Restaurant Technology and AI

Building a Weekly Restaurant Dashboard

Build a one-page weekly restaurant dashboard: the metrics, formulas and data sources that matter, a spreadsheet setup and a 30-minute review agenda.

AssistCall research teamPublished Reviewed by Faruque Amin, VP of Sales and Marketing6 min read

On this page
  1. Why weekly, and why one page
  2. The metrics that earn a place
  3. Start with prime cost
  4. Track profit by channel, not only sales
  5. Add the front-door numbers
  6. Build it in a spreadsheet first
  7. Targets that mean something
  8. A 30-minute weekly review
  9. Common mistakes
  10. Make it readable in two minutes
  11. Example: reading one week
  12. Frequently asked questions
  13. References
Restaurant Technology and AI: Building a Weekly Restaurant DashboardArticle coverRESTAURANT TECHNOLOGY AND AIBuilding a WeeklyRestaurant DashboardAva for RestaurantsInsightsRestaurant Technology and AI: Building a Weekly Restaurant DashboardArticle coverRESTAURANT TECHNOLOGY AND AIBuilding a WeeklyRestaurant DashboardAva for Restaurants

Key takeaways

  • Food and labor each take roughly 33 cents of every sales dollar[1]; put both on the first line of the dashboard.
  • Track margin by channel: 47% of full-service operators say they make no profit on third-party delivery orders[2].
  • Keep it to one page and about 15 numbers, each with a formula, a source and an owner.
  • Review it in a 30-minute weekly meeting that ends with no more than three actions.

Why weekly, and why one page

Daily numbers are too noisy to act on and monthly P&Ls arrive too late to fix anything. A weekly dashboard sits in between: long enough to smooth out a rainy Tuesday, short enough that a cost creeping up can be caught before it eats the month. It also matches how restaurants actually run, with schedules, orders and menu changes planned week by week.

The margin for error is thin. In the first half of 2026, 33% of operators said their restaurant was not profitable[3]. A dashboard will not fix pricing or rent, but it makes sure the manager and the owner are looking at the same numbers every Monday, and it turns arguments about how the week felt into a conversation about what happened.

The metrics that earn a place

Every metric needs a formula, a source system and an owner who can explain it. If nobody can change a number, it does not belong on the page.

A starting set. Add one or two concept-specific lines, such as catering leads or bar sales, and drop anything nobody acts on.
MetricFormulaSourceOwner
Net sales vs same week last yearNet sales this week / same week last year - 1POSGM
Covers or transactionsCount of guests (dine-in) or orders (counter)POSGM
Average checkNet sales / covers or transactionsPOSGM
Food cost %(Opening inventory + purchases - closing inventory) / food salesInventory count, invoicesChef
Labor cost %Hourly wages + payroll taxes / net salesPayroll or schedulingGM
Sales per labor hourNet sales / hours workedPOS + schedulingGM
Sales mix by channelDine-in, takeout, own online, third-party as % of salesPOS order sourceGM
Third-party marginThird-party sales - fees - packaging - food costDelivery statementsOwner
Missed callsCalls not answered / total callsPhone systemFOH manager
Reservations and no-showsBooked covers; no-show covers / bookedReservation platformFOH manager
Average rating and new reviewsWeighted rating; count of new reviewsReview sitesGM
Comps and voidsValue of comps and voids / net salesPOSGM
Staff turnoverLeavers this week; open positionsPayroll, HRGM

Start with prime cost

Food and labor are a restaurant's two biggest costs, each taking roughly 33 cents of every dollar in sales[1]. Together they are usually called prime cost, and they are the first two lines most owners read. Show both as a percentage of sales and as dollars, because a percentage can improve while dollars rise if sales grow faster.

Inputs keep moving, so give each cost line some context. USDA forecast beef and veal prices to rise 8.7% in 2026[4], which is enough to move food cost on a burger-heavy menu by itself. On the price side, full-service meal prices rose 3.8% in the 12 months to September 2026 while grocery prices rose 1.8%[5][6]. When restaurant prices outpace groceries, guests notice, so watch covers and average check together after any menu price change.

3.8%

rise in full-service meal prices, 12 months to September 2026[5]

Track profit by channel, not only sales

Channel mix is where many dashboards stop too early. A rising share of delivery looks like growth until you subtract the fees. In a May 2026 survey, 47% of full-service and 43% of limited-service operators said they do not make a profit on third-party delivery orders[2]. Put third-party margin on the page in dollars, and compare it with orders through your own site.

Operators making no profit on third-party deliveryShare of US operators, May 2026 survey
Operators making no profit on third-party deliveryComparison. Full-service: 47%; Limited-service: 43%.47%Full-serviceFull-service: 47%43%Limited-serviceLimited-service: 43%Operators making no profit on third-party deliveryComparison. Full-service: 47%; Limited-service: 43%.47%Full-serviceFull-service: 47%43%Limited-serviceLimited-service: 43%

Source: National Restaurant Association, 2026[2]

Show the data
ItemValue
Full-service47%
Limited-service43%

If you cannot get fee data weekly, pull the delivery platform statements monthly and enter the average fee rate as a fixed assumption until the next statement arrives. An estimate updated monthly is better than a blank.

Add the front-door numbers

Sales tell you what happened inside. Calls, bookings and reviews tell you what happened at the front door, often a week or two before it shows up in sales. A rise in missed calls on Friday evenings or a drop in the average rating is an early warning. Most phone systems report answered and missed calls; tools that log every call with a summary, such as Ava, also show what callers wanted, which helps you see whether missed calls were bookings, takeout or questions.

Staffing deserves a line too. The quits rate in accommodation and food services was 5.0% in September 2026, more than double the rate for all private industry[7]. A leaver count and open positions, reviewed weekly, tell you whether next month's schedule is at risk.

Build it in a spreadsheet first

  1. Create one tab per source

    POS export, labor export, inventory and invoices, delivery statements, phone report, reservation report. Paste raw data in; do not edit it.

  2. Build a calculation tab

    Write each formula from the metrics table once, referencing the source tabs. Label every cell with its formula in a note.

  3. Design the one-page view

    This week, last week, same week last year and target for each metric. Use conditional formatting sparingly: red only when a number misses target.

  4. Set targets

    Use your own last 12 weeks as the starting target, not industry averages. Adjust each quarter.

  5. Schedule the data pull

    Same time every week, ideally Monday morning for the week ending Sunday. One person owns the pull.

  6. Automate later

    After a few months, move to a BI tool or your POS reporting only for the metrics you actually used.

Targets that mean something

A target is only useful if the team believes it can be hit and knows what to do when it is missed. Industry averages make poor targets because rent, concept, service style and location vary so much. Start from your own history: the median of the last 12 comparable weeks for each metric, excluding holidays and one-off events. Then set a modest improvement for the quarter on two or three lines at most, and leave the rest as guardrails that should simply stay inside a range.

  • Use ranges for volatile lines such as food cost, so one late invoice does not trigger an alarm.
  • Use seasonal targets for sales and covers, built from the same weeks last year.
  • Hold one person accountable per target, and write their name next to it on the page.
  • Revisit targets each quarter, and after any big menu, price or hours change.

A 30-minute weekly review

  1. Minutes 0 to 5: headline sales, covers and average check against last week and last year.
  2. Minutes 5 to 15: prime cost lines, with the chef and GM explaining any line off target.
  3. Minutes 15 to 20: channel mix and third-party margin.
  4. Minutes 20 to 25: front-door numbers: missed calls, bookings, no-shows, reviews.
  5. Minutes 25 to 30: agree up to three actions, each with an owner and a date, and check last week's actions.

Three actions, not ten. A dashboard that produces a long to-do list every week stops being read.

Common mistakes

  • Too many metrics: if the page needs scrolling, cut it.
  • No same-week-last-year comparison, so seasonal swings look like trends.
  • Food cost from purchases only, without an inventory count, which swings wildly week to week.
  • Labor without payroll taxes, which understates the real cost.
  • Numbers nobody owns, so nobody explains or fixes them.

Make it readable in two minutes

Owners read dashboards on phones between other jobs. Design for that. Put the five most important numbers at the top in large type, with this week, last week and the same week last year side by side. Use one color for misses and nothing for hits, so attention goes where it is needed. Add a short notes line for context the numbers cannot show, such as a road closure, a private buyout or a broken fryer. When a number moves sharply, the note should explain it before anyone has to ask.

Example: reading one week

Example (fictional): a 120-seat restaurant's dashboard showed sales up 4% on last year, but food cost up two points and covers flat. The chef traced the food cost jump to a short-rib special priced before the latest beef invoice. Third-party margin was negative for the second week, driven by a promotion the delivery app had opted the restaurant into. Missed calls on Saturday had doubled after a host left. The three actions: reprice the special, opt out of the promotion, and cover the Saturday phone gap until the new host started.

Clean inputs make all of this easier; see connecting POS, reservations and ordering. If the dashboard shows staffing swings, use it alongside AI sales and labor forecasting. Counter-service operators can add menu board tests from digital menu boards for quick service, and lane metrics from drive-thru voice AI pilots.

Frequently asked questions

What should be on a restaurant dashboard?

Sales against last year, covers, average check, food and labor cost, sales per labor hour, channel mix and margin, missed calls, bookings and no-shows, ratings, comps and voids, and staff turnover. Keep it to one page.

What is a good prime cost?

Food and labor each average roughly a third of sales[1], but targets depend on concept, rent and service style. Use your own last 12 weeks as a starting target and improve from there.

Do I need special software?

No. Start with a spreadsheet and the exports your systems already provide. Move to a BI tool once you know which metrics you actually use.

Who should attend the weekly review?

The owner or operator, the general manager and the chef at minimum. Keep it to 30 minutes and end with up to three owned actions.

References

  1. National Restaurant Association (2026). Restaurants remain resilient despite challenging business conditions (opens in a new tab) ↑

  2. National Restaurant Association (2026). National Restaurant Association comments to the FTC on the Food Delivery Fees ANPRM (Project No. P267101) (opens in a new tab) ↑

  3. National Restaurant Association (2026). Restaurants remain resilient despite challenging business conditions (opens in a new tab) ↑

  4. USDA Economic Research Service (2026). Food Price Outlook - Summary Findings (opens in a new tab) ↑

  5. US Bureau of Labor Statistics (2026). Consumer Price Index: series CUUR0000SEFV01 (opens in a new tab) ↑

  6. US Bureau of Labor Statistics (2026). Consumer Price Index: series CUUR0000SAF11 (opens in a new tab) ↑

  7. US Bureau of Labor Statistics (2026). Job Openings and Labor Turnover Survey: series JTS720000000000000QUR (opens in a new tab) ↑

Book a demo

See Ava handle your calls.

Book a short demo. We’ll walk through your calls and workflow, and agree what Ava should take care of.

Book a demo