Restaurant KPIs: the six numbers to check every morning
Six numbers, one screen, five minutes. Last night sits in a line at the top of the dashboard: net sales before GST, orders, average bill. The cards under it are today as it builds, with net sales set against the same weekday last week, beside them what you were paid in, what sold most and what is still open.
Last night is one line at the top. Today is the cards under it.
Open the dashboard at nine in the morning and the first thing on the page is last night in a single line: Yesterday · Tue 8 Sep, with net sales before GST, the order count and the average bill. That is the read you came for and it takes ten seconds. Everything below it is the store's business today, which at nine is nearly empty and fills as the day goes.
The comparison lives on those cards. A restaurant's week has a shape: Tuesday lunch is quiet, Friday dinner is not, and the first weekend after salary day is different again. Measured against the day before, a perfectly normal Tuesday reads as a collapse and you spend the morning looking for a problem that does not exist. So the net sales card carries the same weekday last week beside it, printed as vs last Sat, and by the middle of service you can see whether today is tracking a normal one. For last night, hold the top line against the same weekday a week ago yourself.
Which night a bill belongs to matters here too. The trading day starts at 4am by default and the owner can move it between midnight and 6am in Settings, so a bill rung at 12:40am counts on the night it was earned instead of splitting one service across two dates.
Six numbers, and what each one is allowed to make you do.
Six is what an owner can read standing up, before the vegetables arrive. The first three are on the Yesterday line for last night and on a card for today; the rest are today's only. Each one answers a question, and each one is allowed to trigger a specific, small action. A number that is not allowed to change anything is a number you will stop reading by the end of the month.
| # | The number | What it answers | What it is allowed to make you do |
|---|---|---|---|
| 1 | Net sales: last night on the Yesterday line, and Net sales today (before GST) on the card, against the same weekday last week | Was that a normal night for this weekday? | Ask what was different: rain, a dug-up road, one cook missing, a party that did not come. Not a menu change, on one night's evidence. |
| 2 | Orders, on the Yesterday line and again as Orders today | Fewer guests, or the same guests spending less? | Read it with line 1. Sales down and orders flat is a ticket-size problem. Both down is a footfall problem, and they have different answers. |
| 3 | Average bill (net, before GST), on the Yesterday line and on a card | What one table is worth to you now. | A conversation about portions, add-ons and what the captain offers before the guest asks. A price rise needs more than one morning's number. |
| 4 | How you were paid today, by tender, GST included | Did the money arrive the way the bills say it did? | The card is today's, so take last night's split from the morning email: cash against what was counted, UPI against the bank credits. A gap gets a name today, not at the end of the month. |
| 5 | Top sellers, in units, over the last seven days | What the kitchen should be prepping for. | Set today's prep list, and mark the two or three dishes worth checking against what they cost you to make. |
| 6 | Open right now | Which checks were never settled. | One phone call to the floor while somebody still remembers the table. See below, because at 10am this should almost always be zero. |
A seventh line is worth a glance if you spend from the drawer: Expenses shows what went out today and this month, and it is the number that turns a busy month into a thin one without anybody noticing. The weekly close is where it gets dealt with properly.
Six, not thirteen. orderitnow.in, an Indian POS, publishes a 13-KPI daily list (read 2026-09-09) and it is a good list to grow into. Toast makes the point that decides how many you keep: a small set of decision-ready reports (read 2026-09-09) beats a dashboard nobody opens.
Average bill moves without anyone deciding to move it.
Average bill is net sales divided by orders, before GST, and it drifts on its own. Parcel share goes up and the average falls, because a parcel order is usually one or two dishes and no second round of drinks. A captain stops offering the full portion when the guest hesitates. A cheap dish becomes the thing everyone orders. None of that is a decision anybody took, and all of it shows up here first.
There is no honest number to compare it against nationally. The industry's own survey is NRAI's India Food Services Report, which is paid, and the public summaries carry market size rather than ticket size (checked 2026-09-09). Anyone quoting you an average bill for "Indian casual dining" is quoting a guess. So compare yourself to yourself: this month against last month, weekday lunch against weekday lunch. When it moves, the two places to look are what you charge and what gets offered, which is menu pricing and menu engineering, in that order.
How you were paid includes GST. Net sales does not. They will not match.
The two money cards answer different questions and are meant to disagree. How you were paid today is what actually arrived, split into cash, UPI and card, GST included, because that is what the guest handed over. Net sales today (before GST) is the sales figure you compare week on week, with the tax taken out. A standalone restaurant charges 5% GST without input tax credit (IndiaFilings, read 2026-09-09), so the collected figure is the larger of the two. Two numbers that differ by the tax you owe is the correct picture, and the GST bill post has the rest.
Both of those cards are today, so at nine in the morning they are nearly empty. Last night's split arrives in the morning email, and the longer view is the Sales tab inside Accounts.
The tender split is where a quiet problem announces itself. UPI carried 24,508.96 million transactions worth ₹29,82,355.95 crore in August 2026 alone (NPCI product statistics, read 2026-09-09), and RBI research puts cash at under 60% of consumer spending in 2024, down from around 80% in 2021, while still making up 50% to 60% of household expenditure (Data For India, RBI Bulletin October 2024). Restaurants sit cash-heavier than that. How much heavier we cannot tell you: two Indian POS vendors publish different restaurant cash shares with nothing behind either, so there is no figure here worth printing.
What is worth watching is your own swing. If cash share climbs on one person's shifts while UPI falls, or the UPI total on the report does not match the credits in your bank the next morning, the usual explanation is a personal QR code standing next to the till. The staff theft post covers that one, and the cash handling routine covers the drawer.
Open right now should be zero by mid-morning.
Open right now counts checks that were started and never settled. During service it is just your floor. At ten in the morning it is a list of last night's loose ends, and each one is either food that went out and was never billed, or a stray tap that nobody cleared. Both are worth one phone call while the memory is fresh, and neither is worth arguing about a week later.
The register clears them at the close rather than on the dashboard. An empty check gets discarded in one tap; a check with food on it is written off with a typed reason and a manager's PIN, so the write-off carries a name instead of being quietly zeroed at 1am. The day close guide is the button-by-button version, and the day-end report post is the eight lines to read at night.
If you never open a dashboard, read the morning email.
The summary lands in your inbox each morning and carries last night's money lines: orders, average order, cash, UPI, card, the total collected including GST, and net sales before GST. It is switched on per store in Settings, alongside the day-close cash-up mail that follows a closed shift on an hourly sweep (gross, refunds, net, opening float, expected, counted, variance). Between the two, an owner who is at the other shop all week still knows what happened at this one.
None of it depends on the shop having internet. The register bills, prints and settles on its own database and syncs when the line is back, which is what offline billing means in practice. The dashboard and the email need a connection, so on a bad day the numbers arrive late rather than wrong.
Tomorrow morning, whatever till you run: pick the same weekday last week, write both net sales figures on the same line, then the order count, then the tender split. Three minutes, on paper if you like. EasyKOT is ₹299 a month or ₹2,990 a year plus 18% GST, with three months free and no card, if you would rather the six numbers were waiting for you.
Straight answers.
If I only check one number, which one?
Why compare against the same weekday instead of yesterday?
What is a healthy average bill for an Indian restaurant?
Do I need internet to see these numbers?
Can I get this by email instead of opening a dashboard?
3 mahine chalao, phir paisa lagao.
Three months free, full product, no card. Made in India, for the counter: built inside a running restaurant.