A contract development and manufacturing organization (CDMO) can grow revenue and still become a weaker business.
It can win customers, fill its production calendar and announce a new facility while margins deteriorate, cash tightens and delivery becomes less reliable. Growth does not prove scalability. It reveals whether the company's structure can carry more load. Law I — Structure Must Exceed Load™.
For a CDMO, two uses of capital shape that structure: operating expenditure and capital expenditure. Understanding them is foundational. Choosing between them is strategic.
◆I. First, understand the money
Operating expenditure, OPEX, is the cost of running and improving the business. It is the recurring spend: people, training, quality systems, maintenance, utilities, sales, software, and the daily work of developing, making, testing and delivering product. Some manufacturing cost flows through cost of goods sold rather than the line labeled "operating expenses" on the income statement. Here, OPEX means the broader operating investment decision, wherever the accountants file it.
Capital expenditure, CAPEX, is money invested in assets expected to serve the business over several periods: production equipment, laboratory instruments, facility improvements, automation, new lines. The cash leaves today. The asset's cost generally reaches the income statement over time, through depreciation or amortization, under whatever accounting rules apply.
That accounting difference matters. It does not change the most important fact.
◆II. Both consume cash
A five-million-dollar line, in the model I use, does not become inexpensive because its cost is depreciated over seven years. A quality leader does not become less valuable because her salary is expensed this year. Accounting classification tells us how a cost is reported. Strategy asks what capability the spending creates, when it creates it, and whether customers will pay enough for it.
Law III — Growth Eats Cash First™. A CDMO that grows into a new facility pays for the facility, the people to run it, the validation to qualify it and the working capital to feed it before the first profitable batch leaves the dock. The order of those payments is the risk itself.
◆III. OPEX builds the operating system
In a CDMO, operating investment is usually the first path to scale.
A better production schedule can release capacity the company already owns. Stronger technical transfer can shorten the path from formulation to a repeatable commercial batch. Better maintenance can reduce downtime. Better purchasing can protect gross margin. Stronger quality assurance and quality control can prevent deviations, rework, release delays and lost customer trust.
These are not "overhead" decisions. They determine how much useful output the existing asset base can produce.
But adding OPEX without a defined constraint is dangerous too. More managers do not create throughput. More salespeople do not fix a weak transfer process. More quality staff cannot compensate forever for a badly designed process.
Every meaningful operating investment should answer four questions:
- Which constraint does it remove?
- What measurable result should change?
- How quickly should that result appear?
- Does the improvement repeat as volume grows?
The measures are the ones a plant already keeps or should: right-first-time rate, schedule adherence, batch cycle time, yield, deviation closure, on-time delivery, contribution margin, cash conversion. The point is to connect spending to evidence. Principle 14 — Fund the Constraint, Not the Ambition™.
◆IV. CAPEX builds the physical ceiling
Eventually a well-run system reaches a real limit of capacity or capability. That is when CAPEX can change the company's trajectory.
A new line may raise throughput. An instrument may bring a critical test in house. Automation may cut labor per unit or improve consistency. A facility may open a category the company could not serve before.
The investment can create more than volume. It can change which customers the CDMO can win, the complexity it can handle and the reliability it can promise.
It also creates a new burden. Equipment needs operators, maintenance, validation, quality oversight, utilities, working capital and enough orders. A larger facility raises the revenue required to cover fixed cost. Capacity that cannot be sold profitably is a claim on cash, not an advantage.
Before approving expansion, leadership should be able to show:
- The existing constraint is real and persistent.
- Forecast demand has a credible path to contracted volume.
- The programs expected fit the proposed equipment and facility.
- Price and contribution margin support the full investment, not the equipment quote.
- Ramp time, qualification, staffing and working capital are funded.
- The downside case leaves the company able to operate.
◆V. A purchase order is evidence
A purchase order is evidence. A forecast is a hypothesis. A market-size slide is context. They should not carry equal weight in an investment decision, and in most boardrooms they do.
The public record shows what happens when the weights slip. One large public CDMO reported investing more than a billion and a half dollars in its facilities over three fiscal years, then disclosed productivity problems and higher-than-expected costs at certain of those sites, and a single customer at roughly a sixth of its net revenue. Capacity, cost and concentration arrived together, as they usually do. The names do not matter; the pattern does, and it is not confined to large companies. Law II — Capital Follows Evidence™.
◆VI. The strategic sequence
The question is rarely "OPEX or CAPEX?" A scaling CDMO needs both. The question is which investment comes first, at what size, and against what proof.
Start by making existing capacity visible. Find the constraint at the level of the line, the process, the shift, the laboratory or the release step. Determine whether lost output comes from physical limits or from scheduling, changeovers, yield, staffing, maintenance and quality delays.
Then improve the operating system. Put capable people, process controls, data and customer discipline around the assets already owned. Measure the throughput and margin gained. Principle 05 — Reinforce Before You Load™.
Only then commit to a larger physical ceiling, unless a specific customer requirement or capability gap justifies an earlier investment. Stage the commitment wherever possible: a pilot capability before a full line, a modular expansion before an entire facility, a customer-backed buildout before speculative capacity.
This is the law behind the sequence: reduce uncertainty before increasing commitment. Law V — Buy Information Before You Buy Scale™.
◆VII. Two ways to scale, two ways to fail
OPEX-led scaling extracts more reliable, more profitable output from the present footprint. It can improve the customer's experience and raise the return on existing assets. It fails when spending becomes a permanent layer of cost with no measurable gain in throughput, quality, pricing power or retention.
CAPEX-led scaling adds capacity or capability the current footprint cannot provide. It can widen the customer base the company can serve and create operating leverage as utilization rises. It fails when leadership builds for projected demand that arrives late, at lower margin, or not at all.
The strongest CDMOs combine the two in the right order. They develop a commercial pipeline and an operating process capable of filling new assets, then invest before the constraint damages service or turns away attractive business. They know utilization alone is not enough: a fully booked line running low-margin, high-complexity work can destroy value. They measure profitable, reliable utilization, and nothing less.
◆VIII. Concentration, transfers, release and cash
Four things turn impressive revenue growth into a fragile CDMO, and none of them shows up on the utilization chart.
Customer concentration: one program that is a third of the plant is a third of the plant's risk, and a buyer prices it that way. Trap 05 — The Big Customer Trap™. Complex transfers: every program that takes three runs to reach a repeatable batch consumes the capacity the plan assumed was free. Slow release: product that is made but not released is inventory wearing a label, and the customer's invoice waits with it. Working capital: a CDMO carries raw material, work in process and finished goods for customers who pay in sixty days; growth widens that gap before it closes it.
An expansion plan that does not name its exposure to each of these is a plan for a bigger version of the same fragility.
◆IX. The capital test
For every major OPEX or CAPEX proposal, put one page in front of the leadership team.
| Question | Evidence required |
|---|---|
| What constraint are we solving? | Baseline capacity, downtime, yield, cycle time or missed demand |
| What changes after the investment? | Specific operating and customer outcomes, with the measure named |
| What is the full cash commitment? | Initial spend, ramp losses, hiring, maintenance and working capital |
| Where does the return come from? | Incremental contribution, avoided cost, retention or new capability |
| When does the cash come back? | Ramp schedule, payback and the downside case |
| What must be true? | Demand, qualification, pricing, execution and customer-concentration assumptions |
| What is the next decision gate? | The milestone that earns the next tranche of capital |
If the answers are vague, the investment is not ready. That does not mean the opportunity is bad. It means the company has more uncertainty to remove.
◆X. What we believe
We believe a CDMO is built through the interaction of assets, people, process, quality and commercial judgment.
We believe operating discipline can create capacity before a company buys more of it.
We believe physical expansion can create substantial value when demand and execution support it.
We believe a facility is not a strategy, utilization is not profit, and EBITDA (earnings before interest, taxes, depreciation and amortization) is not cash.
We believe customer concentration, complex transfers, slow release and working capital can turn impressive revenue growth into a fragile enterprise.
We believe capital should follow evidence.
And we believe the goal is larger than building a bigger CDMO. The goal is a company whose capacity is dependable, whose economics improve with scale, and whose value a buyer can verify.
- We know which constraint we are funding before we fund it.
- We treat depreciation as a reporting convention, not a discount.
- We release the capacity we already own before we buy more.
- We weigh a purchase order, a forecast and a market slide differently, and say so.
- We stage every commitment: pilot before line, module before facility, customer before speculation.
- We measure profitable, reliable utilization, never utilization alone.
- We name our exposure to concentration, transfers, release and working capital in every expansion plan.
- We put one page in front of the leadership team, and if the answers are vague, we wait.
Dr Scott Kimball · 28 September 2026. Public examples from a filed annual report, cited at the end; the point of view is Scalewrights'. Illustrations are illustrations.
Sources: Catalent, Inc., Annual Report on Form 10-K for the fiscal year ended 30 June 2024 (filed with the SEC): "We invested $1.56 billion in our manufacturing and development facilities since fiscal 2022 for improvements and expansions, including $327 million in capital expenditures during fiscal 2024"; risk factor on "productivity issues and higher-than-expected costs at certain of our facilities"; "one customer accounting for approximately 17% of net revenue." Read 28 Sep 2026. The example is used for the pattern, not as a judgment of the company. Every figure in the body other than these is an illustration.