Masterestaurant Occupancy-by-Daypart Index 2026: The Hourly Map of the Urban Restaurant

Finding 1 — The afternoon valley is the brand storefront no owner is using

A restaurant that expands into retail and branded product turns the cold seat at 4:30 pm into a point of sale, not a sunk cost. Across the 8,400 audited accounts, the 3-6 pm window drops to 19% average occupancy while rent runs the full 24 hours. Diego F. Parra says it in every Masterestaurant engagement: those empty tables already cost the same as the full ones at the peak. The difference is what you do with them. A house-brand specialty coffee, a jar of signature sauce or a logo-stamped chocolate bar turns the after-meal lull into extra ticket. In the 47 locations that tested this lever, the valley added between 8% and 14% in new sales without touching the peak or hiring more kitchen staff. The storefront already exists; it was just missing merchandise to sell in it. 38% of an urban restaurant's lost margin lives between 3 pm and 6 pm, not at the rush hour.

Finding 2 — Measuring by day hides the valley; measuring by window exposes it

When you average a whole day's occupancy into a single figure —say 52%— the number lies by omission: it blends a lunch peak at 71% with an afternoon valley at 19%. The Masterestaurant team measured seats occupied against available capacity in 30-minute windows over three years, and the pattern repeats across the audited Spanish-speaking cities. The owner who only reads the daily close believes the house performs well; the one who cuts the day into windows discovers it bleeds for four straight hours. That's the mistake I see over and over: money goes into filling an already-full peak when the virgin ground —and near-pure margin— sits in the dead afternoon nobody audits window by window. Raising lunch occupancy from 71% to 78% costs expensive marketing and brushes the room's physical limit. The average peak across the 8,400 accounts settles at 71% of capacity, and those seven missing points to the real ceiling demand more advertising, more discount and more forced table turns for diminishing marginal return.

Finding 3 — The peak already hit its physical ceiling: fighting for seven more points is expensive

The valley, by contrast, starts at 19%: every point you gain there is near-clean margin because rent, power and the manager are already paid. At Masterestaurant we run the cold math with the owner: seven points at the saturated peak equal, in real money, less than half what five points in the valley yield. The arithmetic is counterintuitive but brutal. Stop pushing against the ceiling and start filling the basement; that's where the money nobody picks up sits. The peak's lever is table turnover; the valley's is CX and suggestive selling with branded product. Confusing them burns money. At rush hour —lunch at 71%, dinner above 60%— every minute of an occupied table matters: the goal is to serve fast, cash out and free the seat for the next diner. There, retail branding gets in the way because it slows the turn. But in the 3-6 pm window, with 19% of seats filled, hurry is your enemy: you want the customer to stay, order a second specialty coffee and take home a jar of your signature preserve.

Finding 4 — Each window has its lever: don't use the peak's hammer in the valley

Diego F. Parra frames it this way in board meetings: at the peak you sell time, in the valley you sell experience and merchandise. Applying the turnover strategy to the dead afternoon scares off the one customer who actually had room to buy more. This index exists so an owner knows exactly which occupancy percentile their restaurant falls in and which lever to pull in each window. With 8,400 audited accounts we built the real curve: the 50th percentile at the lunch peak is 71%, but the 25th starts at 58% and the 90th reaches 82%. In the afternoon valley the median is 19%, and the top quartile —those already working the window— touches 34%, nearly double. The gap between your house and the 90th percentile isn't luck: it's a list of concrete levers. If your peak sits in the 40th percentile, the problem is table turnover.

Finding 5 — Which occupancy percentile your house falls in

If your valley sits at or below the median, the lost money is in CX and branded product. Masterestaurant built this map as primary research, not a summary of other people's figures, so you stop comparing your restaurant against a blind average. Expanding into branded product turns idle capacity into revenue without cannibalizing the peak service. The typical objection I hear is «I don't have kitchen capacity to make sauces and preserves». False: the kitchen running at 71% during lunch falls to 19% use in the afternoon, and those same four dead hours are when you jar, label and prep the branded merchandise. In the 47 locations that tested it, the marginal production cost was low because they used labor already hired and present during the valley shift. The product then sells in the very 4:30 pm storefront, by delivery or in local shops. A jar that costs 2.10 USD to produce sells at the table for 9 USD; no menu dish gives you that margin.

Finding 6 — Branded retail doesn't compete with the kitchen: it uses it when idle

The key is that the restaurant's brand already carries trust with the diner: selling its product is the natural extension, not a new business from scratch. Real occupancy is measured in hot seats over cold seats, check by check, not in reservations or intent. For three years the Masterestaurant team recorded seats occupied against available capacity in 30-minute windows, and that discipline changed the diagnosis of dozens of restaurants. The owner usually looks at the day's reservations and believes the house is packed; reservations lie because they don't discount the no-show or the party of four that showed up as two. What matters is the physical seat filled against the physical seat empty, window by window. When you put that raw figure on the table —71% at the peak, 19% in the valley— the conversation stops being about marketing and turns to what you do with the afternoon's 81% of cold seats.

Finding 7 — Hot seats against cold seats: the metric nobody keeps

That's where branded product stops being a whim and becomes the obvious answer to the margin that was being thrown away. Measuring by day hides the valley; measuring by daypart exposes it: 38% of lost margin lives between 3:00 and 6:00 PM, not at the lunch rush. The peak is at its physical ceiling (71% average); fighting for 7 more points costs dearly. The valley sits at 19%: every point gained there is nearly pure margin because the rent is already paid. The peak lever is table turnover; the valley lever is CX and suggestive selling —specialty coffee, lingering with branded product, an afternoon format—. The restaurant that expands into retail and branded product turns the valley into a showcase: the cold 4:30 PM seat becomes a point of sale for its brand, not a sunk cost.