Capacity vs. demand planning is the practice of laying what customers are asking for against what the plant can actually make, week by week, to catch the moment demand quietly overtakes your available hours. The tool is a load-versus-capacity chart; the win is seeing the overload weeks early, while you can still do something.
Overloads rarely announce themselves. Order intake looks fine, the schedule looks full, and then one week the plant simply runs out of hours and a ship date slips. A load-versus-capacity view makes that week visible in advance, because it puts demand and true available capacity on the same chart before the floor discovers the gap the hard way. This post covers the difference between the two plans, how to read the chart, and what to do when the bars cross the line.
What is the difference between capacity planning and demand planning?
Demand planning forecasts what customers will order; capacity planning works out whether the plant can make it. One looks outward at the market, the other inward at hours and rates. Neither is useful alone, a demand forecast nobody can produce is a wish, and capacity nobody needs is idle cash.
The two meet in a single comparison: required load against available capacity, over a rolling horizon. Demand planning fills in the load side from forecasts and the order book. Capacity planning fills in the capacity side from your capacity planning metrics effective capacity, not nameplate. Put them on one chart and you can see, week by week, whether the promise is keepable. That comparison is the heart of rough-cut capacity planning and the sales-and-operations meeting it feeds.
| Demand planning | Capacity planning | |
|---|---|---|
| Question | What will customers order? | Can we make it? |
| Looks | Outward at the market | Inward at hours and rates |
| Built from | Forecast plus the order book | Effective capacity, not nameplate |
| Fails when | The forecast ignores what the plant can do | The capacity number is optimistic |
| They meet at | Load versus capacity, week by week, on one chart | |
What is a load-versus-capacity chart?
A load-versus-capacity chart is a simple bar chart: each bar is the demand load for a week, and a horizontal line marks available capacity. Where a bar tops the line, demand exceeds hours, an overload. Where bars sit well under it, you have slack. One glance shows which weeks are in trouble.
The power is in the honesty of the two inputs. Load has to include everything the week really demands, new orders, backlog, planned changeovers. Capacity has to be effective capacity, the hours you can truly run after planned downtime, not the theoretical maximum. Draw the line at the nameplate and every week looks fine right up until it isn't. Draw it at effective capacity and the overload weeks show up honestly, in time to act.
How do you spot the week demand quietly exceeds capacity?
Watch for the bar that creeps over the line three or four weeks out, not the one that is over this week. By the time an overload arrives, your options are all expensive; weeks ahead, they are cheap. The whole point of the chart is to move the discovery earlier.
The quiet overloads are the dangerous ones. A week that is 5% over does not look like a crisis on the order board, but 5% over on effective capacity means overtime, a missed date, or a scramble, and it often hides behind an average that looks fine. Averaging demand across a quarter can show plenty of capacity while individual weeks whipsaw over and under the line. Look at the peaks, not the mean. A quarter that balances on paper can still have three overload weeks that each cost you a customer. Set a simple rule so nobody has to eyeball it: flag any week that crosses a threshold, say, 90% of effective capacity, and review every flagged week in the planning meeting while the fix is still cheap.
What do you do when demand exceeds capacity?
You have two families of moves: raise capacity or reshape demand. Raising capacity buys more hours; reshaping demand moves the load to where the hours already are. Work them in order of cost, cheapest first, and start early enough that the cheap options are still on the table.
- Confirm the capacity line is real. Before spending a dollar, check that the overload is against effective capacity, not a pessimistic number. A wrong line invents overloads that are not there.
- Pull load forward into slack weeks. If earlier weeks sit under the line, build ahead to stock where it makes sense. The cheapest fix for a peak is a valley next door.
- Smooth the demand. Level the schedule with heijunka so peaks and valleys flatten. Sometimes the overload is self-inflicted by lumpy release, not real demand.
- Add hours at the constraint. Overtime or an added shift on the bottleneck step buys real capacity and real throughput. Aim it at the constraint, not the whole plant, so you pay only for the hours that move output.
- Offload or subcontract the peak. Move the overflow to another line, another plant, or a subcontractor for the peak weeks only. You are renting capacity, not buying it.
- Negotiate the dates. If none of the above closes the gap, move delivery dates with the customer now, while there is still notice, instead of missing them silently later.
- Only then consider permanent capacity. If the overloads are structural and recurring, not one-off peaks, you have earned the conversation about a new line or shift, sized to the real, recurring load.
Why use effective capacity, not nameplate, on the chart?
Because a chart drawn against nameplate capacity lies in the direction that hurts most, it hides overloads. If your capacity line assumes the plant runs at its theoretical maximum, weeks that are really over will sit comfortably under a fictitious line, and you will discover the shortfall on the floor instead of on the chart.
Effective capacity, demonstrated performance minus planned losses, is the only line worth drawing. It already accounts for the changeovers, downtime and slow running that eat real hours. Tie it to the constraint with line balancing so the line reflects the limiting step, and check it against takt time to confirm the plant can beat the customer's required rate at all. A load chart is only as trustworthy as the capacity line under it.
What do the numbers say about the capacity-demand gap?
At the national level, U.S. plants usually run with a cushion, but they pay for peaks with premium hours anyway. Per the Federal Reserve's G.17 Industrial Production and Capacity Utilization release total-industry capacity utilization ran 76.2% in May 2026, about 3 points under its 1972–2025 average near 79%. Yet per the BLS Employment Situation the average manufacturing workweek in mid-2026 ran about 40.3 hours with roughly 3.2 hours of overtime, plants buying peak-week capacity with premium labor even while a quarter of installed capacity sat idle on average. That is the capacity-demand gap in miniature: the hours are there in total, but not in the weeks they are needed. A load chart is how you move work from the idle weeks to the overloaded ones before the overtime bill lands.
How does capacity-demand planning connect to the rest of the plan?
Capacity-demand planning sits above the detailed schedule and feeds it. Once the load chart shows the plant can meet demand week by week, the master production schedule commits the plan and production scheduling sequences the day-to-day. Skip the load check and the schedule inherits overloads no sequencing can fix.
It also connects downstream to how you defend the plan on the floor. When a week runs hot, a sized buffer keeps the constraint fed through the crunch, the job of buffer management and a live production dashboard shows whether you are holding pace against the loaded plan or falling behind. Whether you build ahead or hold stock for the peak depends on your make-to-stock vs. make-to-order posture. The load chart decides the strategy; these tools carry it out.
Where does real-time data fit in?
The capacity line on your chart is only as honest as the production data behind it. If effective capacity is a guess, the whole load-versus-capacity comparison is a guess, and the overload weeks you most need to see are the ones a soft capacity number will hide. That is the case for measuring demonstrated capacity from the floor rather than from memory.
Plants like CLS replaced paper production logs with real-time capture, so the capacity line comes from what the plant has actually held, not from an optimistic average. Feed that number into the load chart and the overload weeks show up early and true. To pin down the effective capacity your chart depends on, run your constraint through a free OEE calculator and use the plannable rate it gives you as the line on the chart.