Licensing Against Manufacturing: The Real Trade-Off

The question is usually framed as ambition against caution, which is the wrong frame entirely. A royalty is a small share of a large number produced by somebody with existing distribution. A manufacturing margin is a large share of a small number produced by somebody who has to build the distribution first. Licensing versus manufacturing is therefore a question of which of those two numbers is larger, and that depends on unit economics, working capital and the calendar — three things that can be modelled on a single page before any decision is made.

What follows prices both routes on the same product so the comparison is like for like: a housewares item with a landed manufacturing cost of $3.20, a wholesale price of $9.00 and a retail price of $19.99. Every figure below moves with the category, but the structure of the comparison does not.

Route One · The Licence

What a Licence Actually Pays

Under a licence, the inventor's revenue is a percentage of the licensee's net sales — net meaning after returns, allowances, freight and trade discounts, which is why the definition of the base is negotiated harder than the rate itself. On a consumer hardware item, four per cent of net wholesale sales is an ordinary outcome. At 150,000 units a year and $9.00 wholesale, that is roughly $1.35 million in net sales and about $54,000 a year to the inventor.

Against that income sits almost no cost. There is no tooling to buy, no inventory to finance, no returns to absorb, no product liability exposure sitting on the inventor's own balance sheet, and no staff. The licensee carries the entire operating apparatus. In exchange, the inventor gives up control of every commercial decision that follows: pricing, packaging, which retailers are approached, whether a second size is introduced, and — most consequentially — whether the product is actively promoted at all.

• 2% – 7%Typical royalty band on net wholesale sales for consumer hardware
• $5k – $25kAdvance against future royalties, proving committed budget rather than an option
• 6 – 18 moFirst approach to signature, before any tooling or retail slot exists

Three clauses decide whether a signed licence is worth anything. An advance, commonly between $5,000 and $25,000 and credited against future royalties, demonstrates that a budget line has been opened rather than an option quietly taken. A minimum annual royalty is the only real defence against an exclusive licence held by a company that then decides not to launch — without it, exclusivity can freeze a product for years at no cost to the holder. A defined sublicensing split, usually thirty to fifty per cent of what the licensee receives from third parties, stops the value leaking one layer downstream.

The honest weakness of licensing is conversion. A licence has to be sold to somebody with an existing product line, an existing factory and an existing shelf, and most such companies review far more submissions than they take. Accounts of how independent products eventually reach shelves, such as the development story behind the MixAid device, tend to describe a long sequence of qualified conversations rather than a single lucky introduction.

Route Two · The Factory

The Cost of Owning the Product

Manufacturing pays the whole gross margin. On the same item at $3.20 landed and $9.00 wholesale, that margin is $5.80 a unit, or about seventy per cent above cost — some twenty-six times the four cents a royalty would have produced on the same sale. That ratio is why the route is tempting, and it is also where most modelling stops.

The costs begin before the first unit exists. Injection tooling for a moulded housewares part runs from $12,000 to $45,000 depending on cavity count and finish; a single tool is normal for a first product and two or three are common once packaging inserts are counted. Design for manufacture, sampling and revisions add $4,000 to $15,000. Safety and compliance testing, packaging design, barcodes and artwork add several thousand more. A first production run at a factory's minimum order quantity — frequently 3,000 to 10,000 units — ties up $10,000 to $32,000 in stock at the cost figure above, payable in large part before the goods ship.

Then come the costs that recur. Warehousing, pick-and-pack and freight commonly consume eight to fifteen per cent of revenue. Retail chargebacks, co-op advertising allowances and slotting demands can take a further five to twenty per cent from a buyer who has leverage. Returns on consumer goods run two to eight per cent depending on channel, and marketplace fees on direct sales take twelve to fifteen per cent of the gross before shipping. Product liability insurance, at $1,500 to $6,000 a year for a low-risk item, is not optional once a physical product carries a name.

The decisive constraint is working capital rather than margin. Retail terms of net sixty are ordinary, which means stock is paid for four to five months before it is paid for by anybody else. Growth makes that gap worse, not better: doubling orders doubles the cash locked in inventory, and a profitable product can exhaust a small operation precisely because it is selling well. That is the mechanism behind most failures of the manufacturing route, and it has nothing to do with the product being wrong.

Comparison · Break-even

Where the Two Curves Cross

Set the two side by side at the same volume and the shape becomes clear. At 150,000 units a year, licensing returns roughly $54,000 with essentially no outlay and no operational load. Manufacturing the same volume produces $870,000 of gross margin, from which perhaps $130,000 of warehousing and freight, $90,000 of retail allowances and returns, $180,000 of marketing and $120,000 of staff and overhead are deducted — leaving something in the region of $350,000 before tax, on a business that required $80,000 of upfront capital and now permanently holds $150,000 or more in inventory and receivables.

Manufacturing wins that comparison decisively, and it wins it on the assumption that the volume happens. Reverse the volume and the ranking reverses with it. At 8,000 units a year, a licence pays about $2,900 — disappointing but costless. Manufacturing 8,000 units generates roughly $46,000 of gross margin against fixed costs of warehousing, insurance, compliance and the amortised tooling that do not shrink with volume, and the result is close to nothing, or below it. The crossover for a product of this type generally sits somewhere between 15,000 and 40,000 units a year.

Which means the real comparison is not between two income figures but between two probability distributions. Licensing has a low ceiling, a floor at zero and very little variance in between. Manufacturing has a high ceiling, a floor well below zero once borrowed money is involved, and enormous variance. Framing the whole project as a ledger rather than a lottery ticket makes this legible: the question is not which route pays more at its best, but which distribution the household can actually survive.

Category · Fit

Some Products Only Work One Way

The choice between licensing versus manufacturing is often made by the product rather than the inventor, and recognising that early saves years. Four attributes decide it.

1. Where the value sits. If the invention is a mechanism that improves an existing product category, licensing is natural — the buyer already has the tooling, the distribution and the brand, and needs only the mechanism. If the invention is the entire product and there is no incumbent whose line it slots into, there may be nobody to license it to.
2. Regulatory weight. Anything touching medical claims, food contact, children's products or electrical safety carries an approval burden that only an established operation absorbs comfortably. That burden is an argument for licensing regardless of margin.
3. Shelf economics. Products that need a physical retail slot are licensing candidates because the slot is the asset and the incumbent already owns it. Products that sell adequately through direct channels can be manufactured, because the distribution constraint is softer.
4. Iteration speed. If the design will change three times in two years, tooling is a liability and a licensee will not tolerate it either. Freeze the design or expect the route to be forced.

Category context also cuts across the arithmetic. In fields where the buyers are large industrial purchasers rather than shoppers — the pattern visible in reporting on independent invention in the environmental technology sector — licensing is frequently the only realistic route, because the sales cycle and certification requirements are beyond anything an individual can fund. In fast-moving consumer categories with short tooling cycles, the balance tips the other way.

Hybrid · Sequence

Manufacturing First, Licensing Later

The two routes are usually presented as alternatives, and for most independent inventors the more useful framing is sequential. A small self-funded run of 500 to 2,000 units does two things a pitch deck cannot. It produces a landed unit cost verified by an invoice rather than a quote, and it produces sell-through data — units moved, reorder rate, return rate, the price a real buyer paid. Those two numbers convert a licensing conversation from a speculative discussion into a negotiation over percentages, and they typically move the achievable royalty rate up by a point or two, along with the size of the advance.

Run in that order, the risk is bounded. A 1,000-unit run at $3.20 with a single soft tool or a bridge process costs perhaps $8,000 to $20,000 all in — real money, but a fraction of a full launch, and it either produces evidence or produces an early stop. Anything the run reveals about cost, breakage or buyer indifference is cheaper to learn at this scale than at fifty times it. The way experienced practitioners describe the arc, including in this conversation about how independent inventors reach market, is a series of small tests rather than one committing bet.

Two habits keep the sequence honest. Model both routes on the same page before committing, using the same volume assumptions for each, and write down the volume at which the ranking flips. Then decide in advance what evidence would change the answer, because the temptation once tooling money is spent is to defend the route rather than the product. Whichever way the comparison falls, both paths depend on knowing what the underlying right is worth: valuation before a product exists sets the floor for a royalty negotiation, and the capital required for a first run sets the ceiling on how much manufacturing is even available as an option.

Read properly, licensing versus manufacturing is not a question about character. It is a question about which distribution of outcomes an individual can carry, and the answer changes with volume, category and the size of the buffer standing behind the decision.

Price both routes at the same volume, find the crossover, and let the number decide — then spend the smallest amount that would prove the number wrong.