Restaurant sales forecast template in Excel

A restaurant sales forecast Excel template multiplies expected covers or orders by their matching average check, then adds the sales across locations and services. Save those assumptions before trading and compare actual POS results with that saved version so the workbook can explain the difference.

The layout below is reproducible in your own workbook; it is not a downloadable restaurant file. For choosing comparable trading days and setting the assumptions, see restaurant sales forecasting.

Give every forecast row an operating identity

Use a Plan sheet with one row per date, location, service period and channel. Keep a separate Actuals sheet for the POS export and a Review sheet for the comparison. Record the forecast version and approval date so later edits cannot silently change the plan being evaluated.

Build live operating plans daily, then summarize them into weeks. The worked table uses weekly totals for compactness: every row covers the week beginning October 19, 2026, and Takeout combines service periods under an explicit All label. Do not mix daily and weekly rows in the same total.

Format the range as a table with headers following Microsoft's table instructions. Keep input cells visually distinct from formulas. Covirage is not affiliated with Microsoft.

Enter a complete worked forecast

A fictional US restaurant group forecasts two locations. All checks use net sales after discounts and refunds, excluding tax and tips. Takeout sales use orders; dine-in sales use covers.

Downtown lunch is 400 × $25.00 = $10,000. The other rows use the same multiplication. Downtown totals 10,000 + 14,000 + 4,200 = $28,200; Riverside totals 7,200 + 9,500 + 2,600 = $19,300. The group total is $47,500.

The volume checks are separate: 400 + 350 + 300 + 250 = 1,300 covers, and 150 + 100 = 250 takeout orders. Adding those two quantities would obscure the operating units.

This differs from a B2B sales forecast template, where a baseline and opportunity pipeline may be appropriate. A restaurant workbook needs trading-volume assumptions that an operating manager can review before service.

Match POS actuals to the same rows

Aggregate the POS export by the same business date, location, service and channel. Record actual volume and net sales separately; compute actual average check as actual sales divided by actual volume. Use the restaurant's business-day convention for late-night service.

The completed week produces these actuals:

In Review, retain Plan columns A:H and add I for actual volume, J for actual average check, K for actual sales and L for variance. Enter these formulas in K2 and L2, respectively, then fill down:

=I2*J2
=K2-H2

In a live workbook, import actual sales into K and calculate J as =IF(I2>0,K2/I2,NA()). Keep full precision, displaying checks to two decimal places. That avoids reconstructing actual sales from a prematurely rounded check.

To retrieve actual sales from a detail sheet whose columns A:D match the forecast identity and F contains net sales:

=SUMIFS(Actuals!$F:$F,Actuals!$A:$A,A2,Actuals!$B:$B,B2,Actuals!$C:$C,C2,Actuals!$D:$D,D2)

Microsoft documents SUMIFS as summing amounts that meet multiple criteria. This formula assumes the detail rows have already been normalized to the same period and service labels. Keep unmatched actual categories on an exception list.

Prove the variance and explain its drivers

Actual sales total $47,410 against $47,500 planned, so variance is -$90. Percentage variance is -90 / 47,500 = -0.2%, rounded to one decimal place. A nearly matched group total can conceal a missed service: Riverside dinner is $530 below plan.

Split each row's difference into volume and check effects. Use M for the volume effect, N for the check effect and O for the reconciliation, entering these formulas in M2, N2 and O2:

=(I2-F2)*G2
=I2*(J2-G2)
=M2+N2-L2

The reconciliation must be zero. This convention assigns the interaction between volume and check to the check effect.

For Riverside dinner, (230 - 250) × $38.00 = -$760 from volume; 230 × ($39.00 - $38.00) = $230 from check. Together they explain -$530. Use forecast accuracy and bias when evaluating repeated saved forecasts, rather than calling a single week's small net variance reliable accuracy.

Change one assumption before approving a revision

Suppose Downtown dinner's planned covers fall from 350 to 315, with its check unchanged at $40.00. Planned dinner sales become 315 × $40.00 = $12,600, down $1,400 from $14,000. The group forecast becomes 47,500 - 1,400 = $46,100.

Save that revision under a new version. Managers can then distinguish the earlier expectation from the revised staffing and purchasing basis. A separate sales dashboard workbook can present the review, while this schedule retains the assumptions and proof.

Where it goes wrong

• Net and gross sales differ. Tax, tips or inconsistent discount treatment can move actuals without a trading change. Reconcile the sales basis first.
• Orders become covers. A takeout order may serve several guests. Use the matching volume unit and average check for each channel.
• Refunds disappear. Keep refunds in the agreed sales measure and document how their timing affects the actual comparison.
• Forecasts are overwritten. Replacing the original plan with actuals removes the evidence needed to evaluate assumptions.
• Unmapped services vanish. Reconcile summarized actual sales to the POS export total, including unmatched labels and exceptional trading days.

Review your restaurant workbook in Covirage

Bring a sample of your own completed forecast and POS actuals to Covirage, which maps the columns. Its tools compute and check the supplied sales comparisons; the external AI model explains the results with every figure traced to its rows. Create a free Covirage account and set up your business with a sample of your own data.

See Covirage for hospitality for the wider operating context.