Nobody recommends their water company
A customer metric for categories where recommendation doesn’t work.
The problem with the metric you’re probably using
Net Promoter asks one question: would you recommend us? It’s a fine question for a restaurant or a pair of running shoes. It’s a strange one for a water utility, an electricity network, a super fund your employer chose, or a motoring club your parents joined. Nobody recommends these because nobody chooses them, people don’t talk about the category, or switching is rare, painful, or impossible. So the score measures something, but not the thing you need. Three failures show up again and again.
It swings. NPS is answered through the lens of the last interaction, so it moves with the last bill, the last outage, the last hold time. A board sees a six-point drop and demands an explanation for what was really a bad fortnight.
It’s a complaints channel, not a value channel. Satisfaction and NPS are built to find what’s going wrong. They tell you very little about what you’re actually strong at, why customers stay, or what to protect.
It can’t see the relationship. A customer can be irritated by a call centre and still, underneath, trust the organisation. NPS collapses both into one number.
Ask how they’d feel about losing you
The Customer Value Indicator turns the question around. Instead of asking whether customers would recommend you, it asks how they’d feel about losing you.
If a different organisation took over supplying your water tomorrow, how would you feel?
That question can’t be answered from the last bill. It reaches for the accumulated relationship: whether there’s any value in the organisation beyond the commodity it delivers. Which is precisely what a monopoly, a mutual or a regulated business needs to know, and precisely what the regulator, the board or the members will ask.
Three customer types, and one score to track
Supporters
Would be sorry to lose you. They carry your reputation, absorb a price rise and forgive a bad month. Understanding what they value tells you what to protect.
Potentials
Are neutral. Something about the relationship hasn’t landed yet. They’re the group most worth studying, because what would move them to Supporters is your growth agenda in one list.
High effort
Customers get little from the relationship beyond the service itself. They’re the ones a competitor, a regulator or a bad news cycle will find first.
The headline CVI score tracks wave on wave and, where we’ve built one, against a category benchmark, so you can see both movement and relative position. Because the question isn’t tied to the last interaction, the score is stable enough to rally an organisation around and sensitive enough that a real improvement shows up as a trend rather than noise.
Where it’s been used
Lady Gowrie Tasmania uses it to track how parents value its long day care centres.
The same logic transfers directly to water corporations, electricity networks, member-owned insurers and banks, early learning and aged care providers, and any organisation whose customers didn’t really choose it or a category that people wouldn’t recommend.
What a project looks like
Add CVI to what you already run. One question and a short diagnostic block added to your existing survey, reported alongside your current metrics so you can see what each one is and isn’t telling you. Cheap, fast, and the quickest way to find out whether your satisfaction score has been hiding something.
A value study. A standalone piece of work that measures CVI, profiles the three types, identifies what Supporters value and what would move Potentials, and turns that into a short, defensible narrative about the value the organisation creates. This is the version that goes into a price submission, a member communication or a board strategy day.
Either way, it’s senior researcher access from start to finish. No account management layer.
What would your customers say about losing you?
If your satisfaction score isn’t telling you what you need, I’d love to hear what you’re tracking. No deck, just a conversation.











