Manufacturing demand forecasting for production plans

Manufacturing demand forecasting estimates customer requirements by product and delivery period, then tests those estimates against the order book. Compare the output plan and manufacturing capacity separately so a supply constraint does not masquerade as weaker demand. A useful review explains both the forecast miss and whether the factory plan matches the current customer requirement.

Define the demand measure before opening the workbook

An illustrative US valve manufacturer reviews September 2026 demand after the month closes. Its frozen forecast estimates units customers will request for September delivery. The comparison uses confirmed order quantities assigned to that requested delivery period, including orders booked before September.

That differs from a report of orders received during September. The Census Bureau's manufacturing definitions distinguish shipments, new orders net of cancellations and unfilled orders. Use those distinctions to label your exports, then state the timing basis of your own planning measure. Do not compare a delivery-period forecast with an order-receipt report simply because both contain a month column.

Keep requested dates available when a promised shipment date changes. Moving a late order into the following month can make a capacity problem disappear from the original demand comparison. Account managers should identify cancellations separately from factory-driven rescheduling.

The broader demand forecasting guide covers method selection. Here the job is to validate a supplied customer forecast and the business decisions built around it.

Prepare matched product rows

Use the approved forecast snapshot, confirmed customer-order export and output plan for the same products and period. Keep quantities in saleable units, converting cases or packs before aggregation. Retain the snapshot dates rather than replacing the approved plan with the latest spreadsheet.

For an Excel workbook, a Raw sheet can hold product in A, requested month in B, order status in C and net units in D. Each order line appears once. Net units reflect approved quantity changes and cancellations without retaining an additional active copy of the original line.

On a Review sheet, put product in A, frozen forecast in B, confirmed demand in C and output plan in D. Aggregate the confirmed September units in C2 with:

=SUMIFS(Raw!$D$2:$D$1000,Raw!$A$2:$A$1000,$A2,Raw!$B$2:$B$1000,DATE(2026,9,1),Raw!$C$2:$C$1000,"Confirmed")

Store requested month as its first calendar date. Microsoft's SUMIFS documentation explains how multiple criteria restrict the values added; all referenced ranges must have matching dimensions. Covirage is not affiliated with Microsoft.

Confirm the aggregated demand equals the corresponding filtered order report before calculating error. Resolve missing product codes and duplicate order lines first. A spreadsheet total can be internally consistent while still counting the wrong records.

Calculate forecast error without netting products together

Put the following example in Review, with headers in row 1 and product rows 2 through 4. All products belong to the same valve family, and quantities use individual finished units.

Enter these formulas in E2, F2 and G2 and fill down:

=C2-B2
=ABS(E2)
=D2-C2

The net demand miss is 950 minus 900 = 50 units. Absolute error is 60 + 30 + 20 = 110 units. Weighted absolute percentage error uses actual confirmed demand as the denominator:

=SUM(F2:F4)/SUM(C2:C4)

The result is 110 / 950 = 11.6%, rounded to one decimal place. Keep unrounded calculations in the workbook. If the denominator is zero, flag the result as undefined and report absolute units instead.

The output plan also has a mix problem. It falls short of Valve A demand by 30 units and Valve C demand by 10, while planning 20 more Valve B units than required. Gross shortages are 40 units; the surplus is 20. The identity is 20 surplus minus 40 shortage = the net output gap of -20 units.

Calculate shortages and surplus in separate helper columns with =MAX(C2-D2,0) and =MAX(D2-C2,0). Do not let an unwanted Valve B offset a missing Valve A in the exception list. Repeated directional misses belong in a forecast bias review by product and site.

Check the plan in a common capacity unit

Units alone do not establish whether production fits. These products share a constrained work center but require different standard processing minutes. Assume the planning team has already validated the standards and a net available schedule of 6,000 minutes for the same period.

Use a Capacity sheet with the following columns A through F:

Calculate E2 and F2 with:

=B2*D2
=C2*D2

Place net available minutes in H2. The plan's capacity gap is:

=SUM(F2:F4)-$H$2

The output plan exceeds availability by 6,320 minus 6,000 = 320 minutes. Confirmed demand exceeds it by 6,340 minus 6,000 = 340 minutes. The planned surplus of the slower Valve B uses time while shortages remain elsewhere.

This is a screening comparison. It does not replace a production schedule or material requirements calculation. Here, opening usable finished inventory is assumed to be zero; in your review, reconcile available inventory and other committed supply before deciding how much demand requires new output. Include setup and other allowances either in the standards or the available-time calculation, consistently.

Where manufacturing forecasts go wrong

• Shipments replace demand. Unfilled customer requirements can disappear when output or delivery capacity limits shipments. Keep fulfillment and requested demand distinct.
• Orders are added twice. Some customer forecasts already include firm orders. Confirm whether the forecast is total demand or an additional unbooked estimate before combining them.
• Dates change the question. Order receipt, requested delivery, promised delivery and shipment dates describe different events. Label the basis of each comparison.
• Capacity uses mixed units. Finished units, machine minutes and labor hours cannot share a denominator without an explicit conversion and matching scope.
• Sparse products distort percentages. Investigate absolute quantities and the timing of orders for slow movers; intermittent demand forecasting covers that separate pattern.

Sales owns the explanation for changed customer requirements. Operations owns output feasibility, and finance owns the reconciliation between the supplied versions. Use the demand forecasting process to record who approves the next forecast and which exceptions remain unresolved.

Compare your manufacturing files in Covirage

Upload a file or Google Sheet containing product forecasts, confirmed demand, output plans and capacity assumptions; Covirage maps the columns, and its tools compute and check the comparisons. The external AI model explains the results with every figure traced to its rows. Create a free Covirage account, set up your business and bring a sample of your own data; creating the account takes about a minute.

See Covirage for industrial manufacturers for the industry context.