Unibail-Rodamco-Westfield counted 1.9% more footfall across its shopping centres in 2025. Tenant sales in those same centres grew 3.9%, more than double the pace, so the crowd barely moved while the money did. That divergence is the quiet event now reorganising how stores measure themselves: footfall volume is being retired as a vanity metric, and the quality of each visit is taking its place. Quality here is concrete, meaning how long a shopper dwells, what that dwell says about intent, and whether it ends at the till. The change runs deeper than any dashboard, into how malls lease their space, how chains staff their floors, and what a good location is worth.

The first sign that volume lost its authority is that the count itself became something to correct. Sensormatic’s ShopperTrak now sells a feature it calls Employee Exclusion, which strips staff movements out of the door count so conversion can be measured against shoppers alone. A company does not build tooling to remove people from a number it still believes. The count misleads by place as much as by headcount: a door reader can log fifty thousand visits in a week while only a fraction of them reach any given display, as RetailNext reminds the brands now buying in-store media. Conversion is transactions over visitors, and the denominator had to be cleaned before the ratio meant anything. Counting was never the goal; it was the cheapest thing the first sensors knew how to do.

The metric that replaced raw traffic has a name, and it folds quality into a single figure. RetailNext calls it shopper yield: conversion rate multiplied by average transaction value, a number that climbs only when visitors both buy and spend more. Footfall does not enter the calculation. At the menswear chain Boggi Milano, shopper yield rose 40% over five years on RetailNext’s sensors, while the raw door count was treated as weather rather than a target. The change is subtle and complete: the store stopped asking how many came through and started asking what each arrival was worth. A metric you cannot inflate by propping the door open is one that changes what a store does.

Grading visits rewrites the staffing model before that logic reaches anything else. When a store tracks traffic by dwell rather than by time of day, it can place associates at the moments a salesperson turns a browser into a buyer, rather than spreading bodies evenly across opening hours. Boggi Milano cut labour costs by 15% even as conversion climbed, because the rota stopped being a calendar and became a forecast. The job itself is redefined, since a shift now exists to catch demand rather than to fill the room. Once conversion is the grade, labour follows the signal rather than the rota.

Leasing is where the new metric has the most money riding on it. The mall’s old promise to a tenant was footfall: we put bodies past your door, and you do the rest. Unibail-Rodamco-Westfield signed its 2025 leases at a 6.7% uplift on passing rents, in a portfolio whose footfall barely moved. The higher rent cannot be a reward for traffic; it is a charge for the quality of visit a location now delivers. Rent measured as a share of a tenant’s sales is the occupancy-cost ratio that landlords have always underwritten to. A good location stops being the busy one and becomes the one that converts.

The risk in grading visits is that a store learns to trust a number that looks like intent and is only motion.

That risk has a real basis, and naming it precisely is the only honest test of the shift. A head count is cheap, objective and comparable across every store and every year; dwell and intent are only inferences, and a long pause can mean desire, a queue, or confusion. If visit quality cannot be measured with enough signal to act on, footfall keeps its throne, because a blunt number you trust beats a rich one you cannot. That is the condition under which the whole move collapses. Instrumentation is the answer to it, since employee exclusion, zone-level paths and conversion tied to a basket now let a queue read as a queue and a sale read as a sale. The harder answer is that volume and quality have begun to move apart: Placer.ai recorded US mall traffic rising across every format in early 2025 even as the share of repeat visitors fell — a finding Placer.ai reads as new-customer growth, yet one that equally shows a growing crowd is no longer a proxy for a converting one.

The sensors have forced a choice that footfall let stores avoid. When the door count was the only number, every visit weighed the same, and a crowded store was a good one by definition. Grade the visit and that equivalence breaks: a quiet shop that converts can out-earn a thronged one that does not. From there the lease, the rota and the site search all follow the grade rather than the crowd. If that logic holds, the price falls on the location that was only ever busy — the transit-hub unit, the tourist strip, the store that sold proximity in place of persuasion — which will find that its footfall no longer spends. The store has not stopped watching the door; it has stopped believing what the door alone tells it.