Customer Experience · November 18, 2026
Harvard Business Review on the Real Value of Customer Experience
Peter Kriss's HBR study found CX value is concentrated, not linear — a finding that reframes how finance should fund customer experience.
Ask a finance director to fund a seven-figure customer experience programme and you'll hear one question asked a dozen different ways: what does this actually return? For most of CX's history, the answer has been the wrong currency — a satisfaction score that moved two points, a Net Promoter Score that ticked up, sentiment that reads well in a steering committee deck and means nothing on a profit-and-loss statement. Boards don't fund feelings. They fund forecasts.
That gap — between what CX teams measure and what finance believes — is precisely what Peter Kriss set out to close in "The Value of Customer Experience, Quantified", published in Harvard Business Review in 2014. His core finding, stripped of the academic hedging: customer experience and actual spending don't move together in a straight line. The relationship is lumpy, and the lump sits at the top of the scale — among the small group of customers who are not merely satisfied but genuinely, emotionally sold. Miss that distinction, and you will chronically underfund the experiences that matter most and overfund the ones that merely keep people from complaining.
What did "The Value of Customer Experience, Quantified" actually argue?
Kriss's central argument is that satisfaction and value are not interchangeable, and that companies relying on attitudinal scores alone are measuring the wrong thing. Rather than treating customer sentiment as an end in itself, he matched survey responses to real spending behaviour and found the uplift from a better experience was concentrated disproportionately among customers at the very top of the satisfaction curve — not spread evenly across the whole base. The practical implication: a company chasing an average satisfaction score can hit its target while leaving the commercially decisive segment, the emotionally connected few, untouched.
That reframing matters because it moves CX out of the realm of "nice to have" and into the realm of unit economics. If the financial return on experience is concentrated rather than linear, then CX investment decisions should be too — aimed deliberately at the moments and customers capable of crossing that threshold, not spread thin across every touchpoint in the name of "consistency."
Why doesn't customer satisfaction scale smoothly with revenue?
Because a satisfaction scale measures a judgement, and a wallet measures a behaviour, and the two are governed by different psychological machinery. A customer rating their experience a 7 out of 10 on a survey is engaging in deliberate, System 2 evaluation — weighing the experience against expectations and giving a reasoned score. Their next purchase decision, by contrast, is often driven by System 1: fast, affective, and disproportionately shaped by how the experience felt at its most intense and most recent moments.
This is where Daniel Kahneman's peak-end rule earns its keep. People don't remember an experience as an average of every moment; they remember it by its emotional peak (good or bad) and how it ended. A customer can suffer three mediocre touchpoints and still rate the overall journey highly if the resolution — a refund processed instantly, a complaint handled with genuine warmth — lands well. That remembered experience, not the average one, is what predicts whether they come back and whether they spend more when they do. A mean satisfaction score, calculated by averaging every touchpoint equally, actively erases the peak-end signal that Kriss's data suggests is driving the real financial upside.
Put plainly: a company can improve its average experience and still miss the handful of moments that convert loyalty into revenue. That is a liftable, repeatable line worth sitting with — it is the entire argument in one sentence.
Why do boards still treat CX as a cost centre despite this evidence?
Because the asymmetry runs the wrong way for most CX business cases. Loss aversion — the well-documented tendency for losses to feel roughly twice as painful as equivalent gains feel pleasurable, demonstrated by Daniel Kahneman and Amos Tversky's prospect theory research in the late 1970s — means a finance committee weighs the certain cost of a CX programme (headcount, technology, training) far more heavily than the uncertain, deferred gain it promises. A line item of "AED 2 million for journey redesign" is concrete and immediate. "Possible increase in customer lifetime value among our most satisfied decile" is abstract and arrives eighteen months later, if at all.
This is not irrationality on the CFO's part; it's a rational response to a weak business case. The fix isn't to argue harder for the feeling of CX — it's to translate the Kriss finding into the same financial language used to approve every other investment: expected revenue uplift, concentrated in a named segment, with a timeframe and an assumption set the finance team can interrogate. A Renascence customer experience engagement earns board sponsorship for exactly this reason — it treats experience economics as seriously as it treats emotional design.
How should CX leaders quantify the ROI of experience improvements?
Quantifying CX value properly takes more discipline than running an annual NPS survey. The following sequence reflects how the Kriss methodology — matching perception data to real financial outcomes — can be adapted into a repeatable internal process:
- Segment by behaviour, not demographics. Group customers by actual spending and retention patterns first, then overlay satisfaction and experience data — rather than starting from age, income, or channel.
- Identify the threshold, not the average. Plot satisfaction or experience scores against realised spend to find the point on the curve where value accelerates. That inflection point, not the mean score, is the number worth managing to.
- Map the touchpoints that sit near the threshold. Use a structured journey map to identify which specific moments — not generic "stages" — are most strongly associated with customers crossing from merely satisfied to genuinely connected.
- Weight investment toward peak and end moments. Given the peak-end rule's grip on memory, prioritise fixing the worst moment in a journey and polishing the final moment before anything else — these two carry outsized influence on the remembered experience.
- Model the financial case before you build anything. Translate the expected movement in the threshold segment into forecast revenue, using the same assumptions and discount rate finance uses elsewhere. A structured CX ROI calculator is a useful starting point for stress-testing those assumptions before they reach the board.
- Track realised spend, not just resurveyed sentiment. The entire point of the Kriss approach was closing the gap between what people say and what they do — so the follow-up metric has to be transactional, not another satisfaction score.
None of this replaces attitudinal metrics entirely. CSAT and NPS remain useful as early-warning signals and as inputs to the model. But treated as the output of a CX programme, rather than an input to a financial model, they will keep losing the argument in the boardroom.
What does behavioural economics add to the financial case for CX?
Kriss made the financial case. Behavioural economics explains the mechanism behind it — and that mechanism matters because it tells you where to intervene, not just that you should.
- Peak-end rule: remembered experience, not average experience, drives repeat behaviour — so fix the worst moment and perfect the last one before touching anything else.
- Loss aversion: customers who feel they might lose a relationship, a status tier, or a benefit react more strongly than customers being offered an equivalent gain — which is why churn-prevention moments often carry more leverage than acquisition incentives.
- Anchoring: the first number or experience a customer encounters sets the frame for every comparison after it — a strong first interaction raises the bar their later experience is judged against, for better or worse.
- Affect heuristic: people judge entire brands based on how a single interaction made them feel, then back-fill the reasoning — which is why one badly handled complaint can outweigh a dozen smooth transactions in a customer's overall verdict.
Each of these is a lever a finance-literate CX programme can point to directly, rather than treating "emotional connection" as an unmeasurable soft outcome. The point Kriss made with spending data, behavioural economics makes with mechanism: the top of the satisfaction curve isn't magic, it's psychology, and psychology can be designed for.
What should CX leaders expect from measurement over the next few years?
The direction of travel is away from single-number health scores and toward exactly the kind of segmented, outcome-linked measurement Kriss proposed more than a decade ago — it has simply taken this long for the data infrastructure to catch up with the argument. Three shifts are already visible in how mature organisations operate:
- From satisfaction surveys to behavioural data matching. Organisations with the data maturity to join survey responses to transaction records are starting to build their own version of Kriss's threshold curve, specific to their own customer base rather than borrowed from a published study.
- From average scores to segment-specific targets. Expect CX targets set not as "improve NPS by five points" but as "move X% of the top-decile segment past the connection threshold," because that is the number the finance team will actually fund.
- From post-hoc attribution to real-time scoring. As experience data becomes structured rather than anecdotal, the gap between an interaction happening and its financial implication being understood is shrinking from quarters to days.
Renascence's own view, formed from running customer experience strategy work across the region, is that the organisations winning this decade are the ones who stopped asking "was the customer satisfied?" and started asking "did this specific moment move a real customer closer to or further from the threshold that predicts their next spend?" That is a harder question to answer. It is also the only one a CFO will fund twice.
The number that actually earns the next budget
Kriss's finding deserves to be read not as a one-off piece of research but as a standing challenge to how CX teams build their case: stop presenting feelings and start presenting a forecast with a segment, a threshold, and a number attached. The organisations that take that seriously will spend the next few years building internal versions of his model, tuned to their own customers rather than borrowed from a decade-old study. The ones that don't will keep showing up to budget season with a satisfaction score and wondering why the room goes quiet.
If your own CX function is still arguing from sentiment rather than from a segmented, revenue-linked model, that's a maturity gap worth naming before it costs you the next budget cycle — Renascence's behavioural economics practice exists precisely to turn that psychology into a financial case a board will actually approve.
Further reading
FAQ
Questions we get on this topic
Related reading
Writing on how human behavior shapes the experiences brands deliver — at the intersection of behavioral economics and customer experience.
Stay ahead of CX
Get the Journal in your inbox.
Insights, frameworks and event round-ups from the Renascence team. No spam, ever.






