Coming from Latin America, being succinct is by no means something that comes naturally - we are really long winded. During my first job, I remember being asked to try to write an email in one sentence; condensing what I thought was at least a page of information into one tiny sentence was a real challenge. Yet now after living in the US and Britain for over 9 years, I have come to appreciate the power of less.
Never does this notion of ‘the power of less’ hold truer than when used for ranking scales in survey design. The power of less, in this case, means higher response rates and more accurate results.
At Fizzback, we have discovered that when asking an NPS® question via SMS, response rates are between 30 to 50 per cent higher and results are more accurate when asking the question on a 1 to 5 scale (versus the traditional 0 to 10 NPS scale proposed by Reichheld).
There are a number of intuitive reasons for this; fewer thumb clicks (scoring a 10 takes 2) and simpler instructions are just two. But I think there is a deeper reason than this:
As many of you will know Reichheld’s classic 0-10 NPS scale segments promoters as customers that score 9’s and 10’s, passives as customers that score 7’s and 8’s, and detractors as customers that give scores between 0 and 6.
But is this distinction cross-industry applicable? Is an 8 a “real” passive and is a 6 a “real” detractor?
By feeding comments and ‘classic NPS’ scores through Fizzback’s Natural Language Processing (NLP) and looking at the results, we can gain an insight into whether the 8’s and 6’s actually follow the segmentation that the scale posits.
From a cross-industry perspective, there is only around a 50 to 70 per cent chance that Reichelds’s segmentation proposition is accurate. We believe that utilising the classic NPS scale via SMS - rather than adding granularity - adds confusion. By reducing ambiguity in the scale, we see improvements in correlation between the scores and the verbatim sentiment captured by our NLP.
Obviously, as with everything in life, there are trade-offs in going for the 1-5 scale. The classic NPS scale comes with its considerable legacy and traction in boardrooms worldwide, driven largely by the benchmarking potential it affords.
But for me – in a trade-off between legacy and like-for-like benchmarking and core Customer Experience Management (CEM) principles of gaining volume and accuracy – there can only be one winner. So in the search for accurate and robust results I suggest that less is more.