Share article
A few years ago I sat in a meeting where marketing proudly presented how they’d improved “overall satisfaction with the customer journey” by three percentage points. It looked great on the slide. The problem was that nobody had noticed complaints about late deliveries had risen by a tenth in the meantime. We’d improved the average. Customers dealing with an actual problem couldn’t have cared less.
This is the most common mistake I see in CX: chasing average satisfaction across the whole journey instead of admitting that not all touchpoints carry the same weight. A handful of them decide whether a customer falls in love with a company or runs from it. The rest is noise.
What counts as a moment of truth, and what’s just another touchpoint
The term “moment of truth” was popularised by Jan Carlzon, head of Scandinavian Airlines, in the 1980s. His observation was simple: each of SAS’s ten million customers came into contact with an average of five employees, and each contact lasted around 15 seconds. That’s 50 million moments a year in which the company is either built up or torn down in the customer’s eyes. Carlzon built SAS’s entire reorganisation on this insight, handing decision-making power to frontline staff, because they are the ones who actually hold those moments of truth in their hands.
Twenty years later, P&G introduced two further terms that are still used in marketing today: the First Moment of Truth (FMOT), those few seconds when a customer stands in front of the shelf or on a product page deciding whether to buy, and the Second Moment of Truth (SMOT), the experience with the product after purchase that determines whether they come back.
In practice, this means the following: a touchpoint is any contact a customer has with a company, logging into an app, waiting on the phone, receiving an invoice by email. A moment of truth is a touchpoint with an above-average impact on whether the customer stays, recommends the company, or leaves. You can recognise one by three things:
- there’s high risk or strong emotion involved: something has gone wrong, something is uncertain, something is happening for the first time
- the customer is actively making a decision at that moment, rather than just moving mechanically through a process
- the impact on behaviour (retention, recommendation, spend) is measurable and noticeably higher than at surrounding points in the journey
Almost everyone makes the same mistake here: treating all touchpoints as equally important, because that’s easier to report on. In reality, a long journey might have thirty touchpoints, but only five to eight of them actually decide the outcome.
How to find critical moments in the data, not by guesswork
You can’t just guess this from your desk. You need to combine two types of data.
First, a correlation or driver analysis between ratings of individual journey touchpoints and the outcome metric, NPS (Net Promoter Score, a measure of willingness to recommend), retention, or repeat purchase. The points where a small drop in satisfaction drags down the overall result the most are your candidates for a moment of truth. A point with a low rating that barely correlates with the outcome at all is simply annoying, not critical.
Second, an analysis of open-text feedback (verbatims) and contact centre calls. You’re looking for places where strong emotion shows up: words like “desperate”, “cheated”, “finally”, “I didn’t expect that”. Emotion is a sign that something critical is happening there.
Add to that an analysis of drop-offs along the journey: where customers most often stop buying, cancel an order, hang up the phone, or delete their account. A sharp drop-off at one specific step almost always points to a moment of truth.
I recently spoke to a colleague who works in CX at a bank. Her team had long believed the key moment was signing the mortgage agreement. The data showed something different: most people dropped off (and rated the experience worst) while waiting for the loan approval decision, days of uncertainty with no information at all. Not the signing. The waiting. Without data, they’d never have found this, because it isn’t a point you’d normally map as a “milestone”.
Why it pays to invest in a few places, not spread thin
Here’s the uncomfortable truth: the CX budget is always limited, and improving “everything a little” across the board almost never shows up in the business numbers. Improve ten touchpoints by five percent each, and NPS moves by a tenth of a point, because you spread the money over things the customer didn’t actually care about.
McKinsey’s “Consumer Decision Journey” framework has long argued exactly this: companies should define the specific “moments that matter” and deploy resources where the marginal impact is greatest, rather than spreading them evenly across the journey. In practice, this means a focused improvement to five key touchpoints usually shows up in retention more than a diluted improvement of a few percent spread across thirty touchpoints at once.
It’s also a question of organisational capacity. Tell a team to “improve the customer experience” and nobody knows where to start, so the energy gets dissipated into reports and dashboards. Tell them “this one specific step needs to look different within a year” and it becomes something you can plan, measure, and hold someone accountable for. A moment of truth is a unit of work you can actually act on. “Overall satisfaction” is not.
Moments of truth across industries
In practice, this is what it looks like across different sectors:
In air travel, it isn’t check-in at the desk, even though a great deal of money gets spent on it. It’s communication during a delay or cancellation, specifically how quickly and clearly the airline explains what’s happening and what comes next.
In banking, it’s usually the approval or rejection of a loan and the first disputed transaction. Not opening an account, which is largely smooth everywhere these days.
In e-commerce, it’s resolving a delivery problem or a complaint, not the checkout itself. A customer whose problem gets fixed quickly and without a fight often ends up more satisfied than if nothing had gone wrong in the first place (this is known as the service recovery paradox, though treat it with caution: it doesn’t excuse problems that keep happening again and again).
In healthcare, the moment of truth is delivering a diagnosis or waiting for test results, not patient registration.
In telecoms, it’s the first month after switching to a new provider, not the first bill, the first outage, the first call to customer service. The experience in those first 30 days strongly shapes whether a customer stays for years or leaves soon after.
In hospitality, it isn’t check-in itself, but how a problem gets handled: the wrong room, a missing reservation, noisy surroundings. How reception responds decides the rating far more than how clean the sheets are.
Notice the pattern: it’s almost always about moments of uncertainty, problems, or first-time experiences. Not the routine steps that work the same way every day.
How to get started on Monday
You don’t need mature CX analytics to get started, just three steps. Take your customer journey map and, for each point, note whether there’s high risk or emotion involved, whether it’s a first-time experience for the customer, and what happens if it goes wrong. Points with at least two “yes” answers are your candidates.
Then check those candidates against the data you already have: touchpoint ratings against NPS or retention, drop-offs in the journey, emotion in the verbatims. It doesn’t need to be a sophisticated model, a spreadsheet with correlations is enough.
And finally: pick three, five at most, and put your entire CX improvement budget for next quarter into those. Leave the rest alone. It sounds simple. It isn’t, because everyone in the organisation will want to add their favourite point to the list. Stick to the data, not to gut feeling.
