NPS: The holy grail or Emperor’s new clothes?
Peter Grist takes a hard look at NPS, its value and limitations, and how it can be used as it was originally intended.
Some people soak in a warm bath. Others practice yoga. For me? Unwinding after a long day’s work means curling up with a copy of APRA’s latest information paper, Self-assessments of governance, accountability and culture.
As someone whose bread and butter is helping people deliver amazing customer outcomes, my attention was piqued by APRA’s discussion of ‘conduct risk’.
ASIC defines conduct risk as: The risk of inappropriate, unethical or unlawful behaviour on the part of an organisation’s management or employees.
After analysing all the self-assessments submitted by the banks, APRA observed that there’s very little rigour around monitoring conduct risk – with the banks heavily reliant on net promoter score.
In APRA’s words: It appears that standalone monitoring of ‘conduct risk’ is relatively new, often with a focus on the customer ‘net promoter score’, with no analysis or reporting of complaints data that could be used as a lead indicator.
Of course, Australia’s banks are not alone, with two-thirds of Fortune 1000 companies using the metric in an attempt to measure success in a customer-centric way.
To me, it’s mind-boggling that the world’s leading organisations aren’t holding themselves to account with leading (input) indicators.
You know, the ones that actually clue people in on what they need to do to drive a result.
In the beginning
NPS was launched by Frederick Reichheld – a Bain & Co consultant – in a 2003 HBR article: The one number you need to grow. What a neat antithesis to the increasingly complex requirements of running a people-facing business. No wonder it took off!
It was billed as the ‘simplest, most intuitive and best predictor of customer behaviour’ and a ‘useful predictor of growth.’
I have often used a company’s NPS increase as a lagging indicator of the impact of a GRIST learning intervention. I’ve also seen firsthand the unifying effect that NPS can have on an organisation. One team, one dream: shift that number!
NPS is also unique in that almost everyone is measuring it, so it becomes a useful benchmark for both industry-wide and cross-industry comparisons.
The ease and simplicity of a single number to track and compare over time is brilliant. It makes absolute sense that organisations got on board with it.
The problem is, it’s now accepted as conventional wisdom. Conventional wisdom sucks us into groupthink, which stifles innovation and often defies logic.
It’s time to look at the limitations of NPS. It’s time to pull back the curtain.
The journey away from NPS
Two years ago, a Big Four bank data and analytics executive confessed they’d never found a correlation between NPS and the metrics that really matter.
Since then, I’ve increasingly started to question the value of NPS. I suspect that there are a lot of you who – on the surface – are slavishly following the cult of NPS, but who, deep down, aren’t buying it.
I’m certainly not the only one writing about whether a high NPS is indeed the one number to rule them all, the holy grail of CX.
The Wall Street Journal published a takedown of NPS earlier this year in the article, The dubious management fad sweeping corporate America.
In it, they quote Reichheld as he laments how the metric is being misused and abused:
He is astonished companies are using NPS to determine bonuses and as a performance indicator. “That’s completely bogus,” Mr Reichheld, who still consults for Bain, said in an interview. “I had no idea how people would mess with the score to bend it, to make it serve their selfish objectives.”
Unimpressed with the general tone of the WSJ article, Reichheld came out swinging on Bain & Co’s blog to clarify what NPS is a measure of:
Instead of NPS, perhaps I should have called the system ‘NLE,’ for Net Lives Enriched. After all, that is what NPS really measures. Of all the lives you touch, how many are enriched, how many diminished? By focusing on that question each day, you can make the world a better place—especially if you measure, test and learn.
When employees know they regularly make their customers’ lives better, they bring energy, enthusiasm and creativity to work, spurring further innovation and learning. When you have such a positive impact that a customer gives you an unqualified recommendation, saying he or she would want a loved one to have a similar experience, this is clear evidence you made someone’s life better. And that’s what it’s all about.
You’ll note there’s no mention of its predictive power of growth. Instead, Reichheld touches on how NPS can motivate customer-facing people to deliver a positive experience.
Certainly a noble intent. In practice, however, it rarely manifests in such a customer-centric way.
Well, what then?
To be clear, I’m a big proponent of using data insights to drive an improved customer experience.
It’s just that NPS isn’t the right measure.
While NPS was initially well-researched and validated by Bain & Co, in a practical sense, NPS has become a vanity measure. A measure that gives organisations an excuse to not bother with the hard work of listening to and learning from customer interactions.
Rather than driving change that improves the customer experience, NPS is more likely to drive change that improves the NPS score, with many organisations focusing on the NPS as the outcome.
Whatever happened to the dogged pursuit to deliver amazing customer experiences?
When NPS goes rogue
Here’s a snapshot of what I’ve seen in the pursuit of improved NPS scores:
Agents coaching customers on why they should give a better score, or explaining how the scoring works
Organisations choosing which interactions they evaluate, such as only successful sales interactions
Manipulating the timing of the survey to ensure the best results (for a contact centre, for instance, this would be at time when wait times are low)
Another way
A better foundation for measuring success is a framework that clearly defines what a great customer interaction looks like.
This framework needs to be validated through voice-of-customer research and correlation analysis against the desired outputs (lagging indicators).
Doing so requires organisations to commit to the hard work of observing, assessing and developing the behaviours that are proven to drive a positive customer experience, rather than simply gaming a number.
For the better part of a decade, GRIST’s research team has been assessing customer conversations at a behavioural level to understand what really drives customer and financial outcomes.
As an example, after assessing thousands of conversations and tracking the inputs to the outputs, we know that there is a strong link between specific, measurable micro-behaviours, such as empathising with the customer, and a positive sales outcome. The old adage ‘we buy from those we like’ is true.
Of course, leaders can’t just coach their people to ‘be more likable’ – but they can measure when and how often their people empathise with customers, and then coach them to demonstrate that behaviour more consistently.
Organisations can’t beat their people over the head with sales targets anymore but they are just as beholden to their shareholders as they’ve ever been. Understanding that customer-centricity and solid financial results are not incongruent, traditional sales-based scorecards can be translated into the corresponding behaviour-based leading indicators.
In doing this, scorecards become customer-centric by design, and organisations have clarity into which behaviours achieve the desired outcomes.
Time for bed
NPS is so ubiquitous it seems we’ve lost track of what it actually means. Sure, its simplicity has its benefits. However, this is no substitute for good science. What customer-facing team members really need is clarity into which levers to pull to drive growth via better customer outcomes. Lagging indicators like NPS just don’t cut it.
And on that note, thank you and good night.
Contact us to learn how GRIST’s micro-behavioural approach can help your organisation achieve excellence by degrees.
