Crowd & occupancy
Compare your rooms
Every venue’s typical night drawn on one shared scale, so the eye compares shape rather than size. Venues without enough measured nights are named, not drawn.
What it is
Compare your rooms is a small-multiples grid: every venue's typical same-weekday crowd curve drawn as its own little panel, and all the panels forced onto one shared scale — the same time-of-night window across the bottom, the same crowd axis up the side. Beside the grid sits a short playbook of deals proven at one venue and never tried at another.
The shared scale is the whole idea. When repeated panels share identical axes, your eye does the comparison for you and reads shape — where a room fills, peaks and empties — without confusing it with scale, the raw size of the crowd. A 90-capacity bar and a 400-capacity club can sit in the same grid and still be honestly comparable, because you are comparing the arc of the night, not the height of the numbers.
How to use it at your venue
Operators often use the grid to tell a shape problem apart from a scale one — the distinction the shared axes are built to expose.[1] A slow room whose curve peaks early and dies has a programming or timing problem; a room with the same arc as a strong venue, just lower, has a scale problem — marketing or capacity, not pacing. Those point at different fixes, and side-by-side raw crowd numbers would hide the difference, since crowd patterns genuinely differ in both level and shape from room to room.[2]
The playbook beside the grid is the second, blunter use: a lever that measurably worked at one venue and was never tried at a comparable one is a low-cost thing to test. A grounded way to read the whole card is as a question generator — which room's shape is unlike the others, and which proven move hasn't crossed the street yet — rather than a ranking that tells you which venue is winning.
How to read it
Read shape, not height. Each curve is that venue measured against its own same-weekday history, so a tall panel is a bigger room, not a better one, and a short panel is a smaller room, not a failing one. The comparison the grid earns is across the arcs — when rooms fill and empty relative to each other — not across their peaks. Treat a curve that looks unlike its neighbours as something to explain, not a score to act on.
What it doesn't mean
⚠ What it can't tell you
There is no group total and no ranking here: venues are never scored against each other, only drawn side by side so shapes can be compared. The panels carry no dollars — this is built entirely from scan history, with no POS anywhere in it, so a taller crowd curve says nothing about money earned. And a venue only appears once it has enough measured nights; thin venues are named rather than drawn, because a curve built on too few nights would look authoritative while meaning nothing. A named-but-not-drawn venue is a coverage state, not a poor result.
How it's calculated
Built client-side from each venue's own scan history — nothing on a server computes or freezes this. For a chosen weekday, every venue's typical crowd curve is placed on one shared time window so the panels line up, and a line is drawn only for venues with enough measured nights on that weekday. Venues below the floor are listed by name instead of being drawn. The proven-levers playbook is pulled from each venue's own deal history — moves that measurably worked. No figure here is added into any total.
The exact method
The grid is produced by smallMultiples(rows, wd, focusIds). A single x-window is shared across all venues so the panels are on identical axes; a per-venue polyline is drawn where nights ≥ 2, with each curve's peak taken from rcTypicalPeak. The reliability floor is RC_PEAK_FLOOR = 4 measured same-weekday nights — below it a venue is named rather than drawn. Inputs are per-venue scan history for the curves plus deal/lever history for the playbook, whose proven levers come from rcDealLevers.proven (rendered ). There is no POS input. The suppression floor is a data-honesty convention chosen for sample reliability, not a threshold drawn from any published standard.
The research behind it
When you put a bunch of charts next to each other and force them onto the same scale, your eye naturally does the comparison for you — that's a well-established data-visualization rule (Tufte's "small multiples"), not something we invented. Separately, the hotel industry has been comparing properties against peers on an indexed, apples-to-apples scale for decades (STR's comp-set reports) because raw occupancy numbers alone are misleading — a Cornell study confirms occupancy patterns genuinely differ shape-by-shape and market-by-market, so you can't just eyeball two numbers. Restaurants have their own version of the same idea (same-store sales comparisons). RollCall's "compare your rooms" feature applies both principles to nightlife: same shared scale so shapes are honestly comparable, and it borrows the peer-benchmarking instinct from hotel/restaurant industry practice. The "name thin venues instead of drawing them" rule is our own data-honesty convention for low-sample venues, not something pulled from a paper — flagged as unverified-by-literature.
-
Envisioning Information (Chapter 4: "Small Multiples") ✓ verified
Read the source ↗ -
Natural Occupancy Shifts in Hotel Markets ✓ verified
Read the source ↗
Confidence & caveats
The shared-scale grid rests on a well-established data-visualization rule — repeated panels on identical axes are the correct way to make shape comparable across units[1] — and on the documented fact that crowd patterns differ in shape and level from place to place, which is why raw side-by-side numbers mislead.[2] That second source studied hotel markets over time rather than one venue against another, so the leap to comparing individual rooms is our extrapolation, not its finding. The "name thin venues rather than draw them" rule is our own honesty convention for low-sample rooms, not something a paper validates. And every curve is a description of past same-weekday nights — it is not a forecast, and it never says why a room's shape looks the way it does.