a contemporary, real-world framework that's more actionable – customer advocacy.
Here’s the bottom line. It would be attractive to have a single metric, but its analytical limitations can be dangerous. Marketers want to know more about their customers’ perceptions and actions. They need to understand why levels of loyalty behavior are occurring, not just what is occurring or what the indicators are, on as sub-segmented and individual a basis as possible. Again, referral and recommendation are important outcomes of loyalty but are not just the end goal, or Holy Grail, in and of itself. Magic and potions may work for Harry Potter; however, marketers must understand customer behavior in the real world; and this must incorporate the leveraging effect of brand impression and positive and negative word-of-mouth.
Though many market researchers have been dismissive of the value of the recommendation question, expressed as net positive or net negative, they admit that it has created attention and acceptance; and, it has continued to endure due to the vacuum created by the absence of a better way of understanding customer behavior. Now, there is a better, more real-world, and more actionable, consistent, and monetizing way. It is customer advocacy.
Unsuitable or offensive? Report this comment
r.murray | Mon, 21 Feb 2011 11:53 am
Unless you contact your full range of business contacts predicting customer satisfaction is innacurate as upset customers will not even be bothered to contact you back.
Often the customers most upset are not interested you for survey work.
Too many marketing companies surveying on your behalf give you the story you want to hear and don't give all the negatives.
I have always made the point of chasing lost customers plus the ones that get in touch and then do not trade with me, only when you survey completely can you be sure.
An old principle I was taught was to work on "one in ten".
For every one customer that tells you your faults there are nine others that are upset who never tell you!
Unsuitable or offensive? Report this comment