Why is death of NPS inevitable? Power of Predictive Analytics and how companies are using it?
Since its introduction by Fred Reichheld from Bain & Company, in the 2003 Harvard Business Review article, Net Promoter Score or NPS has been one of most popular and widely accepted marketing metric across the world. The period from 2003 to 2013 was even dubbed as the “NPS Decade” by a score of professionals. Probably no other marketing concept made as much impact to putting customer satisfaction at the core of corporate strategies as NPS.
Even though, it is still the reigning metric as far as marketing insight studies and customer satisfaction surveys go, there has been an increasing concern over the relevance and accuracy of NPS as a metric in today’s times. There is an increasing debate over whether to continue using NPS or to look for something else. There are questions being raised about its effectiveness and marketers seem more and more divided when it comes to using NPS.
So what is it about NPS that has gone wrong? Or were there concerns and limitations from the beginning and it’s becoming more evident now? Or given the technology advancement, is it simply time to retire NPS? Let’s take a tour through NPS and its place in the world of marketing while looking at the future of customer satisfaction metrics.
The game changer of the new century : NPS
NPS stands for Net Promoter System which is a way to measure the interaction of businesses with their customers. At the end of the process, businesses get a net score which indicates the likelihood of customers recommending (or not) a business within their immediate networks. Hence, NPS is also often referred to as Net Promoter Score.
The power of NPS lies in its simplicity. There is only one question that customers need to answer:
How likely are you to recommend Company X [or Product X] to a friend or colleague?
Customers are expected to respond on a 11-point scale from 0 to 10, 0 being the lowest score (completely unlikely) while as 10 being the highest score (highly likely).
What really made NPS really tick was its superiority over everything else that was available to marketers to measure customer satisfaction powered with its extraordinary simplicity. The traditionally available methods of long, boring, tiresome and too vague customer surveys were not making anyone happy, be it the surveyors or the respondents.
NPS came around as the perfect one-question solution in sync with new generation of both marketers and consumers which needed everything fast. And the “Generation Now” lapped it up with aplomb and quite rightly so. NPS essentially answered the favourite question of the new century in business context : wassup?
NPS gave the required sharpness to customer metrics. In the wake of the times that it originated in and became immensely popular, it enabled marketers to quantify the economic answers to some valuable customer loyalty questions which could provide a direct and more accurate correlation as against the metrics available earlier.
Marketers could now answer questions like what is the value of customer loyalty, how does it impact business performance and growth, what is the value of increasing NPS score, where does the business stand for customer loyalty against the competition etc.
The changing times: need for something “beyond NPS”
The single question of NPS has been hailed as the “ultimate question” to be asked for gauging customer surveys since the introduction of NPS. And in that simplicity and a complete awe for NPS is where it is falling short to deliver and answer real marketing problems for today’s marketer.
NPS has become the irony of marketing analytics studies.
Image source : Openview
There have been concerns about NPS right from the beginning. However, in the new marketing world where marketers are armed with customer data and cutting-edge analytics tools to dig into this data to gain valuable insights into business problems, NPS alone no longer suffice.
NPS was a great tool and still continues to be so to answer some very important basic questions but there is a need to look for something beyond NPS, primarily to make the right use of where NPS begins.
Modern marketing world seems to be in agreement that the death of NPS is inevitable. Both the NPS proponents and opponents agree that NPS will no longer be enough to answer customer loyalty questions for businesses. And the simple reason for that is that times have changed significantly since NPS was introduced.
Technology has ushered into new era and keeps doing that. The customer behavior has changed drastically over the years. A new generation has taken over. And most important than above all the ability of businesses to collect, mine, analyze data and to gain insights from it has grown extraordinarily since 2003.
One could argue that nothing is wrong except that NPS is simply becoming obsolete, is slowly phasing out and dying a slow yet definitive death. This is true to a large extent. However, there are some other reasons emanating from the structure of NPS itself which have worked against it and kept marketers on a lookout for something beyond it.
Let’s have a deeper look into what has led NPS to it’s imminent death, apart from the natural time-bound obsoleteness:
- It’s not the “ultimate question” and the answer lies in “Why”
NPS, on its own, doesn’t provide any new insight. Merely knowing what is your NPS, will do nothing for you. The prime reason is that NPS merely indicates the probability of positive (or negative) recommendation expressed by the customers. NPS fails to answer the “Whys” behind it.
Why customers have provided a certain rating?
Why there is the difference in NPS scores for measurements of different periods, e.g. if a company scored 32 last year and has scored 36 this year, why this has happened?
Why do customer think that they would want to (or not to) recommend the business?NPS is not the “ultimate question” by any means. If anything, it just provides a direction from where marketers need to devise a solid follow up strategy. Relying solely on NPS wouldn’t serve any real purpose for businesses and will remain like just another meaningless metrics.
- NPS (?) of the non respondents?
The biggest limitation of NPS is its complete dependency on responses from customers. Studies suggest that even in the best of cases only 20% customers respond, leaving a giant blind spot behind. This might be a massive opportunity for designing actionable marketing strategies but NPS leaves it completely out of the equation.Won’t it be great to have a technology that can measure every customer’s inclination and also take respondents dependency away?
- It is way too linear and ignorant of important factors
For its own convenience and in the name of simplicity, NPS categorizes respondents in three categories, Detractors, Passives and Promoters. While defining the underlying assumptions are too blasphemous to make peace with. The terminology is not even representative of how we’re trying to understand customers. Anybody who is not considering to recommend you for whatever reason, doesn’t really become a detractor, similarly not everybody scoring a 9 or 10 will go on becoming your promoter. It’s way too much generalization and would mislead the whole point of understanding customer understanding.NPS is also culturally and diversity insensitive. In some cultures, scoring a 7 or an 8 is considered paramount performance, it is equivalent to a score of 9 or 10 elsewhere. Not taking such detailing into account and stuffing everybody in three large stuffy buckets doesn’t bring brands closer to customer, on the contrary, it drives them as far away as possible. Businesses need to be mindful of segment differences be it cultural, regional or even perceptional.
- It’s not actionable and relies completely on the respondent’s mercy
Since NPS doesn’t provide anything apart from a recommendation likeability by customers, it is at best a “loyalty metric” or an “outcome metric”. If not followed up properly with other actionable strategy, it doesn’t provide anything to the marketers to work with. Unless one digs deeper and knows whats, whys and hows, it’s not of much use.
NPS completely depends upon the respondent’s action. Inherent bias kicks in when people are asked to respond to something. Some people might choose to become diplomatic while scoring or turn opposite and become unreasonably scathing. Humans by themselves are poor predictors of their own future behaviors. It makes the NPS responses week and way too subjective.
While we can all agree for the need to have measures that are easily understood and used by managers, that is completely irrelevant to the issue at hand. Regardless of whether or not one chooses to believe in Net Promoter, we all must insist that the evidence used to support the metric be unbiased…[otherwise] there is no reason to believe anything we say.
Why ask when you can predict?
The modern marketers are increasingly turning towards more scientific and inclusive approaches like Predictive Analytics to find the answers about brand-customer relationships. The comprehensive data collection methods, scientific and precise processing of this data, strong and mathematical data and result correlation, unbiased pattern spotting is empowering the marketers to take better decisions armored with deep customer understanding. Afterall, patterns are better indicators of future behavior than individual data points. As a result, more and more marketers are turning towards methods like Predictive Analytics.
Predictive analytics is the branch of the advanced analytics which is used to make predictions about unknown future events. Predictive analytics uses many techniques from data mining, statistics, modeling, machine learning, and artificial intelligence to analyze current data to make predictions about future.
To put it simply, Predictive Analytics attempts to answer all the questions which remained unanswered by the limiting and fast becoming outdated methods like NPS. The strength of Predictive Analytics lies in the strong correlation between the historic digital footprints left behind by customers and the emerging patterns for likely future behavior. Additionally, since data is automatically mined across various systems and networks as opposed to relying on the customer responses, the friction and inherent biases of the data collection processes are reduced to a great extent.
Image Source : http://www.wikiwand.com/en/Prescriptive_analytics
As the image above showcases the important questions answered by Predictive Analytics form the inputs for robust decision making for businesses eventually leading to the desired effect. Predictive Analytics leading to Prescriptive Analytics solutions demonstrate the actionability of this approach.
How Promoto is using Predictive Analytics to shape the future of customer-brand relationships?
Promoto uses three key sources of data:
- Internal Systems such as CRMs, Helpdesks, Social interactions, Feedbacks, Surveys etc.
- Social Systems, we mine the social data and get to the individual-level details of their interactions with your brand
- Human Touch points data. We enrich the data based on frequent feedback from various human touch points.
All of this data is then processed by a unique custom Predictive Model Algorithm.
An uncomplicated Promoter Score and Influence Scores is calculated for every customer. Enabling you to rank customers and then prioritize relationships in ways that were never possible before.
In simple words, Promoto is using its unique data models and algorithms to crunch immense amount of customer data left behind in the form of digital footprints across the web to predict who is showing the signs of becoming your promoter. This goes far beyond the limitations of NPS. This is a powerful actionable information for any marketer as methods like this enable brands to convert customers into promoters.
Isn’t that like a marketer’s nirvana?