Are You a One Dial Pilot?

Two posts ago, we looked at dashboards that display data but never trigger action. Last week, we looked at why most VoC programs only have one working wing. This week: why one score was never going to be enough in the first place.

Would you fly with just one dial?

Would you get on a plane if the pilot announced, “I’ll be flying using just the altimeter today”? No airspeed indicator, no attitude indicator, no compass, no radio. Just the one dial, glowing reassuringly in the dark. I’d want to get off that plane before it left the gate.

Yet this is exactly how many companies run their feedback programs. One score. One dial. Usually Net Promoter Score (NPS), sometimes Customer Satisfaction (CSAT), occasionally Customer Effort Score (CES). Whichever it is, one number is expected to tell leadership everything it needs to know about the health of the customer relationship, and by extension, the health of the business.

It can’t. And it was never meant to.

Like every other feedback platform, SurveyVista makes it easy to build a dashboard full of these single-number gauges. That’s not the problem. The problem is when the score becomes the purpose, and the single number becomes the only landmark leadership navigates by.

The one-number promise

Fred Reichheld’s 2003 Harvard Business Review article, “The One Number You Need to Grow,” made the case that a single question — “how likely are you to recommend us?” — correlated with company growth better than other survey questions tested at the time. It’s a genuinely useful finding, and NPS has earned its place in the boardroom because of it. But a useful finding is not the same as a complete instrument panel, and plenty of people have since read “correlates with growth” as “tells you everything.”

Reichheld isn’t the only one to make a case for a single defining metric. In 2010, Matthew Dixon, Karen Freeman, and Nicholas Toman argued in “Stop Trying to Delight Your Customers” that Customer Effort Score was the better predictor, at least in service interactions, and that reducing customer effort mattered more to loyalty than trying to delight people.

Both pieces of research are real and useful. Neither was designed to fly the whole plane alone, and I doubt either set of authors would claim otherwise.

The problem isn’t the metric. It’s the expectation loaded onto it: that a complex, multi-dimensional relationship between a customer and a company can be steered using a single dial.

Complex systems need more than one instrument

Aircraft pilots monitor altitude, airspeed, attitude, heading, vertical speed, and turn — six key factors, each with its own instrument, each telling the pilot something different, none of them sufficient alone. Altitude without airspeed tells you nothing about whether you’re about to stall. Heading without attitude tells you nothing about whether you’re flying level or spiraling into the ground.

Customer relationships are at least as complex as a plane. A single relationship score can’t tell you why it moved, where in the journey the damage was done, whether it’s an early warning or a lagging scar, or what to do next. Feedback programs need their own instrument panel:

• A relationship metric like NPS, CSAT, or a similar one is your altimeter. It tells you the altitude of the relationship, roughly, but nothing about your heading or speed.
• An interaction or effort measure is your airspeed indicator. Effort scores and transactional feedback tell you how hard customers are working to get what they need, moment to moment.
• Usage and engagement data is your attitude indicator. Product usage, adoption, and engagement trends often move before sentiment does, showing the angle of descent before passengers feel it.
• Operational and service data is your compass. Complaints, escalations, response times, and resolution rates tell you whether you’re heading toward or away from the outcomes that matter.
• Financial and commercial data is your fuel gauge. Retention, expansion, churn, and LTV: CAC are the only true tests of whether any of the above is actually working.

None of these dials flies the plane on its own. Together, they let you pilot the relationship.

Leading and lagging, not either/or

Part of the confusion is that most feedback metrics are lagging indicators. They tell you what has already happened. The NPS score reported this quarter reflects relationships built, or damaged, over the preceding months. By the time it moves, the moment to act on the original cause has usually passed.

Leading indicators like in-product usage signals, support ticket trends, effort at the point of interaction, and front-line employee sentiment give you the chance to act before lagging metrics confirm the damage is done. 

A feedback program that only tracks lagging measures is a pilot reading the altimeter after the mountain, not before it. It’s the difference between a survey program and a feedback program.

Data, insight, decision, action

Even with a full panel of dials, none of it matters if nobody’s hand is on the controls. There’s a link in too many feedback programs that breaks somewhere in the middle:

• Data is what the dials show. 
• Insight is understanding what the data means and why it moved. 
• Decision is choosing what to do about it. 
• Action is actually doing it.

Most organizations invest heavily in the first link and not enough in the last. They can produce beautifully cross-tabulated data and impressive dashboards. Fewer can tell you the insight it points to. Fewer still can tell you the decision that insight drove. Vanishingly few can point to the action that decision produced, and the result it delivered.

Take a SaaS business watching renewal risk. The data is a dip in product usage among a cohort of accounts. The insight is that the dip follows a support ticket about a feature that didn’t work as expected. The decision is to fix the feature and proactively contact every affected account before renewal, not just the ones who complained. The action is doing exactly that, then watching usage and the next relationship score to see if it worked. Skip any one of those four links, and you’re back to reading a dial for its own sake.

Are you a one dial pilot?

If your feedback program reports a single number, ask what it isn’t telling you about heading, speed, and rate of descent. If it reports many numbers but nobody owns turning them into decisions and course corrections, you have dials but not an effective pilot.

Build the panel. Assign the pilot. Make sure someone’s hand is always on the controls. Because a plane doesn’t stay in the air because the altimeter is accurate. It stays in the air because someone reads all the dials and flies it.

What’s next in this series

An instrument panel is only useful if every dial is reading from a rich enough picture to trust. Part 4 looks at why most feedback data is a handful of dots pretending to be a finished picture.

Frequently Asked Questions

1. Is NPS a bad metric for measuring customer experience?
   No, NPS is a genuinely useful correlate of growth, as Fred Reichheld’s original research showed. The problem isn’t the metric itself, but treating a single relationship score as sufficient to steer a complex, multi-dimensional customer relationship on its own.
2. What metrics should a feedback program track besides NPS or CSAT?
   A relationship measure (NPS/CSAT), an effort or interaction measure (CES or transactional feedback), usage and engagement data, operational data (complaints, escalations, resolution times), and financial data (retention, expansion, LTV: CAC) used together as an instrument panel, not in isolation.
3. What’s the difference between a leading and a lagging indicator in customer feedback?
   Lagging indicators, like a quarterly NPS score, reflect what has already happened. Leading indicators, like usage trends or support ticket volume, can signal a problem before it shows up in a relationship score, giving teams the chance to act before the damage is fully done.