Supply Risk and Dual Sourcing

Why it matters

Supply risk examines not only whether material arrives under average conditions but also how much of the need can be met in a defined disruption. Buying from two different firms appears to provide assurance. But if the two firms take the same critical component from the same upstream source, a shared disruption can affect both at once. The same transport gateway, power connection or special tool can create a similar dependency. A source evaluation therefore does not end with counting firm names. Material, capacity and access dependencies are traced all the way to the point of need. An unknown dependency is not accepted as independence. The aim is not to count every hazard but to show clearly in which events the chosen sourcing arrangement will work and in which it will fall short.

Capacity is not a single number either. A supplier's total capacity for all its customers, the capacity set aside for the part under review, or the capacity it can sustain under the approved quality condition can differ. A high quantity in a quotation document does not prove that it will be ready for use in the week of the disruption. Tool changeover, worker skill, material access and other orders all limit the same capacity. The comparison uses the same time unit; units/day and units/week are not added directly. Whether the share that looks idle in the normal plan is truly usable is verified. The difference between capacity left on paper and the material that can be produced in an emergency is one of the most important tests of the plan. A secondary source is only a possibility when its preparation is not complete.

Using two sources does not mean that each source can meet the whole need alone. Even if total capacity is sufficient in the normal case, the stoppage of one can create a large gap. The normal case and the disruption cases are therefore calculated separately. During the disruption, usable stock can buy some time; but it does not close the capacity gap permanently. If preparation time is needed for an emergency capacity increase, the gap during that time must also be visible. Options such as postponing demand, switching to another product or opening an independent additional source are evaluated under their own approval conditions. That one option is better does not mean it is flawless. The choice is made once the remaining gap and the response time are known.

A probability calculation is useful only if the dependency structure of the events is known. Directly multiplying the disruption rates of two sources contains an independence assumption. If the stoppage of the same upstream source affects both firms, this assumption is invalid. Where reliable probability data is lacking, a scenario calculation can still be done: which event takes how many units of capacity out for how long? This question often provides enough clarity to make a first decision. A risk plan should show which information has been verified and which is open, instead of producing the appearance of numerical certainty. A backup source that has been tried, a dependency monitored over time and a current capacity commitment together create assurance. The mere presence of a second firm in the contract does not.

Formula