Crowd & occupancy
Eyes on the room
How much of each room you can currently see — the number of rostered people who have posted at least one scan since tonight’s doors opened. This is coverage, not a staffing-level judgment.
What it is
Eyes on the room is a coverage figure: per venue, how many of your rostered staff have posted at least one scan since tonight's doors opened, alongside the number of scans that team has posted. It answers a single, literal question — how much of this room can we currently see through the people scheduled to look at it.
It is deliberately narrow. It is not a headcount of who is working, not a measure of how busy the night is, and not a judgment about whether the staffing level was right. It is the reporting fact underneath tonight's crowd read: the more rostered staff actually generating scans, the more of the room the rest of your numbers are drawn from.
How to use it at your venue
Operators tend to read this as a data-trust signal rather than a staffing verdict — a quick check on how much confidence tonight's crowd read deserves before leaning on it. Survey-measurement work draws exactly this line: whether a unit was observed at all (coverage) is a separate error category from whether what you measured is accurate or sufficient, and the two are not interchangeable.[1] A low number here means you are seeing less of the room, so it is a grounded prompt to treat the night's other figures as thinner evidence, not to conclude the floor was short.
A grounded next step, when you do want to ask about staffing adequacy, is to pair this with a demand or occupancy signal rather than reading it alone. Hospitality labor-scheduling frameworks separate "who reported for duty" from "was that count adequate for the demand," treating them as distinct steps answered by different data;[3] and staffing-measurement research outside nightlife has shown empirically that a headcount- or hours-based figure does not reliably track adequacy — the two correlate only weakly and capture different things.[2] So operators often use this to decide how much to trust a read, and use a demand signal beside it to ask whether they were actually staffed for the crowd.
How to read it
Read it against your own roster for that venue, and read it as presence, not sufficiency. A higher number means more of the room is being observed tonight; a lower one means you can see less of it. It resets each night, so it describes tonight only. Because an owner's own scan does not count as a team scan, this reflects the team actually reporting for duty, not you standing in for them.
What it doesn't mean
⚠ What it can't tell you
This is coverage, not staffing. Someone can be on shift all night and never scan, so a low number means we can see less of the room — not that the room is understaffed. It does not say whether the count was right for the crowd, does not measure how busy the night was, and is never compared against another venue's roster. A high number is not proof of good staffing, only of good observation.
How it's calculated
Counted per venue and reset each night, entirely from scans actually posted. When a rostered staff member posts their first scan of the night through their staff link, they count toward the coverage number; the scans that team posts are tallied alongside it. It is compared only against your own roster for that venue. No sales data and no crowd headcounts feed it — it reads the record of who scanned, nothing more.
The exact method
Both figures are drawn from per-venue staffState, which is written only by the staffScan callable — so a staff member is counted only once they post via a staff link, never from being on the schedule. scanning is the number of rostered staff who have posted at least one scan since doors opened tonight; teamScans is how many scans they posted. The count resets each night and is directly observed (no estimate, no band). An owner's own scan is excluded from the team-scan count. Inputs are the roster size, the scanning count, the team-scan count, and the live crowd band; there is no POS input.
The research behind it
This number answers "did the people scheduled tonight actually show up and check in enough to generate data" — a presence/reporting fact. It deliberately does not answer "was that the right number of staff for how busy it got." Research on measurement (survey science) and on staffing (hospitals, hospitality labor scheduling) treats those as two different questions on purpose — mixing them up is a documented measurement mistake, not a subtlety RollCall invented.
-
Total Survey Error: Past, Present, and Future ✓ verified
Read the source ↗ -
Hospital Nurse Staffing: Choice of Measure Matters ✓ verified
Read the source ↗ -
Workforce Scheduling: A Guide for the Hospitality Industry (Cornell Hospitality Report, Vol. 4, No. 6) ✓ verified
Read the source ↗
Confidence & caveats
The count itself is exact — it is a direct tally of scans that were actually posted, with no estimate or confidence band. What it is sure of is narrow on purpose: it is sure how many rostered staff reported by scanning, and nothing more. It cannot see a staff member who is present but not scanning, so "no reading" from a person is an absence of data, not a zero on the floor. It carries no adequacy verdict and no cross-venue comparison, and treating a coverage number as a staffing-level judgment is a documented measurement mistake, not a subtlety of this product.