Restaurant Analytics: Metrics, Decisions and a 30-Day Plan

What decisions restaurant analytics helps you make

Tie every metric to an action. If you do not know what to do after seeing a number, remove it from the main screen. Keep clear deviations, assign an owner, and agree on how they will investigate the cause. That way, you can end every meeting with a concrete decision.

Plan shifts. Review revenue and receipt count by hour, broken down by weekdays and weekends. Compare workload against staff schedules. Account for kitchen prep time and delivery packaging. These tasks take up shift hours and drive staffing needs even without an increase in dine-in sales.

Manage purchasing. Gauge demand from dish sales and calculate ingredient requirements using recipe cards. Check stock balances, shelf life, and scheduled deliveries. If you simply repeat the previous purchase, you may reorder the same surplus. Clear it before placing your next supplier order.

Review the menu. Compare portion volume with the amount left over after food costs. A popular item might generate a small margin. For a slow-moving dish, you might be buying a dedicated ingredient that later gets written off. Discuss demand, assortment, and prep workflow with the kitchen before revising the menu.

Compare locations with context. Suppose you are reviewing a coffee shop in Almaty and a cafe in Astana. Check operating hours, format, foot traffic, rent, and shift structure. Choose comparable days. Then determine how much of the performance is explained by local conditions and how much can be changed by manager decisions.

Divide responsibilities before the first review. The procurement manager checks deliveries, the general manager reviews shifts, and the accountant verifies source documents. The analyst connects the data. Decisions on the menu, purchasing, and schedules are made by the manager in charge of operations.

What a restaurant dashboard is

A restaurant dashboard is a screen displaying key metrics for your venue. You use it to monitor sales, ingredient usage, and deviations over a selected period. From the total figure, you drill down into a location, day, dish, or document to check the cause.

Start with the owner's question. For example, you need to understand why one location lags behind target. The screen needs revenue, receipt count, and the ability to view specific days and hours. Checking food waste requires recipe cards, stock balances, and write-offs. Choose the screen layout based on the decision you will make.

Agree on which data is finalized and which is still being updated. Show the upload date and the source owner. Then you can distinguish an actual metric change from a file delay or unposted document and discuss the specific cause with your team.

Which restaurant metrics to track first

Start with sales, food costs, waste, and direct channel expenses. For each metric, document the definition, timeframe, and data source. Separately decide what you call revenue. Gross sales before discounts, net sales after returns, and bank deposits answer different questions.

Compare with format in mind. In the Saby article, 30-33% is described as a legacy food cost benchmark. Many establishments operate with lower numbers. Set your target based on your menu economics and venue format. Saby article on dish cost

Align recipes, purchase valuation, revenue type, and expense structure. If you calculated one month with packaging and another without it, first standardize your calculations to a single rule. The formulas and benchmark boundaries for this metric are covered in the article how to calculate food cost.

Review metrics together. An increase in write-offs may mean that employees simply started recording them. The average check can grow while the number of purchases drops. In your meeting, ask for an explanation of both changes and request the documents to verify them.

Where to get data for restaurant analytics in Kazakhstan

Compile a list of available sources. For each export, record the owner, period, generation date, and correction procedure. Check whether there is enough detail for your chosen task. Ask vendors once for the list of fields, access credentials, and transfer methods across all required systems.

POS. To analyze sales from iiko or r_keeper, you need receipt line items, time, location, dish, discounts, and refunds. Check the availability of these fields on a sample export. Then reconcile the total against the POS report for the same day and save your transaction selection rule.

Accounting and inventory. Take purchases, expenses, and required documents from 1C or another accounting system. Match stock balances and inventory counts with the sales period. If you close inventory later than the POS, label the ingredient calculation as preliminary. Once the documents are closed, check the total again.

Plans from Excel. A manager can track revenue targets, shift schedules, and the budget in a spreadsheet. Assign a person in charge, maintain a change history, and use a single template. Standardize arbitrary column names and summary rows into a structure that can be imported regularly.

Kaspi payments. Match confirmed receipts with POS sales according to your established rule. Check refunds, settlement dates, and batched payments. Find an operational explanation for every discrepancy. Save the original amounts so your accountant can repeat the reconciliation.

Delivery. Collect available reports from Wolt, Glovo, and Yandex Eda on orders and settlements. Aggregator commission depends on the contract. Separately check packaging, restaurant-funded discounts, and withholdings. Using a sample report, determine which expenses you can attribute to a specific order or time period.

Guest acquisition. Use inquiries and clicks from 2GIS if you have access to such data. Connecting them to sales requires a verifiable basis. Track profile views, visits, and receipts separately. Note which part of the guest journey you can confirm with available data.

Why POS, inventory, and bank reports do not match

Take one location and one closed day. Check receipt details, refunds, invoices, and the date of each transaction. First explain the discrepancy for this period. After agreeing on the rules, repeat the reconciliation for the remaining days.

• Different business days. The shift ends after midnight. The POS assigns the receipt to the operating day, while the spreadsheet assigns it to the calendar day. Choose a common rule and keep the original transaction time.
• Different catalogs. The same dish is named differently in the POS and accounting systems. Create a mapping table with consistent IDs. Check for renames and new portion sizes whenever an item is added.
• Duplicate delivery counting. The order is already in the POS. Use the aggregator report for channel analysis and settlements, making sure the sale is included in the total only once.
• Different units of measurement. Convert deliveries received in packages into recipe units. Keep the original unit and conversion factor. Then the purchasing manager can verify any discrepancy that looks like a shortage.

Assign an owner for master data. Add any new dish, warehouse, or location following a single procedure. Maintain a list of entries that do not yet have a match. Display them next to the total and specify which part of the calculation requires clarification.

How analytics helps find losses and potential theft

Calculate ingredient usage based on sales and recipe cards. Add documented write-offs and compare them to actual usage from the inventory count. Check for document errors, yield discrepancies, unit mismatches, and process violations. Keep the supporting rationale for each explanation so you can repeat the calculation.

Examine specific transactions. The numbers allow you to select an incident to investigate. Concluding that an employee is at fault requires verifying source records. When comparing shifts, account for workload, menu mix, and assigned duties. Analyze these differences before discussing an individual's actions.

Build a queue of events that the manager needs to review.

• Receipt cancellations after payment.
• Frequent discounts.
• Write-offs without justification.
• Recurring discrepancies within a product group.

Next to each item, list the period, document, and person responsible. It is harder to reconstruct an individual transaction from a monthly total, so preserve all available details. The investigation process and its limitations are described in the article how to detect restaurant theft using data.

What to choose for restaurant analytics

Choose among four approaches based on your objective, available data, and team capabilities. Test on a single closed period how you get answers and who can update the calculation. Factor in the time needed to prepare data sources and handle post-launch support.

In Kazakhstan, options include Choco «Smart Restaurant» and BEAST. Add them to your list of industry-specific options to test. On your own test dataset, check the required integrations, terms, and alignment with your accounting records. This will allow you to evaluate whether the service fits your task.

Yashcher implements Power BI dashboards and is interested in such projects. You can start with an agreed-upon spreadsheet or a POS report. If you already get the answers you need and regularly make decisions based on them, stick with that approach. The selection criteria are gathered in our comparison of Power BI and POS analytics.

What should be on a restaurant owner's dashboard

Build the first screen around daily questions. Keep revenue and receipt counts by location, variance from plan, ingredient usage, and unexplained events. Show which documents have been closed. Label food cost as preliminary until the physical inventory count is completed.

Provide access to detailed breakdowns. If revenue drops, look at days, hours, and menu categories. If costs rise, drill down into specific ingredients, documents, and recipes. The manager must be able to verify the root cause using source data and bring a concrete explanation to the meeting.

Suppose a hypothetical chain of four coffee shops sees an increase in the average check at one location. This is a fictional example. Check the number of receipts and the share of delivery. A shift in order composition might explain the increase. Investigate any drop in purchase volume separately before evaluating dining room performance.

Display the date of the last successful data load and the completeness of sources. When opening the report, the executive should clearly see which day the sales data covers. On a mobile phone, verify key metrics and drill-down navigation. Open complex tables on the owner's device and check their readability.

Keep a brief decision log. Record the variance, verified explanation, action taken, and follow-up review date. In our hypothetical coffee shop scenario, you could check shift scheduling during peak delivery hours. Keep all other conditions unchanged during the test period so you can attribute any metric shift directly to your action.

When evaluating a manager, consider the expenses and resources they control. If the owner changes purchase valuations or operating hours, display this alongside the location's results. In your meeting, examine which decisions influenced the bottom line and who is responsible for them.

How to implement restaurant analytics in the first 30 days

Launch your data operations in thirty days. The integration timeline depends on access permissions and data source quality. Set verifiable goals for the first month. Agree on the rules, reconcile a benchmark period, and conduct regular management reviews.

1. Week one. Choose your questions. Write down the decisions you currently make blindly. Choose a location, a closed period, and a small set of metrics. Gather POS exports, invoices, inventory balances, recipe cards, and expenses. Assign owners for any missing data.
2. Week two. Reconcile the period. Verify revenue, refunds, and receipt counts. Match dishes with ingredients, recalculate units, and isolate documented write-offs. List the reasons for discrepancies. Agree on the formulas with the general manager and accountant.
3. Week three. Build the report. Use Excel, your POS system, an industry service, or BI. Show sources, completeness, and drill-downs to source documents. Meet with users. For each metric, define the action that should follow after reviewing it.
4. Week four. Establish the rhythm. Schedule update times and review sessions. For significant variances, record the decision, the owner, and the review date. Then expand calculations to other locations and automate repetitive operations.

At the end of the month, verify whether you can find a document explaining any discrepancy and name the owner responsible for resolving the root cause. For unresolved issues, compile a list of constraints and next steps. It will help you continue working with data sources and evaluate report readiness.

All articles in the series

1. Restaurant analytics: metrics and action plan.
2. Restaurant theft: how to track losses with data.
3. Food cost: how to calculate target and actual usage.
4. Power BI for restaurants or POS reports.
5. Menu ABC analysis and menu engineering.
6. Coffee shop analytics and comparing chain locations.
7. Delivery margins for Wolt, Glovo, and other channels.
8. Restaurant dashboard: five screens for the owner.
9. iiko API, 1C, and Power BI: data preparation.

Frequently asked questions

Where do you start with analytics for a small coffee shop?

Take a closed period, revenue, receipt count, and the usage of key ingredients. Check recipes and stock balances, and assign someone to keep them updated. A POS system and a spreadsheet may be enough to start. Add metrics when a specific management question arises.

Are iiko or r_keeper reports enough?

Yes, if you get the answers you need without constantly merging files manually. Check reports on a closed day. When you need expenses from 1C, targets, and multi-location comparisons, assess the effort required to combine your data sources.

How do you know if delivery through an aggregator is profitable?

Compare channel sales with food, packaging, and contractual deductions. Check discount funding and refunds. Calculate the contribution margin after selected expenses. To calculate net profit, also agree on how to account for general overhead.

Can you detect theft using a dashboard?

You can highlight discrepancies and reconstruct available transactions. Verifying intent and root causes requires source documents, physical inventory counts, and process reviews. When consumption happens off the books, on-screen data may not be enough. Save available records and investigate the incident on-site.

Does a single restaurant need Power BI?

Assess your goals and data sources. A single restaurant might need to merge data, while a chain might be fine with standard reports. Check calculation frequency and assign someone to maintain it after launch.