· Orqestra Team · 8 min read

Restaurant KPI Dashboard Template: 11 Weekly Metrics

A restaurant KPI dashboard template for small restaurants: the 11 numbers to track weekly, the formula for each, benchmark targets and a filled-in example week.

A restaurant KPI dashboard template is a one-page sheet with 10 to 12 numbers, each with a formula, this week's value, last week's value, and a target. For a small restaurant the list is: net sales, sales vs last year, average ticket, covers, food cost %, labor cost %, prime cost %, sales per labor hour, table turns, comps and voids %, and online order share. Fill it in once a week, on the same day, and most operating problems show up as a number moving the wrong way before they show up on the bank statement.

The sheet below is available as the restaurant KPI dashboard template. This article gives the formula for each KPI, which targets are backed by published benchmarks, and one full week filled in for a single-location restaurant.

The 11 KPIs and their formulas

Every number on the dashboard comes from three sources you already have: the POS sales report, the payroll or scheduling system, and the weekly inventory count. Nothing requires a new system to start.

KPIFormulaSource
Net salesGross sales − discounts − refunds (before tax and tips)POS
Sales vs last year(Net sales this week − net sales same week last year) ÷ net sales same week last year × 100POS
Average ticketNet sales ÷ number of paid ordersPOS
CoversNumber of guests servedPOS or host stand
Food cost %(Beginning inventory + purchases − ending inventory) ÷ food sales × 100Inventory count + invoices
Labor cost %Total labor cost (wages, payroll taxes, benefits) ÷ net sales × 100Payroll
Prime cost %Food cost % + labor cost % (or (COGS + labor) ÷ net sales × 100)Both
Sales per labor hourNet sales ÷ total labor hours workedPOS + schedule
Table turnsParties seated ÷ (tables × days open)Host stand or POS
Comps and voids %(Comps + voids + manager discounts) ÷ gross sales × 100POS
Online order shareOnline order net sales ÷ net sales × 100POS or ordering platform

A few definitions matter:

  • Use net sales, not gross, for every ratio. Clover's own Sales Overview defines average ticket size as net sales divided by paid orders, and its net sales figure excludes tips, taxes and service charges (Clover Help). Use the same convention so your numbers match the POS.
  • Food cost % needs a weekly inventory count. Without one you only have purchases, which swing with delivery timing. The food cost percentage guide covers the count-based method.
  • Labor cost should include payroll taxes and any benefits, not just hourly wages. If you only include wages, your labor % will read two to four points low.

Which targets are backed by benchmarks

Only three of the eleven have published benchmarks that two or more independent sources agree on. Put a target in the dashboard for those. For the rest, your target is your own trailing average, improved a little each quarter.

KPIBenchmarkSources
Food cost %25% to 35% of sales for most restaurantsComplete Controller, Sculpture Hospitality
Labor cost %25% quick service, 25% to 30% casual dining, 30% to 35% fine dining; 7shifts reports two-thirds of restaurants land between 20% and 30%Restaurant365, 7shifts
Prime cost %About 60%; 60% to 65% full service, 55% to 60% quick service; 65% and above is a warning signRestaurant365, Complete Controller

Prime cost is the one to watch hardest, because it combines the two costs you control week to week. The prime cost guide explains why 60% is the usual line and what to do when you are above it.

For average ticket, covers, sales per labor hour, table turns, comps and voids, and online share, there is no published range that applies across concepts. A $14 average ticket is excellent for a taqueria and a problem for a steakhouse. Set the target from your own last 8 to 12 weeks and move it as you make changes. The average ticket guide shows how to set that one.

The dashboard layout

The template has five columns. Keep it to one page. If it needs scrolling, it will stop being read.

KPIFormulaThis weekLast weekTarget
Net salesGross − discounts − refunds
Sales vs last year% change vs same week LY
Average ticketNet sales ÷ paid orders
CoversGuests served
Food cost %Usage ÷ food sales
Labor cost %Labor ÷ net sales
Prime cost %Food % + labor %
Sales per labor hourNet sales ÷ labor hours
Table turnsParties ÷ (tables × days)
Comps and voids %(Comps + voids) ÷ gross sales
Online order shareOnline net sales ÷ net sales

Add a sixth column for "same week last year" if you have a year of history. Add a notes line under the table for anything that explains a swing: weather, a catering order, a staff call-out, a road closure.

How to fill it in each week

  1. Pick the week boundary your POS uses, usually Monday to Sunday, and keep it.
  2. On Monday morning, pull the sales report for the week: gross sales, discounts, refunds, paid orders, comps and voids, and online orders.
  3. Take the inventory count before the first delivery and total the ending inventory. Pull last week's ending inventory and the week's food invoices.
  4. Pull total labor hours and total labor cost from payroll or the scheduling system, including payroll taxes.
  5. Count parties seated from the host stand or the POS table report. Count the tables you actually had in service.
  6. Compute each row with the formulas above. Move last week's "this week" column into "last week".
  7. Circle anything more than two points or 10% off target and write one line about why in the notes.

Steps 2 through 5 take about 30 minutes by hand. A sales analytics tool that reads your POS, such as the sales analytics feature in Orqestra, fills in the POS rows automatically, and the labor and inventory rows are the only ones left to type.

Worked example: one week filled in

A single-location full-service restaurant, 24 tables, open six days, doing about $107,000 a month. The week is Monday October 19 to Sunday October 25, 2026.

Inputs from the POS and payroll:

  • Gross sales $25,640, discounts $468, refunds $372, net sales $24,800
  • Paid orders 1,240, guests 1,860, parties seated 860
  • Same week last year net sales $23,100
  • Online orders net sales $5,208
  • Comps and voids $372
  • Beginning inventory $4,860, purchases $6,240, ending inventory $4,920, food sales $20,400 (the rest is beverage)
  • Labor cost $7,440 including payroll taxes, labor hours 520

The calculations:

  • Sales vs last year = ($24,800 − $23,100) ÷ $23,100 × 100 = 7.4%
  • Average ticket = $24,800 ÷ 1,240 = $20.00
  • Food cost % = ($4,860 + $6,240 − $4,920) ÷ $20,400 × 100 = $6,180 ÷ $20,400 = 30.3%
  • Labor cost % = $7,440 ÷ $24,800 × 100 = 30.0%
  • Prime cost % = 30.3% + 30.0% = 60.3%
  • Sales per labor hour = $24,800 ÷ 520 = $47.69
  • Table turns = 860 ÷ (24 × 6) = 860 ÷ 144 = 6.0 turns per table per day
  • Comps and voids % = $372 ÷ $25,640 × 100 = 1.5%
  • Online order share = $5,208 ÷ $24,800 × 100 = 21.0%

The filled-in dashboard:

KPIThis weekLast weekTarget
Net sales$24,800$23,900$24,000
Sales vs last year+7.4%+4.1%+5.0%
Average ticket$20.00$19.60$20.00
Covers1,8601,8101,800
Food cost %30.3%32.4%30.0%
Labor cost %30.0%31.2%29.0%
Prime cost %60.3%63.6%60.0%
Sales per labor hour$47.69$45.10$48.00
Table turns6.05.86.0
Comps and voids %1.5%2.1%1.5%
Online order share21.0%19.5%20.0%

Reading it: a good week. Prime cost dropped 3.3 points, mostly because food cost came back from 32.4% after a short-ordering fix. Labor is still a point above target even though sales were up, so the schedule did not flex with the volume. Online share passed 20% for the first time, which is worth a note because online orders usually carry a lower average ticket and change the labor mix in the kitchen.

Common mistakes

  • Mixing gross and net. Ratios computed on gross sales read lower than the benchmarks, which are all on net. Pick net and stay there.
  • Using purchases instead of usage for food cost. One big protein delivery on Friday makes a 28% week look like 36%. Count inventory.
  • Leaving payroll taxes out of labor. Your labor % will look two to four points better than it is.
  • Tracking monthly only. A bad week inside a good month is invisible. Weekly catches the problem while the cause is still fresh.
  • Twelve targets copied from the internet. Only food, labor and prime cost have published ranges worth using. The rest should be your own numbers.
  • Changing the week boundary. If one week is Monday to Sunday and the next is Sunday to Saturday, nothing compares.
  • No notes line. A dashboard with "food cost 34%" and no explanation makes next month's version of you guess. Write the reason the same day.