Money & sales
Discounts — share of gross sales
Discounts as a percentage of the gross sales those same orders rang up, on the latest completed night, all channels. Not a refund rate and not the leakage metric — refunds are deliberately not shown here.
What it is
Discounts — share of gross sales is how much of a venue's gross sales came back off as discounts on its most recent completed night. It is the night's discounts divided by the gross the same orders rang up — item prices before any discount was taken off — shown as a percentage, across every sales channel the venue rang.
Gross is the right denominator here for an accounting reason, not a stylistic one: under the restaurant industry's standard chart of accounts, discounts, comps and promotional allowances are recorded as a deduction from gross sales — contra-revenue — rather than as an expense line.[1] So this is the share of what the room rang that never reached net, expressed against the same cohort of sales it was taken from.
How to use it at your venue
Operators often watch this the way they watch food cost or labour as a percentage: a line that should sit at a level they have chosen on purpose, not one that drifts. A grounded use is reading a venue against its own recent nights — when the share spikes on a night with no planned promotion, that usually points at a comp-happy shift or a POS default worth checking rather than a marketing win.
When the share moves because of a deliberate promotion, the revenue-management literature's central question is whether the discount brought in trade that would not have shown up otherwise, or mostly gave margin back to guests who would have paid full price — the incremental-versus-replacement distinction.[2] How a discount is structured and framed also shapes how fairly guests perceive it, which is part of reading whether a given level is a problem rather than just that it changed.[3] Because the figure covers all channels and one exact cohort of orders, comparing it across nights, dayparts or staff can surface where discounting concentrates.
How to read it
Higher means more of what the room rang was given back as discounts. It is not colour-coded good or bad, and it should not be read as either. Comps, promotions, house policy and format legitimately differ between rooms, so read the trend within one venue against the level you meant to run — a jump you did not plan is the signal, not the absolute number.
What it doesn't mean
⚠ What it can't tell you
The ratio is exact, but it cannot carry intent: a comp for a complaint, a promotion, a staff meal and a loyalty reward are all discounts here and are not separated. It cannot say whether a share is too high or too low — no threshold for a "normal" share is defined, because comps and house policy legitimately differ between rooms.
Refunds are deliberately not shown here: a night's refunds routinely settle sales from earlier nights, so dividing them by this night's figures would not be a rate for this night's sales — refund activity lives on the venue's own Sales page. This is also not the venue leakage metric, which divides by occupied guest-hours and stays gated until calibration. And a feed that does not report discounts says nothing, rather than 0%.
How it's calculated
Both figures come from the completed night the sales comparison above already loaded, so nothing extra is read. The discounts and the gross are summed over exactly the same orders, in the same revisions, for the same business night, so the numerator and the denominator describe one cohort of sales. Only completed, settled nights with a positive gross produce a value; a night that reports no discounts is left blank, never shown as 0%. Values are never added into a group total.
The exact method
The numerator is the night's total discounts; the denominator is gross sales — item prices before any discount was taken off — both taken in minor units (discountMinor, grossMinor) from the same orders, same revisions, same business night. The result is rounded to one decimal place. Unlike the other comparisons on this card, which leave POS-classified off-premise orders out, both halves of this ratio cover every channel the venue rang.
A night only produces a value if its gross is positive and its discount figure is present and non-negative: non-positive gross and signed or absent discounts are skipped, with no fallback to net sales and no substitution of 0% for a missing discount figure. The figure is computed client-side over the already-loaded completed night (discountShareFromDocs); the money components are produced server-side by moneySummary. No door scans are involved.
The research behind it
Every dollar you discount is a dollar that never hits the top line — so if you don't track discounts as a percentage of what you actually sold, you can't tell the difference between a smart promotion and a slow bleed of free stuff walking out the door.
-
Uniform System of Accounts for Restaurants: A Guide to Standardized Restaurant Accounting, Financial Controls, Record Keeping and Relevant Tax Matters (8th Edition) ✓ verified
Read the source ↗ -
A framework and model to evaluate promotions: A restaurant cross-promotion in-market study ✓ verified
Read the source ↗ -
Has Revenue Management Become Acceptable? Findings from an International Study on the Perceived Fairness of Rate Fences ✓ verified
Read the source ↗
Confidence & caveats
The ratio itself is exact for the night — one cohort of orders divided by its own gross, so there is no estimate and no band. But "exact" is not "judged": the figure carries no verdict, no target, no benchmark and no group total, and it is never compared to another venue — the comparison it does show carries no currency and no room size.
A blank is a state, not a zero: when a feed reports no discounts, the metric says nothing rather than claiming the venue discounted nothing. And the USAR basis it rests on documents the contra-revenue treatment as standard practice, but the same framework also allows comps to be booked as a marketing expense in some contexts, so "discounts ÷ gross" is the correct convention for this reading rather than the only accounting treatment that exists.[1]