Customer success metrics and KPIs need to be reliable. They need to be telling your business useful truths.
Customer intel you can act upon. Warnings you can heed. Triumphs you can substantiate. Cause and effect you can use to try to predict outcomes.
NPS – Net Promoter Score – is pretty well accepted as the go-to metric for customer experience. So much so that asking customers “On a scale of 0-10, how likely are you to recommend our company?” is sometimes the totality of customer experience measurement for large organizations.
Unfortunately for them, relying on this measurement alone can render your capacity for insight one-dimensional.
How likely are you to recommend NPS?
Evidence has been mounting against the validity of the NPS truly being “the one question you need to ask” for some time.
This 2016 paper from Cambridge University cites research showing that NPS criticisms revolve around two broad failings. The first is NPS’s lack of depth. Depending on when and how often you ask the question (and in B2B, who you ask) you might only scratch the surface of what customers really think.
And besides, what about the risk of failing to measure responses from other people: employees, partners, suppliers…? The NPS question doesn’t really apply to their experience at all.
These issues get worse when you abstract NPS data across a customer base. Overall NPS data could be telling the business all is well, but revenues could be falling and customers leaving in their droves. NPS doesn’t diagnose underlying causal factors.
The second failing is NPS’s fixation on customer attitude rather than behavior. The NPS question invites the customer to consider a future action that they may never act upon. But customers recommend companies all the time, often on the back of recent experiences.
Because NPS isn’t retrospective, it doesn’t accurately record these.
Consumers saying what they’ll do versus doing what they say
The number of people who actually recommend a company has been found to be far smaller than the cohort who say they intend to – less than half according to this study.
Another study looked deeper and found significant instances of consumers both recommending and actively discouraging others from using the same brand.
The reasons given for this were perfectly logical. For example, recommending a brand to people who you think would like it and discouraging others for whom it seems inappropriate. Or loving a brand so much that you rate it a measly zero because of a new, minor aspect you dislike.
The sensible conclusion to draw from all this is that NPS is not perfect.
Useful but not perfect.
But with business leaders invested in the value of this as a ‘single’ metric, where do professionals charged with measuring the people experience go from here?
The Total Experience mindset
One of the three main themes in Gartner’s top strategic technology trends for 2021 is ‘people-centricity’. 2021 is a pivotal year for bouncing back from Covid-19 of course, so we totally agree people-centricity is perfect.
As we covered above, people means more than just customers.
What’s interesting is how Gartner sees this being applied by organizations, namely by unifying multiple people experience disciplines into one: Total Experience (TX).
‘Total experience combines traditionally siloed disciplines like multiexperience (MX), customer experience (CX), employee experience (EX) and user experience (UX), and links them to create a better overall experience for all parties. Not only does this streamline the experience for everyone, because organizations are optimizing across all experiences, it offers an excellent opportunity to differentiate an organization from competitors.” – Gartner
How might this work in practice? Well, it’s pretty clear these different disciplines have been separated until now for good reason. Different audiences, objectives, workflows, cadence – the list goes on.
Tying it altogether requires a single point of visibility and control: a single pane of glass, if you will. One that abstracts the complex interplay of customer touchpoints, employee interactions, user interfaces and automated processes.
And ideally works as an easily-deployable overlay to existing systems rather than a clunky application that takes years to rollout and integrate. This provides granular context for understanding NPS alongside other metrics, and their effect on your business.
Abandoning NPS would be a rash call. Rather, by using NPS alongside a basket of other supporting measures, organizations can make NPS more powerful and predictive.
And apply this to better understand and improve their customer experience.
Here are 4 considerations for what else to put into that basket and why:
A robust way of measuring true customer happiness
Every organization should be able to deduce how happy customers are at any point in time. At CT, we produce a simple ‘happiness factor’ metric which is calculated by:
Number of Happy responses divided by number of All responses
Another way of doing this would be to apply CSAT metrics and correlate this with an internal quality measurement taken from – for example – the customer support team. By measuring this over time, you can track the relationship between positive ‘happiness’ ratings and strong NPS scores.
Tracking satisfaction across actual referrals and purchases
As highlighted earlier, the big falling down with NPS is losing the connection between customers who say they’ll recommend you and the instances where they actually do. Understanding this in more detail through a referral program can uncover some critical insights that you can act upon.
This is all about making NPS a reliable indicator, and identifying how this differs according to variables such as customer demographic and product/service line.
Asking the right CSAT questions at the right time
Think about it from the customer’s perspective. Being asked lots of questions at once is tiresome and inconvenient. Single questions are better as they cause far less fatigue – hence the popularity of NPS.
But what if it’s always exclusively the same question over and over again? Possibly not. And is each customer looking at the question afresh each time, or in the context of being repeatedly asked?
Let’s take an example. Diners are used to being asked “Is everything alright with your meal?” about 5 minutes after their food arrives. Most people don’t go to the same restaurant every day (and order the same food), so the question doesn’t grate. The context feels right for that touchpoint, rather than, say, when you get the bill.
Customer support agents will often ‘inject’ a single question into the email notification that closes the enquiry, like: “How did we do today?” And it’s easy to see why that would be beneficial on so many levels.
But there are lots and lots of potential touchpoints, and many questions that could elicit profound insights into customer satisfaction, intent, emotion and loyalty. All of these boost the efficacy of NPS but also feed into direct improvements on literally any facet of the business from product development to employee onboarding.
In the words of the dishevelled TV detective, Columbo: “Just one more thing…”
Measuring CSAT: how hard are you making customers work?
The shift toward self-service and automation may have accelerated in the wake of Covid-19 as consumers adapt purchasing behavior. This is a great opportunity to enhance the customer experience at the same time as reducing costs of sale. This increases the importance of metrics like Customer Effort Score (CES) that get at how easy customers find buyer journeys and other interactions.
This is especially valuable when rolling out innovative new self-service features into live digital environments.
Research has found 81% of customers reporting ‘high effort’ interacting with a brand say they would speak negatively about that company to others.
This makes CES a useful supplementary layer of insight for determining advocates and detractors. CES works best when directed at specific processes rather than as a blunt instrument for the brand in general. Correlating the score with data about the process itself (e.g. conversion rates, dwell times, etc.) optimizes the validity of the results and makes them more actionable.
It’s unlikely one measure will give you everything you need. It’s also unlikely that anything will really change for your customers unless you act on the ground. We see many businesses with good measurement in place but no processes on the ground to identify individuals who are unhappy and actually change things for them. Or identify people who are delighted) and reward staff members who have caused that to happen.
The best basket of measures is manageable, and allows you to sense and respond to real people, as well as to track and record your progress.
If you need a lightweight basket of measures across your business, give Customer Thermometer a try for free: