Single-supplier risk: how to measure dependency and when to dual-source

01Single, sole and dual sourcing

Single sourcing is not a mistake. It often gives the best price and quality, and many lean supply chains rely on it deliberately. The problem appears when single sourcing is combined with long recovery times, low stock and no plan.

02Two lessons from real events

In March 2000 a fire hit a Philips semiconductor plant in Albuquerque, New Mexico, that supplied radio-frequency chips to both Nokia and Ericsson. Nokia noticed the irregularities in deliveries quickly, escalated, and secured capacity from Philips's other plants and from alternative suppliers. Ericsson, which relied on the plant as a single source, reacted later and suffered severe shortages of a key component in its phone business. The event is still taught as an example of how detection speed and alternatives matter more than the disruption itself.

In March 2011 the Tōhoku earthquake and tsunami in Japan damaged a Renesas Electronics plant in Naka that produced a large share of the microcontrollers used by the global car industry. Many carmakers discovered only then that their tier-1 suppliers, who looked diversified, depended on the same tier-2 source. Production was cut in plants far from Japan for months.

The common thread: the dangerous single sources are often invisible at tier 1.

03Measure 1: concentration (HHI)

The Herfindahl-Hirschman Index measures how concentrated purchases of an input are. Square each supplier's share of spend in percent and add them up. One supplier gives 10,000. Four equal suppliers give 2,500. The inverse, 10,000 divided by HHI, is the "equivalent number" of equal suppliers.

HHI is useful for scanning many inputs at once, but it does not tell you whether the suppliers share the same risks. Two suppliers in the same industrial district, or relying on the same raw material source, behave like one in a flood or a sanctions decision. Add a column for geography and for shared upstream sources.

04Measure 2 and 3: time to recover and time to survive

Researchers at MIT, led by David Simchi-Levi, proposed two measures that turn dependency into a planning number, published in the Harvard Business Review in 2014:

• Time to recover (TTR): how long it would take a supplier node to be restored to full function after a disruption.
• Time to survive (TTS): how long your operations could keep matching supply with demand after the node fails, using stock, alternative sources and reallocations.

Exposure exists when   TTR > TTS
Shortfall (units)    ≈ demand per day × (TTR − TTS)
Lost margin (€)      ≈ shortfall × contribution margin per unit

For a die-cast component with a 10-week recovery after a fire, three weeks of stock and no qualified alternative, TTS is three weeks and the gap is seven weeks of lost output for every product that uses it. For a packaging material with five suppliers within a day's drive, TTS is effectively unlimited even if one supplier fails.

Rank exposed inputs by lost margin. That list is your priority list.

05When a second supplier is worth it

Dual sourcing costs money. Typical costs are a higher unit price from splitting volume, the one-off cost of qualifying a supplier (samples, audits, tests, tooling), and ongoing management. The benefit is the reduction in expected and worst-case losses.

A simple decision rule compares annual cost with annual expected loss avoided, and then checks the bad case:

Expected loss avoided per year = probability of disruption per year × lost margin if it happens
Annual cost of dual sourcing   = price premium × volume + qualification cost / years of use + management
Dual-source if expected loss avoided > annual cost,
or if the worst-case loss is one the company could not survive.

The second condition matters. A disruption with a 3% annual probability may not justify a premium on average, but if it would stop the main production line for three months, the business case is insurance, not expected value. A Monte Carlo simulation gives both numbers from the same model.

06Cheaper alternatives to full dual sourcing

• Qualified dormant backup. Qualify a second supplier and give it a small share, perhaps 10 to 20%, enough to keep the relationship and the tooling alive.
• Strategic stock. Hold extra stock of the exposed input only, sized to close the TTR–TTS gap for the most likely disruption.
• Design for substitution. Redesign so that two different components can be used, which turns a sole source into a single source.
• Supplier's own resilience. Ask the supplier about multiple plants, its own backup sources and its business continuity plan; include obligations in the contract.
• Shared tooling and drawings. Own the tooling and the technical documentation, so moving production is faster.

07Map tier 2 for critical inputs

The Renesas case shows that diversification at tier 1 can be an illusion. For each exposed input, ask your suppliers where their critical inputs come from. You do not need the full network, only the nodes that many of your suppliers share. Our research page on single-supplier dependency shows how these shared nodes appear at country and sector level.