Dual sourcing vs single sourcing

What the two models mean

Single sourcing means one supplier is approved and receives the full volume for a part. Dual sourcing means two approved suppliers share the requirement, either on a fixed split or with one as a primary and one as a standby. The terms describe the supply base, not the number of quotes. Collecting three quotes and awarding all of the business to one supplier is still single sourcing.

The distinction matters because the cost and risk sit in different places. A single source concentrates tooling, learning and volume in one relationship. A dual source spreads the risk but also spreads the volume, which can push both suppliers above their most efficient batch sizes or below the threshold where the part is worth their attention.

The case for a single source

A single source is usually the cheapest way to buy a part. The full volume sits with one supplier, so setup, tooling and quality costs are recovered over more parts. The supplier has an incentive to invest in the relationship, and the buyer has one set of documents, one audit schedule and one point of contact.

For a part with low continuity risk, a stable design and a supplier with proven capacity, single sourcing is often the correct commercial choice. It is also easier to manage. Every additional supplier adds purchase orders, reconciliation, performance reviews and the risk that the two never produce identical parts.

The condition is that the risk is genuinely low or can be covered another way. If a single machine, a single site or a single special process carries the whole requirement, the concentration is a real exposure even when the price is attractive.

The case for two sources

Dual sourcing exists to answer a simple question: what happens if the primary supplier stops? A fire, a flood, a labour dispute, a machine breakdown, an export restriction or a sudden insolvency can interrupt supply with little warning. A qualified second source shortens the recovery time from months to weeks, and sometimes to days.

It also creates a competitive reference. A second supplier with current approvals gives the buyer a credible alternative at review time and a way to test whether the primary price is still fair. That competitive tension can offset part of the cost of splitting the volume, provided both suppliers know that performance, not tenure, decides the next split.

The benefit is only real if the second source is genuinely ready. A supplier that has quoted, signed an agreement and never made the part is not a second source. It is an option that still needs tooling, trials, first article approval and a production run before it can carry volume.

What splitting the volume costs

The obvious cost is the loss of volume advantage at each supplier. The less obvious costs are tooling duplication, duplicate first article and PPAP effort, two sets of audits and the engineering time to resolve differences between the two processes.

The table below sets out the trade-off in the terms that usually decide it. It is a decision framework, not a rule. The weights change with the part.

Qualification is the real cost

The qualification effort is where dual sourcing programmes fail. A second supplier needs the current drawing, the full specification, the approved material and finish, the inspection plan and the same acceptance criteria as the primary. It also needs a first article that is reviewed, not just received.

For an electrical part, add contact resistance, plating thickness and any functional test. For a structural part, add load or fatigue evidence. The second source should produce parts on the intended production tool and process, because a sample made on a different machine proves very little.

Then keep the approval alive. Drawings change, processes drift and tooling wears. A second source that was approved two years ago against revision C is not approved against revision E. Put both suppliers on the same change notification list and review the second source at the same cadence as the primary. If the supplier cannot show that the change was implemented, treat the approval as lapsed and requalify before placing volume.

1. 01

   Define the requirement

   Use the same drawing, specification and acceptance criteria for both suppliers.

2. 02

   Build or transfer tooling

   Decide whether the second source gets its own tool or a validated transfer.

3. 03

   Run a first article

   Produce on the intended process and review the full dimensional and functional report.

4. 04

   Run a pilot batch

   Make a production-sized lot to expose rate, yield and handling issues.

5. 05

   Keep it live

   Place a small recurring order so the process and the approval do not go stale.

A staged approach that limits the cost

Not every part needs a full second source from day one. A staged approach starts with a single source for the main volume and a qualified alternative held in reserve. The reserve supplier receives a small annual order, enough to keep tooling and approvals current without duplicating the full volume.

If disruption risk rises, the split can move toward the second source without a new qualification cycle. If the risk falls, the reserve order can be reduced. This keeps the insurance premium visible and adjustable instead of treating dual sourcing as a permanent 50/50 arrangement.

The staged model also gives the buyer real performance data from both suppliers. A second source that has never shipped in volume cannot be scored on delivery or quality, so the comparison stays theoretical. A small recurring order turns the alternative into a measured supplier and gives the buyer evidence for the next award.

Deciding by part, not by policy

A common mistake is to apply one sourcing policy to every part. A blanket dual sourcing rule multiplies tooling and qualification costs across a catalogue, including parts where disruption would be an inconvenience rather than a crisis. A blanket single sourcing rule leaves the critical parts exposed.

Sort the parts first. Criticality, annual spend, number of qualified alternatives, lead time, tooling ownership and the ease of re-sourcing should all influence the decision. High spend and high criticality justify a live second source. Low spend and easy substitution may not.

Write the reason down for each part. A short note that says why this part is single sourced, and what would trigger a second source, is more useful than a policy that nobody can apply at the drawing level.

Review the decision when the inputs change. A new customer commitment, a capacity constraint at the primary, a change in duty or a shift in annual volume can all move a part from one model to the other. The note should name those triggers so that the review happens before a disruption rather than after it.