Every emotional-loyalty deck contains a multiplier. Emotionally connected customers deliver up to 306 percent higher lifetime value than merely satisfied ones. Emotional connection drives up to eight times more visits and sales. The figures circulate widely, they are quoted confidently, and they are the reason most emotional-loyalty proposals die in the finance review.
They die because “up to” is doing enormous work, because the underlying studies were run on other categories with other definitions of connection, and because no CFO will book a 306 percent uplift against a program whose mechanism has not been demonstrated on this business. The instinct to reject them is correct.
That does not mean the effect is not real. It means the number has to be built locally, and it can be.
Why the borrowed multiplier fails
Three problems, all fatal in a budget conversation.
The construct is not standardised. “Emotionally connected” means one thing in a study using a proprietary connection index and something else in a study using top-box sentiment. Two figures quoted side by side are frequently measuring different things.
Direction of causation is assumed. Customers who visit more often plausibly feel more connected because they visit more often. A cross-sectional correlation between connection scores and spend does not separate the two, and the loyalty program is being asked to fund the causal claim.
Selection is unaddressed. The people who score highest on emotional connection are disproportionately the people who were already the heaviest users. Attributing their entire spend premium to the emotional dimension double-counts the behaviour that produced the score.
What a defensible local figure looks like
The measurable line item is not “emotional loyalty is worth X.” It is the spend difference between matched customers who differ on an emotional measure and are otherwise comparable. Building it takes four steps.
1. Pick one emotional measure and keep it. Brierley’s Loyalty Quotient approach is useful here because it explicitly separates rational loyalty from emotional loyalty rather than collapsing them into one score. Whatever instrument is chosen, the requirement is that it is fielded consistently and can be joined to individual customer records. A brand-tracker score with no customer-level join cannot produce a line item.
2. Match before comparing. Group customers by prior-period frequency and spend band first, then compare emotional-score tiers inside each band. This is the step that removes most of the inflation in the published multipliers. The premium that survives matching is the one worth quoting.
3. Measure forward, not backward. Score at time T, measure spend from T+1. A premium computed on the same period as the score is circular and will be recognised as such.
4. Express it in the unit finance already uses. Incremental margin per member per year, not a multiple. A multiple invites the question of what the base was; a per-member figure goes straight into a program P&L.
The gap is the actual opportunity
The most useful finding in the current research is not the size of the premium — it is that consumers consistently report that brands fall short on the emotional drivers they say matter most. There is a measured distance between what drives connection and what brands deliver against it.
That gap is more actionable than the multiplier, because it is diagnostic. It tells a specific operator which drivers are underperforming for its own customers, and those become the intervention. A program that raises performance on a driver customers rate as important and the brand rates as weak has a testable hypothesis attached. A program justified by a 306 percent claim has none.
Scale changes the arithmetic, not the method
Travel and hospitality runs the largest loyalty programs in existence — Marriott Bonvoy closed 2025 with 271 million members, Hilton Honors with 243 million. At that scale a small per-member premium is an enormous absolute number, which is precisely why the per-member framing matters more than the multiple.
It also means the measurement is affordable. A program with hundreds of millions of members can field an emotional measure on a sample that is statistically trivial as a proportion and still large in absolute terms. The constraint is never sample size. It is whether the score can be joined to transaction records, and whether anyone committed to fielding the same instrument twice.
The one-sentence version
Stop quoting the industry multiplier and start reporting the matched, forward-looking, per-member margin difference on a single consistently-fielded emotional measure. It will be a smaller number than the decks claim, and it will survive the meeting.
