An idle line makes unused capacity feel like a marketing problem: if we had more orders, we could make more product. Sometimes that’s exactly right. But factory headroom only tells you the operation may be able to accept more demand. It doesn’t tell you what that demand will cost to acquire, or what the extra orders will leave behind.
This is for owners who make their own products and are looking at open capacity, wondering whether to turn the ads up.
Unused capacity is a legitimate reason to test more advertising — as long as each additional order still leaves acceptable contribution, and the next customers cost less to acquire than those orders can support.
Capacity creates the opportunity. It doesn’t create profitable demand. The right move is rarely “turn the ads up.” It’s a controlled test: add one increment of spend, measure what the additional customers actually cost and leave behind, and only add the next increment if the economics and the operation still hold.
In this article
How much unused capacity do we really have?
Usually less than the machines suggest. Capacity isn’t one number, and the one that matters for a marketing decision is the smallest.
Managerial accountants have long separated what resources could theoretically do from what they can practically do. In their work on activity-based costing, Robert Kaplan and Steven Anderson define practical capacity as what remains after allowing for breaks, maintenance and similar losses, and make unused capacity visible so managers can decide what to do with it. For a demand decision, three distinctions are useful:
- Theoretical capacityWhat equipment and schedules could produce under near-ideal conditions.
A rating, not a promise.
- Practical capacityWhat you can reliably produce while keeping normal quality, staffing, maintenance, workflow and service.
The operating number.
- Economic capacityHow much more you can produce and sell at acceptable economics before something materially changes.
The number the marketing decision needs.
These are management distinctions, not accounting standards, and every operation draws them a little differently. The point is simple: a machine rated for 10,000 units doesn’t mean the business can responsibly accept 10,000 orders. And keeping some buffer for maintenance, variability, quality problems, rush orders, supplier disruption and absences is a decision each business makes for itself. There’s no universal right utilization.
What does one more order actually contribute?
Build the decision from contribution, not revenue, and count only the costs that actually rise with the extra order.
Two questions for your operations team
If we sold one more unit tomorrow, which costs would actually change?
If we sold 1,000 more next month, which costs would change?
The answers are often different. At low extra volume, equipment, buildings, salaried supervision, software and some labor already exist. At higher volume, overtime, another shift, a hire, expedited materials, outsourced production, warehouse space, customer service or 3PL costs can start to move. Managerial accounting calls these step costs: they hold steady over a range of activity, then jump.
- First-order net revenue
- $100
- Materials and incremental production cost
- −$32
- Packaging, fulfillment, payment fees, shipping
- −$13
- Contribution before acquisition
- $55
- Customer acquisition cost for that customer
- −CAC
- Contribution after acquisition
- $55 − CAC
Which costs belong depends on your operation. Contribution after acquisition isn’t net profit; fixed costs still come out of it.
That $55 is also this order’s first-order break-even CAC: spend more than $55 to win the customer and the order contributes nothing. How to calculate break-even CAC when you manufacture your own products walks through that number, including how to treat overhead.
Doesn’t more volume lower our cost per unit?
Reported cost per unit, yes. Total fixed cost, no. Keep those two ideas apart.
Under absorption costing, the standard approach for financial reporting, fixed manufacturing overhead is folded into product cost. Spread the same overhead across more units and each unit carries less of it.
That’s real, and it’s why unused capacity makes additional demand potentially valuable. But the overhead didn’t shrink; it was spread thinner. It doesn’t make every extra order profitable, and it doesn’t justify accepting any CAC because a machine is idle. For the marketing decision, ask two separate questions, the way managerial accounting separates costs relevant to a decision from the rest:
- What additional contribution does the next order create, after the costs that actually rise with it and after acquisition?
- Separately: does the business’s total contribution cover its fixed costs and profit requirements?
What’s the difference between average CAC and marginal CAC?
Average CAC is what customers have cost across current spend. Marginal CAC is what the additional customers from your latest increase cost. For a scaling decision, the second one matters.
- Before$60,000 spend · 1,500 new customers
$40.00 average CAC
- The increase+$20,000 spend · +350 new customers
$57.14 marginal CAC
- After$80,000 spend · 1,850 new customers
$43.24 average CAC
CAC doesn’t always rise with spend, and there’s no fixed rate at which it does. But it often can. The most responsive customers may already be reached; added spend can move into less responsive audiences; auction costs, creative effectiveness, channel mix, seasonality and promotions change; and demand in a market is finite. Google’s own planning tools let the conversion rate change as budgets rise, to account for diminishing returns as investment increases. Don’t assume the next $20,000 buys customers at the same price as the last. Measure it.
What does this look like in a real decision?
A factory with open capacity, attractive current economics, and a first test that works — just not as well as the average suggests.
- Rated monthly capacity
- 5,000 units
- Current production
- 3,800 units
- Apparent headroom
- 1,200 units
- Practical near-term headroom after maintenance, scheduling, quality buffer and fulfillment
- 800 units
- Contribution before acquisition
- $55
- Current blended all-in CAC
- $32
- Contribution after acquisition
- $23 per customer
Looks attractive.
- Additional media spend
- $15,000
- Additional new customers
- 350
- Marginal media CAC
- $42.86
- Plus added management and creative
- $1,450
- Marginal all-in CAC $16,450 ÷ 350
- $47.00
- Contribution after acquisition, per added customer $55 − $47
- $8
- Total incremental contribution 350 × $8
- $2,800
Blended all-in CAC across all 1,850 customers is now about $34.84 — still comfortable. The added 350 customers cost $47 each and leave $8, not $23.
The test worked: it was still profitable, and the factory still has 450 units of practical headroom. But the margin on the new increment is far thinner than the average suggests, and it’s approaching the line. The conclusion isn’t “we have room for 800 more, so buy 800.” It’s “test the next increment, measure it, then decide.”
- Current output3,800
- Profitably acquired at the next increment+350
- Remaining practical headroom+450 · to 4,600
- Operating buffer to rated capacity+400 · to 5,000
What else can limit capacity besides the machines?
Materials, labor, fulfillment and cash. Any of them can be smaller than production headroom.
- Materials and components
- A factory that can physically make 1,000 more units this month can’t, if a key component takes eight weeks to reorder. Practical capacity for this month may be far lower than the machine rating. Marketing needs the operating constraint, not the nameplate.
- Labor
- Idle equipment doesn’t mean idle people, and existing payroll can mean some extra output costs little in labor at first. Past a threshold, overtime, a shift, temporary staff or new hires can change unit economics materially. Neither “all labor is variable” nor “all labor is fixed” is safe.
- Fulfillment
- Every extra unit still has to be picked, packed, shipped, supported and sometimes returned. A factory that can make 1,000 more but ship only 300 more reliably has 300 units of customer capacity. Demand you can’t fulfill has costs too: in one field study, stockouts reduced customers’ later purchasing as well as the current order.
- Working capital
- Scaling spends cash before it comes back: media, raw materials, labor, freight, packaging and inventory built ahead of orders. A profitable incremental order can still create cash pressure, especially when suppliers want deposits, lead times are long, or acquisition pays back only through repeat purchases. Forecasting is how businesses avoid tying up unnecessary capital in stock; marketing is one of its biggest inputs.
None of this makes Coast333 an operations consultancy or a CFO. It’s the operating context a demand decision can’t ignore. MIT Sloan Management Review research describes the cost of the alternative: companies whose demand and supply plans aren’t integrated end up selling excess product below market or losing sales to shortages.
This isn’t theoretical for us. Coast333’s founder spent roughly eight years inside a direct-to-consumer manufacturer handling about 10,000 orders a month, across operations and eventually as Head of Marketing, where more demand had real consequences for inventory, production, labor, fulfillment and margin.
When does unused capacity justify more marketing?
Sometimes the answer is simply yes: go create more demand. The test is whether the whole system supports it.
- Practical capacity genuinely exists
- Materials and inventory support it
- Fulfillment supports it
- New orders leave adequate contribution before acquisition
- Current CAC is acceptable
- Recent marginal CAC is still acceptable
- Cash can fund the scale
- Conversion is healthy
- Measurement is trustworthy enough to judge the test
- No more obvious bottleneck sits upstream
- The product loses money after acquisition
- CAC is already near or above the economic ceiling
- Conversion is weak
- Materials are constrained
- Fulfillment is overloaded
- Cash can’t fund inventory and acquisition together
- Tracking is too unreliable to judge a test
- The business needs a different product mix, not more orders
- The current channel mix has hit diminishing returns
The second column isn’t an argument for less marketing. It’s a reminder that idle equipment isn’t evidence that more customer acquisition will pay. A strong-looking platform return doesn’t settle it either; see why a 5x ROAS can coexist with no profit.
Is marketing a factory-utilization tool?
No. A factory can have unused capacity because market demand is genuinely limited, the product mix or pricing is wrong, demand is seasonal, sales channels are concentrated, or the economics don’t support buying more demand. Marketing’s job isn’t to keep every machine busy at any cost. It’s to create the right amount and quality of profitable demand for the system the company is trying to run.
How do you scale ad spend responsibly into unused capacity?
In increments. Scale is a series of decisions, not a switch.
Don’t reason “we have 20% unused capacity, so raise spend 20%.” Capacity percentages and spend percentages aren’t interchangeable; the relationship between spend and customers is what you’re testing. Continuing the scenario:
- Current$48,000 all-in · 1,500 customers · $32 CAC
$23 contribution per customer after acquisition; 800 units of practical headroom
- Test 1+$16,450 all-in · +350 customers · $47 marginal CAC
+$8 per added customer, +$2,800 total; 450 units of headroom left. Still clears the line, so consider one more step.
- Test 2+$15,000 media · +250 customers · $60 marginal media CAC
Above the $55 break-even before any management or creative cost: about −$5 per added customer, −$1,250 in total. Illustrative.
- HoldStop at the last increment that cleared the line
Even with 200 units of practical headroom still open. Hold also when capacity tightens, cash tightens or measurement confidence falls.
At that point the constraint has moved. Capacity is still available, but acquisition economics at this channel mix now bind. The next move might be better creative, a new channel, a product-mix change or conversion work, not more of the same spend.
Why does the bottleneck keep moving?
Because relieving one constraint exposes the next. At the start, demand may be the bottleneck. After advertising succeeds, it may be materials, labor, production, fulfillment, inventory, working capital or customer service. The point of scaling isn’t to declare marketing the permanent bottleneck; it’s to relieve the current constraint until the next one binds. That’s the logic behind working out whether marketing is really your growth bottleneck.
Judging each increment needs numbers you can trust. Use actual new customers and company sales, not platform-attributed revenue alone; why Meta, Google, Shopify and GA4 show different revenue explains which number should judge a test.
What should I check before increasing my ad budget?
Seven questions, in order, every time you consider the next increment.
- 01CapacityCould we actually fulfill more orders well?
- 02EconomicsWhat does the next order leave before acquisition?
- 03AcquisitionWhat are the next customers likely to cost us?
- 04CashCan we fund the media, production and inventory required?
- 05MeasurementWill we know whether the test worked?
- 06TestIncrease demand by one increment.
- 07ReassessDid the constraint move?
Most of these numbers belong on the owner’s monthly panel anyway: CAC and its trend, contribution after acquisition and capacity headroom. What marketing numbers an owner should review every month lays that panel out.
So should I turn the ads up?
If you have real, practical headroom, test it. Unused capacity can make additional demand genuinely valuable. But let the test tell you how much: each increment has to clear its own economics, and the answer is often smaller than the factory’s open capacity. Scale until the next increment stops paying or the next constraint binds, then decide what to fix.
Capacity creates the opportunity. Marginal economics decide how much of it to buy.
Frequently asked questions
Should I increase ad spend if my factory has unused capacity?
Test it, rather than simply increasing it. Unused capacity makes more demand potentially valuable, but each additional increment has to leave acceptable contribution after its own acquisition cost, and materials, fulfillment and cash have to support it.
What is marginal CAC?
The acquisition cost of the additional customers produced by an increase in spend: extra spend divided by extra new customers. It can be much higher than your average CAC even when the average still looks healthy.
Should I use average CAC or marginal CAC when scaling?
Marginal CAC for the scaling decision, because it tells you what the next customers cost. Average CAC is useful for tracking overall efficiency, but it hides a deteriorating increment.
Why does CAC get worse when I increase ad spend?
It doesn’t always, but it often can: the most responsive customers may already be reached, added spend moves into less responsive audiences, and auction costs, creative, channel mix and seasonality change. Measure each increment rather than assuming.
Does unused manufacturing capacity lower my break-even CAC?
It can raise it, if extra orders need little additional labor or overhead, so more of each order counts as contribution. Past a capacity threshold, overtime or another shift can lower it again. Break-even depends on which costs actually rise with the order.
What is the difference between practical capacity and theoretical capacity?
Theoretical capacity is what equipment and schedules could produce under near-ideal conditions. Practical capacity is what you can reliably produce while keeping normal quality, staffing, maintenance and service levels. Plan demand against practical capacity.
Can more sales improve manufacturing profitability?
Yes, if additional orders leave positive contribution after acquisition, because that contribution helps cover fixed costs that don’t grow with volume. Spreading fixed costs over more units lowers reported cost per unit, but total fixed costs don’t shrink.
Should I run ads just to keep my factory busy?
No. Advertising should buy demand only when the extra orders leave acceptable contribution. Idle capacity can also mean limited market demand, the wrong product mix or pricing, or seasonality, none of which more ad spend fixes.
How much capacity should I leave unused?
There’s no universal number. Each business sets its own buffer for maintenance, variability, quality, rush orders, supplier disruption and absences. Plan demand against what you can produce reliably, not the equipment rating.
How does working capital affect marketing scale?
Scaling spends cash on media, materials, labor and inventory before customers pay back. Even profitable increments can strain cash when lead times are long, suppliers need deposits or payback depends on repeat purchases. Check cash before each increment.
Sources & further reading
- Time-Driven Activity-Based CostingKaplan and Anderson, Harvard Business School working paper 04-045.
- Identify and Apply Basic Cost Behavior PatternsOpenStax, Managerial Accounting, 2.2.
- Compare and Contrast Variable and Absorption CostingOpenStax, Managerial Accounting, 6.5.
- Identify Relevant Information for Decision-MakingOpenStax, Managerial Accounting, 10.1.
- Contribution Margin: What It Is, How to Calculate It, and Why You Need ItHarvard Business Review, 2017.
- About forecasts in Reach PlannerGoogle Ads Help.
- Integrating Supply and DemandMIT Sloan Management Review, 2015.
- Measuring and Mitigating the Costs of StockoutsAnderson, Fitzsimons and Simester, Management Science, 2006.
Every capacity figure, cost, CAC and dollar amount in this article is illustrative, created to show the reasoning. None is a benchmark, an industry average, a utilization target or client data.
