In 2003, customer loyalty consultant and author Fred Reichheld and Satmetrix introduced Net Promoter, aka NPS (for Net Promoter Score), through a widely-read Harvard Business Reviewarticle. Net Promoter is based on a single question, the likelihood individuals would be to refer a friend to a particular brand or company. On an eleven point scale of recommendation (0 to 10), it subtracts the percentage of low scores (known as Detractors) from high scores (known as Promoters) to produce a single number value. Reichheld has said that “it would be 100% accurate in determining revenue growth”and that NPS was “the single most reliable indicator of a company’s ability to grow.”NPS has proven attractive, especially to executives.
One of the chief values of the one-number recommendation metric, and a benefit upon which many can agree (despite all of the ongoing controversy surrounding it), is the attention and focus it has brought to having a key customer experience and relationship performance indicator. This particular indicator is purported to be stronger and more actionable than long-utilized metrics like customer satisfaction. Its creators also have argued that traditional market research is too complex and resource-consuming, and can produce results that take companies in the wrong direction.
The single number has face validity in that it indicates advantages versus disadvantages on a macro basis, and shows at least some linkage to business outcomes. It is a popular measure which facilitates scoring comparison to other companies. Its simplicity and elegance is compelling to corporate leaders, and its thesis is easy to grasp. Among other things, it created a recognition that:
- Companies need to set clear customer-related business outcome goals as a means of creating organizational focus and action
- Customer behavior is more than customer satisfaction, customer retention, and repeat purchase
- There should be at least a modicum of linkage between customer measures, obtained through a survey device, and actual business outcomes
- Customer data, beyond the merely anecdotal, should be shared within the organization; and this can create greater focus
- Results from this research can be used as a basis for performance comparison with industry peers
Even some of the technique’s harshest and most vocal critics within the market research industry have credited Satmetrix, the company offering the one-number approach, with getting the attention and acceptance of senior management, and then solidifying the position of recommendation question results as a corporate rallying point.
Recommendation has Limitations
Creators and sellers of the single recommendation question as a surrogate measure for profitability and growth also, increasingly, appear to be asking the business community to accept, on faith, the premise that Recommendation = Word of Mouth = Advocacy. This thesis, though the usually unchallenged annexation of the concept of advocacy has audacity, fails on a number of important bases.
Advocacy, as a concept and relative to recommendation, will be covered in more detail later in this article; but here let us address terms. The word ‘advocate’ has French and Latin origins. It has multiple applications, including legal, political, social care, and marketplace. It is the marketplace applications where the business, academic marketing, and management consulting communities have focused. Essentially, advocacy can be defined as active personal espousal or support of a brand, product, service, or company. Recommendation, simply, is one of the potential behavioral outcomes of advocacy.
Recommendation, as a measure and as a business or growth driver, is a different ‘animal’ when compared to customer advocacy behavior.
First, recommendation can be purchased. In some business sectors, companies can, and do, directly incentivize customer referrals; and there are ‘viral marketing’ and ‘buzz creation’ organizations whose sole business proposition is to use recruited, and compensated, individuals who will communicate their endorsement of selected products and services to others. This makes the referral scores for those products and services go up, of course, though the reasons for this increased recommendation are artificial, not voluntary or based on personal experience.
Next, as noted, recommendation is clearly only one of several downstream communication and action behaviors which can take place after a transaction or an experience. Offline and online informal dialogue between peers is one of them, and this behavior is certainly not coequal with either recommendation or advocacy (which also requires strong brand favorability). Neither is recommendation the most prevalent, nor impactful, action taken following transactional experiences. Based on research conducted through Harris Poll and other studies, the incidence of customers’ positive and negative communication on behalf of a brand, as a result of an experience, is far higher (among all age groups, and both genders) than positive, neutral or negative recommendation behavior. In addition, communication influence on customer behavior is as high, or higher, when compared to recommendation.
Findings of national polling studies have determined that customers can, as well, often communicate with the vendor or service provider involved; and this can take place in an array of ways, again both offline and online. Because the creators of the one-number approach have said they favor very short surveys, it is unlikely that the technique would ever generate types of diagnostic information which identify this type of downstream communication, or other behaviors taking place as a result of experiences and transactions.
…customer advocacy is based on such key elements as brand favorability, evidence and frequency of positive and negative voluntary personal communication, and continued consideration and relationship likelihood.
Also, customers can be completely silent after a transaction or experience; and this very lack of post-event communication has an impact on their future behavior (known as ‘self-perception theory’ in academic circles) all the while their failure to communicate has no influence on the behavior of others. Even though generally silent on their preferred brands, these customers (whom we identify as ‘Allegiants’ through our advocacy research framework), also demonstrate fairly strong loyalty. If companies can identify ways to build and bootstrap these quiet loyalists into more vocal supporters, the positive financial impact will be significant. Recommendation, as a single metric, cannot do this; and the ‘neutrals’ that the technique identifies are nothing like silent loyalists in terms of their behavior.
The recommendation question itself, i.e. ‘would you refer’, is found suspect by many, both inside and outside of the research community. It does not address actual customer history of referral or recommendation, but rather the likelihood that this action will take place.
Another key challenge of recommendation as a stand-alone measure is that the same customer can, and often does, actively recommend or refer competitive brands, services, or suppliers in the same business sector. There is nothing in the one number technique which requires a customer to narrow the consideration, or evoked, set or justify why they might recommend one, or multiple, brands or firms offering the same product or service. Customer advocacy measurement, though almost as simple as the recommendation metric, is much more rigorous, correlating highly to future purchase intent and other monetizing behaviors; and, driven by brand impression and evidence of positive or negative word of mouth on behalf of the brand or company, there is often a natural reduction of consideration sets.
Advocacy as Rational and Emotional Measure of Real-World Customer Behavior
As compared to individuals or customers who positively recommend, advocates are the deeply connected and brand-involved, energized, positive and vocal de facto sales force within a company’s, product’s, or service’s customer base. As we measure it, and as guided by ideas and findings published by major consulting organizations over the past decade, customer advocacy is based on such key elements as brand favorability, evidence and frequency of positive and negative voluntary personal communication, and continued consideration and relationship likelihood.
Advocacy identifies the monetizing downstream customer behavioral impact of informal communication, by individuals on a voluntary, active, peer-to-peer basis (and as it influences their own downstream behavior, i.e. the self-perception effect, as a result of personal experiences). Inclusion of the ‘personal experiences’ qualifier is critical, because it represents a depth of individual knowledge unidentified in the one-number recommendation approach.
Consulting firms such as McKinsey have determined that informal, voluntary peer-to-peer communication (i.e. word-of-mouth) drives 20% to 50% of customer decision-making, so it is extremely important; and it is every bit as behaviorally leveraging as recommendation. Recommendation, though, certainly isn’t word-of-mouth, nor is it advocacy behavior. It should be added that, just as recommendation isn’t word-of-mouth or advocacy, neither is it customer loyalty behavior (and can’t give management much detailed decision-making guidance for creating loyalty or pinpointing motivations for individual purchase choices); however, much of true customer loyalty behavior can be identified in drivers of customer advocacy.
When considered as a core measure, or metric, of customer loyalty and business health, advocacy can be expressed as a combination of two key constructs: rational (tangible and functional) and emotional (service and brand impression) toward a brand, expressed through brand preference and narrowed consideration, or evoked, set, high share of spend, and positive, frequent communication behavior on behalf of the preferred company, brand or product, principally through offline and online word-of-mouth. This approach, or framework, is far more robust, rigorous and actionable when compared to the single-number recommendation metric; and, per the above points, the willingness to refer or recommend can be considered one outcome of high-end loyalty or advocacy behavior, rather than the behavior itself.
Again, advocacy behavior isn’t the same as promotion or recommendation, nor is promotion or recommendation the same as advocacy behavior. Most research deals with customer retention and ignores share of wallet and client acquisition in their linkage research. As shown in this graphic from one of our bank customer studies, advocacy impacts all three:
New account activity is a major key performance indicator (KPI) for organizations such as banks, so the greater number, and polarity between customer groups, identified by customer advocacy research (positive and enthusiastic Advocates as the top segment, disaffected and negative Alienated customers as the bottom segment) is particularly significant.
Advocacy, in sum, is a more contemporary, real-world, and robust means for companies to understand drivers of customer marketplace behavior. As noted above, and shown in more detail in this graphic, levels of advocacy can be reliably linked to such KPI’s as retention and wallet share.
In my new book, The Customer Advocate and The Customer Saboteur (ASQ Press, 2011), I’ve devoted an entire chapter to this subject of recommendation and referral, and its value relative to customer advocacy as a performance metric and barometer. This article introduces a much more detailed presentation and discussion.