Measuring Customer Satisfaction: How to Know Your Customers Are Happy

For the owner who already has the records, the systems, and the years behind them: the point here is not another thing to store. It is turning what you already have into one honest signal that tells you whether your customers are actually happy, or only quiet.

The question this answers

You have a good week. The orders come in, the till moves, nobody shouts at anyone. So your customers must be happy. Right?

Maybe. The trouble is that “nobody complained” and “everybody’s happy” are not the same thing, and the day you find out the difference is usually the day the repeat orders have already dried up. If your read on customer satisfaction is a gut feeling, you are guessing. And you cannot improve a guess.

Quick answer

Key takeaways

• If you don’t ask, you’re guessing. Satisfaction happens inside the customer’s head, so a busy shop can sit on top of quietly disappointed people¹.
• “No complaints” is not “satisfied.” Most unhappy customers never complain. They just stop coming back^2,3.
• What customers do beats what they say. Your repeat rate is a truer signal than any form, and you can read it from records you already keep^3,4.
• Track your own trend, not a benchmark. There is no reliable “good score” to chase. The only fair comparison is your own number last month against this month.
• A number you never act on is decoration. A signal earns its keep only when a dip triggers a question, a fix, and a fresh reading.

See: the happy customers you never hear from

Picture a well-run supplier in Lagos. Solid systems, a proper customer list, steady repeat trade. The owner would tell you her customers are happy, and she would mean it, because the phone rarely rings with a complaint.

Now look closer. A handful of her steady buyers have quietly drifted to a competitor over the past two quarters. Not one of them called to say why. They did not argue about a price or file a grievance. They simply placed their next order somewhere else. On her sales report, nothing looks wrong until you line up the months and notice the same names have gone missing.

That silence felt like satisfaction. It was the opposite.

Understand: the pattern — you can’t improve what you don’t ask

Here is the pattern worth carrying into every customer relationship you have.

Two well-tested ideas sit underneath it.

First, satisfaction is a comparison, and it is invisible. The most tested model of how satisfaction works says a customer forms an expectation before they deal with you, then privately weighs what they actually got against it¹. Because that weighing happens in their head, it never rings up. A full day and a busy shop can sit right on top of customers who quietly expected more. The only way to see it is to ask, or to watch what they do next.

Second, most unhappy customers won’t tell you. When a customer is let down, they have a choice: speak up, or slip away. Decades of work on how people respond to decline show that many say nothing and simply leave, or do nothing at all, rather than raise a complaint². The national satisfaction science points the same way: complaints tend to fall as satisfaction rises, so a quiet complaints book can mean happy customers, or it can mean unhappy ones who already gave up on you³. You cannot tell which from silence alone. That is why a complaint is almost a gift, and the dangerous customer is the one who leaves without a word.

Put the two together and the conclusion is plain. If you are not deliberately collecting a signal, you do not know how your customers feel. You are hoping.

Decide: pick one lightweight signal

You do not need a research department. You need one signal you will genuinely keep collecting, chosen to fit how you already meet your customers. Here are the practical options, in plain terms.

Signal	How to collect it	What it tells you	Watch-out
One-question follow-up	After a sale or job, ask a single plain question (“How did we do?” or “Would you use us again?”) by message, on the receipt, or in person	A quick read of how that customer felt, caught before they leave in silence	Keep it to one question asked soon after; a long form gets ignored
Repeat / return rate	Count the share of customers who come back over a period, straight from records you already keep	What people actually do, which is a truer signal than what they say3,4	Return rate also moves with price, location and competition, so always read it next to a “why”
Short review ask	Invite happy customers to leave a public review	A satisfaction signal that doubles as word of mouth for new customers (direction)5,6	The sample is self-selected: the delighted and the furious post, the quiet middle does not. Read it as direction, never a headcount, and never buy or fake a rating
CSAT (a satisfaction score)	Right after one interaction, ask “how satisfied were you?” on a small scale, say 1 to 5, then track the share who are satisfied	How one specific moment landed (a read of the transaction)	It measures that single moment, not the whole relationship
NPS (a recommend score)	Ask “how likely are you to recommend us, on a scale of 0 to 10?” The score is the share who answer 9–10 minus the share who answer 0–67,8	A simple, repeatable read of the overall relationship	Its famous claim to predict growth is not supported by independent research4, and it does not tell you why. Ignore any “a good score is X” benchmark

A note on the last two, since they get thrown around as jargon. CSAT is just “how satisfied were you, right now?” NPS is just “would you recommend us?” turned into one number. Both are fine, simple signals. Neither is magic. The score alone tells you very little without the follow-up question of why, so treat any single decimal with suspicion.

If you already run on records and systems, the honest starting pair is the two you can get almost for free: your repeat rate from your own sales history, plus a one-question follow-up after each job. Add a formal score only once it will actually change a decision.

Act: install one signal into normal life

The reason most measurement fails is not the method. It is that the method sits outside the daily flow of the business, so it quietly stops happening.

So do not launch a “satisfaction programme.” Instead, bolt one signal onto something you already do:

• If every job ends with an invoice or a delivery message, add the one follow-up question to that same message.
• If you already pull a weekly or monthly sales summary, add one line to it: how many customers came back this period.
• If you thank customers at handover, that is the natural moment to ask for a short review.

One signal, riding on a habit you already have, collected the same way every time. That is what survives a busy month. A fancier method you abandon in three weeks tells you nothing.

Measure: watch the trend, then close the loop

A single reading is almost meaningless. The intelligence is in the direction. Is your repeat rate climbing or slipping? Are the follow-up answers warmer or cooler than last quarter? Compare your number to your own past, not to some benchmark you read online, because there is no reliable universal “good score” that holds across trades and countries.

Then do the part that makes measuring worth the effort at all: close the loop. A number you collect and never act on is decoration. So when the signal dips, treat it as a question, not a verdict. Ask a few customers what changed. Find the moment that soured. Fix it. Then read the signal again to see whether the fix worked. Signal, to decision, to action, to a fresh reading. That loop is the whole point.

Measuring is a loop, not a one-off: a signal points to a decision, the decision drives an action, the action gives you a fresh reading, which becomes the next signal.

This is also exactly the kind of signal that belongs in a wider view of your business health, sitting next to your margins, your cash, and your pricing. On its own it tells you about customers. Alongside the rest, it tells you whether the whole business is holding together.

Related reading

• Customer Experience for Small Businesses — the cornerstone this brief sits under: what good customer experience actually is, and how to close the gap you are now measuring.
• Why Customers Never Come Back — a closer look at the silent exit: why customers leave without a word, and what to do about it.
• Customer Lifetime Value — what a satisfied, returning customer is actually worth to you over time.

Check your own business

Find out how healthy your own business is. Take the free Business Health Check. A satisfaction signal tells you about your customers. The Health Check puts that next to your profit, your cash, and your pricing, so you can see how the whole business is really doing, not just one corner of it. The Monyvo Business Library has short companion reads that pair with this brief.

---

Evidence & Confidence

How sure are we of the main ideas here? Rated by the basis of each claim, not by how much we’d like it to be true.

Claim	Confidence	Basis
Satisfaction is a private comparison, so it can’t be read off your sales figures	★★★★★	Foundational, heavily replicated theory (Oliver)
“No complaints” is not the same as satisfied; most unhappy customers leave quietly	★★★★☆	Foundational theory (Hirschman) plus a validated national index (ACSI); the viral “X% never complain” ratios are unreliable and are not published
What customers do (repeat/return) is a truer signal than what they say	★★★★☆	Well-supported direction (ACSI; independent loyalty research)
NPS is a simple, useful signal, but not a proven predictor of growth	★★★★☆	Its definition is settled; the “predicts growth” claim is refuted by peer-reviewed work (Keiningham and colleagues)
Asking happy customers for reviews feeds word of mouth	★★☆☆☆	Direction is well-attested; the specific percentages are commercial-survey figures and are not published
A local benchmark for a “good” African satisfaction score	★★☆☆☆	No reliable Africa-specific figure was found; compare to your own past instead

Sources

1. 1. Oliver (1980), “A Cognitive Model of the Antecedents and Consequences of Satisfaction Decisions,” Journal of Marketing Research 17(4):460–469. The expectation-disconfirmation model: satisfaction is the private comparison of what a customer got against what they expected. Concept and direction.
2. 2. Hirschman (1970), Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States, Harvard University Press. Dissatisfied people often exit quietly, or do nothing, rather than voice a complaint. Concept and direction — not a source of complaint percentages.
3. 3. The Science of Customer Satisfaction, American Customer Satisfaction Index (ACSI), University of Michigan (est. 1994). Validated cause-and-effect model: satisfaction drives loyalty, while complaints move inversely to satisfaction. Anchors the “complaints are not satisfaction” and “behaviour is the truer signal” logic.
4. 4. Keiningham, Cooil, Andreassen & Aksoy (2007), “A Longitudinal Examination of Net Promoter and Firm Revenue Growth,” Journal of Marketing 71(3):39–51. Peer-reviewed: NPS performs no better than other satisfaction and loyalty measures at predicting growth. The binding counterweight to the NPS growth claim.
5. 5. BrightLocal, Local Consumer Review Survey. Commercial survey. Carried for direction only — that reviews influence whether new customers choose a local business, and that asking works. All specific percentages are not published.
6. 6. Same review-behaviour evidence series as the source above, alongside the satisfaction-to-word-of-mouth chain from the cornerstone research. Direction that a review ask feeds word of mouth. Magnitudes not published.
7. 7. Reichheld (2003), “The One Number You Need to Grow,” Harvard Business Review 81:46–55. Origin of NPS: the 0–10 “would you recommend” question and the promoter/passive/detractor bands. Used for the definition only; its predictive-superiority claim is contested.
8. 8. “Net promoter score,” Wikipedia (accessed 2026). The settled definition (the 0–10 question, the 9–10 / 7–8 / 0–6 bands, and %promoters minus %detractors) and a summary of the academic criticisms. Concept only.