This year has been the Net Promoter Score’s (NPS) moment in the shade. The popular metric for measuring customer satisfaction got a drubbing this May from the Wall Street Journal, which noted that while corporate America may be obsessed with the NPS — often using it to determine bonuses and guide investment decisions — the metric’s accuracy is “dubious.”
Forbes, too, chimed in with an opinion essay, the headline of which proclaimed its feelings up front: “It’s Time to Retire the Net Promoter Score.” PwC consultant Matt Egol also expressed the view that the metic is “past its prime.”
While criticism of the NPS has been loud lately, it is certainly nothing new. Multiple studies over the past 15 years have shown there is no validity to the claim that the NPS question is the best predictor of growth.
Devised in 2003, the NPS is based on a single question: “How likely is it that you would recommend [company/product/service] to a friend or colleague?” The NPS methodology separates those who respond most favorably with a rating of 9 or 10 (promoters) from those who respond less favorably with a rating of 6 or lower (detractors). The percentage of promoters minus the percentage of detractors gives us the NPS itself, and a high positive value is deemed desirable.
The question at the core of the NPS itself is valid, but research has shown that it’s not any better than traditional customer satisfaction questions. Neither the scoring technique nor the question itself provides an advantage in predicting business outcomes. In fact, asking just one question about willingness to make a recommendation provides a measure that is too narrow to predict customer retention or increased purchasing. The NPS’s limited scope is much better suited for predicting brand advocacy, if anything.
NPS is okay, but it’s not the predictive panacea its advocates have claimed it to be. Of course it isn’t: Behaviors caused by multiple factors (like customer retention) are best predicted by measures of multiple factors (unlike the NPS).
Exit the Employee Net Promoter Score
Despite the research that has revealed its shortcomings, the NPS quickly gained and has enjoyed widespread acceptance, perhaps because it offers a fast, simple solution to a complex problem. Once the NPS had a firm foothold in corporate America, many executives and chief HR officers adapted it for use in employees surveys. Organizations began to track employee sentiment by asking, “How likely is it that you would recommend [company] to a friend or colleague as a place to work?” complete with the 11-point response scale and its scoring procedure. This employee NPS (eNPS) came to be used as an overall metric that represented an entire set of survey topics. For some companies, eNPS is the engagement score and the only survey score that matters.
Once again, however, research has put a dent in the eNPS’s claims of superiority, while also uncovering practical disadvantages such as limited diagnostic capability, increased cultural bias, and poor applicability to small workgroups. Like the NPS, the eNPS seems to be facing a moment of reckoning, and organizations now have more reasons to replace the eNPS than to continue using it.
If your company’s biggest problem is recruiting new talent, then eNPs will be a very helpful metric, because you are asking workers if they would recruit for you. Talent acquisition, however, is rarely the change with the most potential to improve an organization. The eNPS scoring procedure has also created problems for survey programs: The focus on extremely positive responses (i.e., promoters) ignores other positive responses, causes cultural biases in countries where extreme responses are shunned or encouraged, and complicates data interpretation for small work groups. All of this is to say that using the eNPS technique in an employee survey will complicate the reporting phase without supplying the most appropriate predictive metric the organization needs to improve business performance.
Replacing the eNPS With Something Better
Over and above the eNPS, I’d recommend different survey metrics for organizations that want to drive different outcomes, such as agility, innovation, customer focus, etc. Because we know there are multiple contributors to an organization’s culture, the better bet is to build a multi-item index score that is applicable for various organizational outcomes. Using a multi-item score rather than the eNPS provides a more reliable and precise measure that improves prediction and actionability. I’d also recommend a simpler scoring procedure that calculates a percent-favorable metric that is easier to explain and simpler to use and translate into action.
Of course, organizations should feel free to continue to use the eNPS question if it makes sense for them. However, best practices now demand more varied metrics so that a survey program can evolve to create better outcomes.
Paul Mastrangelo is principal strategist at CultureIQ.