When a contract manufacturer says it has capacity for your programme, the answer is usually expressed as a number of lines, a number of placements per hour, or an overall equipment effectiveness figure. Of those three, OEE is the one that sounds most rigorous and is the one most often presented in a way that is technically true and practically misleading. Understanding how the number is constructed is what lets a buyer tell a genuine capacity position from a favourable calculation.
This matters more than it appears at the quotation stage. Capacity claims decide whether the supplier can support a production ramp, whether a second programme can be accepted without displacing the first, and whether the lead time offered is achievable or aspirational. A buyer who cannot interrogate the utilisation figure is relying entirely on the supplier's self-assessment at exactly the moment when that assessment carries the most commercial weight.
What OEE Actually Measures
Overall equipment effectiveness is a compound metric built from three independent ratios. It is defined as availability multiplied by performance multiplied by quality, and each factor answers a different question about how a machine or line actually behaved.
| Factor | Question It Answers | What It Captures |
|---|---|---|
| Availability | Was the equipment scheduled to run actually running? | Breakdowns, setup and changeover, waiting for materials, unplanned stops, operator absence |
| Performance | When it ran, did it run at its rated speed? | Reduced speed, micro-stops, feeder jams, downstream blocking, small delays not logged as downtime |
| Quality | Of what it produced, how much was good? | Scrap, rework, first-pass yield losses, retest and repair |
Because the three factors multiply, a weakness in any one of them suppresses the overall figure. A line that is available 95 percent of scheduled time and runs at 95 percent of rated speed but yields 90 percent good parts has an OEE of about 81 percent. That multiplicative property is why OEE is useful internally: it prevents a manager from celebrating high availability while ignoring yield. It is also why the same property makes the number easy to manipulate, because each of the three factors depends on definitions that are chosen by the party doing the reporting.
The Four Definitions That Change Everything
Two suppliers can both report 85 percent OEE on comparable equipment and be describing very different factories. The difference is entirely in how four quantities are defined.
What counts as scheduled time
Availability is a ratio, and its denominator is scheduled production time. A supplier who excludes planned maintenance, holiday periods and hours when the line has no work from the denominator produces a higher availability figure for the same physical output. This is defensible for measuring equipment as an asset but misleading for measuring available capacity, because the hours the line stood idle for lack of orders are exactly the hours a new customer would be occupying. The question to ask is whether unscheduled idle time appears in the denominator. If it does not, the OEE figure describes equipment health rather than spare capacity.
What is treated as the ideal cycle time
Performance compares actual output to what the equipment could theoretically produce at its rated speed. The rated speed is usually the vendor specification at optimum conditions, which is reached in short bursts rather than sustained over a production week. A supplier comparing actual output to the vendor's headline placement rate will report a lower performance figure than one comparing against a realistic demonstrated rate. Neither is dishonest, but they are not comparable, and the same physical line can swing several points of performance depending on which baseline is used.
Where the boundary of the line is drawn
OEE is measured over a defined system. If the boundary is a single placement machine, the metric excludes everything that constrains throughput in reality — the printer, the reflow oven, the inspection stations, the depanelisation step. A programme's actual output is limited by the slowest step in the sequence, not by the fastest machine in it. A supplier reporting high machine-level OEE may still be constrained by a printing process or a single reflow oven that all lines share. Buyers should ask for the line boundary, and specifically what the actual bottleneck step is and how it is loaded.
Whether changeover is inside or outside the measurement
Setup time is a legitimate availability loss and belongs in the metric. Some reporting practices treat changeover as a planned event excluded from the calculation, which flatters availability considerably on a high-mix line where changeovers are frequent. For a buyer introducing a new programme, changeover time is not a footnote — it is the single largest source of lost capacity in the first weeks of production, and it is the number that determines how fast the supplier can turn your orders. Ask what the typical changeover time is by product family, and how many changeovers the line performs per week. The measurement and improvement of changeover is covered in our guide to SMT line changeover.
Why a High OEE Is Not Automatically Good News for a Buyer
There is a counterintuitive point here that catches buyers out repeatedly. A supplier operating at very high utilisation has very little slack. If a line is genuinely running at 90 percent of its available hours against its own order book, there is limited room to absorb your ramp, your expedite requests, or a quality excursion that requires rework capacity. High utilisation is attractive on a quotation because it suggests demand and competence; it is a risk on a production schedule because it suggests fragility.
The relationship to look for is not the absolute OEE figure but the relationship between committed load and available hours. A supplier at 70 percent utilisation with a well-measured OEE and a credible improvement plan has real capacity to offer. A supplier at 92 percent utilisation is offering you hours that are already promised to somebody else, and your programme will compete for priority with theirs at every expedite. Neither situation is disqualifying, but they require different commercial conversations. The relevant question is straightforward: how many hours per week on the line that will run my product are uncommitted over the next quarter?
Key Takeaway: OEE measures how well a supplier uses the time it has. It does not measure how much time it has left for you. The figure to request is not the utilisation percentage but the uncommitted hours per week on the specific line and process that will run your programme, over the horizon of your ramp. That number is directly useful; a utilisation percentage is only a proxy for it.
The Auditable Signals Behind the Claim
A supplier who genuinely measures OEE can answer specific questions quickly and with records. A supplier who quotes a figure without measuring it will answer in generalities. The list below is a practical way to distinguish the two, and every item is something a competent operation already tracks for its own purposes.
| Question | What a Measured Answer Sounds Like | What Evasion Sounds Like |
|---|---|---|
| How is OEE captured on this line? | Names the data source — machine counters, MES logging, manual shift records — and the review cadence | Refers to the figure as a general industry benchmark or a target |
| What was last month's first-pass yield on comparable product? | Gives a number per product family, states how it is calculated and where it is recorded | Gives one blended number across all products, or describes yield as very high |
| What is typical changeover time and how many per week? | Gives hours or minutes by product family and a weekly count | Describes changeover as quick and optimised |
| What is the bottleneck step for my product type? | Names the specific process and its loading | Says the line is balanced or all steps have capacity |
| How many uncommitted hours per week next quarter? | Gives a number for the line and process in question | States they can always accommodate new customers |
| What stops the line most often? | Names the top recurring cause and what has been done about it | Says stoppages are rare and not a factor |
The value of these questions is not that any single answer proves or disproves the capacity claim. It is that a supplier with a real measurement system produces specific numbers without preparation, and a supplier without one produces adjectives. That distinction is worth far more than the OEE value itself, because it also predicts how the supplier will behave when a production problem occurs later: an operation that measures its own performance will tell you about a yield problem early, and one that does not will tell you when you ask.
How Utilisation Shows Up in Commercial Terms
Capacity and utilisation are not abstract. They determine four commercial parameters that appear directly in a quotation, and buyers should connect the two explicitly.
Lead time. Lead time is a function of queue, not of line speed. A line running at high utilisation has a long queue ahead of your order, and the quoted lead time reflects it. When a supplier offers an unusually fast lead time at high reported utilisation, either the queue assumption is optimistic, the order will be prioritised at the expense of another customer, or the utilisation figure was calculated on a base that excluded idle hours.
Changeover cost and minimum order quantity. Every changeover consumes productive hours. A supplier with expensive or slow changeovers will price small batches higher and set a higher minimum order quantity, because the setup cost has to be recovered somewhere. Understanding the changeover economics explains the MOQ structure and reveals where a buyer has negotiating room — for example, by accepting a longer lead time in exchange for batching with other work of the same family.
Responsiveness to expedite requests. A supplier with genuine slack can expedite. A supplier with none cannot, regardless of willingness. If expedite capability is important to your programme, the capacity question is the one that determines whether that capability exists, and it should be tested before it is needed rather than during a shortage. The commercial framework for these discussions is covered in our guide to total cost of ownership, and the capacity planning side is set out in SMT line capacity and throughput planning.
Continuity across a ramp. Programmes rarely fail at steady state; they fail during a ramp. The hours available to absorb a doubling of volume over six weeks are what determine whether a ramp succeeds, and those hours are precisely what utilisation measures. A supplier who can present a capacity model showing available hours against your forecast volume trajectory is demonstrating the capability that matters, which is not the same as reporting a high OEE.
What to Ask Before Committing a Schedule
Procurement Tip: Ask six questions and require specific answers. First, what is the OEE definition used, and does the availability denominator include unscheduled idle time. Second, what is the ideal cycle time baseline — vendor rated or demonstrated on this product family. Third, what is the line boundary, and what is the actual bottleneck process. Fourth, what is typical changeover time and weekly changeover count. Fifth, last month's first-pass yield on comparable product, per product family rather than blended. Sixth, uncommitted hours per week on the relevant line over the next quarter. A supplier who answers all six with numbers is managing capacity. A supplier who answers with descriptions is not, and the difference will show up during your ramp.
It is also worth asking how the supplier measures the metric, because the instrumentation determines what can be improved. A line with automatic machine data logging can identify micro-stops and quantify feeder-related losses. A line relying on manual shift records captures only the large stoppages and will systematically overstate performance. This is not a criticism of either approach in isolation — manual measurement is common and workable — but it defines the resolution of the information you are being given, and it is reasonable to know whether a capacity claim rests on automatic counters or on an operator's recollection.
The Bottom Line
OEE is a sound metric that is routinely presented in ways that make the figure look better than the underlying position. The three factors multiply, so the number is sensitive to how the denominator and the baselines are defined, and those definitions are chosen by the reporting party. The practical response for a buyer is not to distrust the metric but to ask for the definitions and the uncommitted hours, which are the quantities that actually constrain a schedule. A supplier at moderate utilisation who measures performance properly is a better partner for a ramp than one at high utilisation whose numbers cannot be interrogated, because the first can be planned against and the second can only be hoped for.
At Huaxing PCBA we track line availability, performance and yield by product family through the production record, review the data on a fixed cadence, and quote lead times against actual loaded hours rather than nominal line capacity. Where a programme has a ramp ahead of it, we will model available hours against the forecast volume so the schedule is based on measured capacity. Read our capacity and throughput guide or contact our engineering team to review your programme's capacity requirements before committing a schedule.
