Single-Source Risk and Second Supplier Calculator

Expected cost, and how it differs from a guess

“It would be very bad if this supplier let us down” does not produce a decision, because there is no number to set against it. Expected cost produces exactly that number: the annual value, today, of an event that may not happen.

Expected annual outage cost = Outage probability × Outage days × Daily cost

Each of the three inputs has its own difficulty. Probability is not measured, it is estimated — but it can be estimated: how many deliveries ran late last year, how many times the supplier said “no stock”, whether production sits in one plant, whether imports come through one port. Days are usually underestimated: finding a new supplier is not the slow part, approving one is. Daily cost is already known in most companies — downtime cost, contractual penalty, or the premium paid when something has to be bought in a hurry.

The tool’s most useful output is not the verdict but the break-even probability, and it is worth knowing exactly how it is built, because it is not simply “cost ÷ exposure”. The residual risk that survives with two sources has to be carried too:

Break-even outage probability = (Annual cost of the second source + Expected annual cost that remains with two sources) ÷ (Outage days × Daily cost)

Read it as a threshold: if your own estimate of the outage probability is above that figure, a second source pays — holding the residual probability you entered constant. Saying which side of a threshold you are on is far easier than defending an abstract percentage.

A risk matrix: which item sits in which cell

You do not build a second source for every item — if you try, the buying team spends its time keeping second sources alive instead of buying. Splitting items on two axes puts the effort where it belongs: the impact of an interruption and how easily an alternative is found.

The matrix is a decision frame for prioritising effort, not a rule from any standard. In most companies the number of items that land in the bottom-right cell is smaller than expected — list them first, then run the calculation only for those.

The costs of a second source that nobody budgets

The set-up cost is usually estimated correctly; the cost of keeping the source alive is the one that gets missed. A second source is not something you build once and put on a shelf — it decays unless it is used.

• The discount lost when volume is split. The same total spend divided between two suppliers drops you a tier with both. This is usually the largest item, and because it never appears on an invoice it is rarely counted.
• A second quality file. Samples, test reports, certificate tracking and audits double, and they repeat every year.
• The keep-alive order. A second source that never receives an order will say “we don’t know you” on the day you need it. The cost of one small order a year is the subscription fee for a second source.
• Two different products. If the two suppliers’ goods are not identical, two separate records appear in the warehouse, in production and in the specification; a substitution difference is not something you write down once and forget.

Against that, a second source also has a benefit on the price side that is usually left out of the sum: the existence of a genuine alternative caps what the first supplier can ask for. That benefit was left out of the calculator because it cannot be measured — but read the net figure knowing that it works in favour of the second source.

“Two sources” is not always two sources

The easy way to work out the chance of both suppliers failing at once is to multiply the two probabilities — 15% × 15% = 2.25%. That multiplication is only correct if the two sources are independent of each other, and in purchasing they usually are not.

• They may both be distributors for the same manufacturer — when the manufacturer stops, both stop.
• They may both import from the same country through the same port; a customs or freight crisis hits both.
• They may both depend on the same raw material; in a shortage the source you were holding as an alternative has none either.
• They may both depend on the same document — when an import licence or a permit is suspended, neither can supply.

That is why the calculator does not do the multiplication for you: it asks. If even one of the points above applies, the residual probability you enter should be markedly higher than the product of the two.

⚠ For the same reason, choose the second source for where it differs: a different manufacturer, a different country, a different port or a different raw material. Finding a second name does not reduce risk; finding a different route does.

Three cheap measures to use instead

When the calculation says no, the answer is not to ignore the risk. Three measures cost a fraction of a real second source and are enough everywhere except the bottom-right cell of the matrix.

1. An approved standby, with no order placed. Identify the second supplier, approve their sample, get their current price and lead time in writing — but place no order. Costs almost nothing and cuts the switch-over from weeks to days.
2. Targeted safety stock. Not on every item, only on those in the right-hand column of the matrix (few alternatives). It does not shorten the interruption, it reduces its impact.
3. A notice clause in the contract. An obligation on the supplier to give advance warning of a stock-out is the cheapest measure that shortens an interruption; it costs one sentence.

What the tool says when the answer is no

The calculator is built to be able to say no, and to say so out loud rather than quietly trimming a number. Five cases are reported in words instead of a figure.

• The second source does not reduce risk. If the expected cost with two sources is not lower than with one — the residual probability and the switch-over days you entered multiply out to as much as, or more than, the single-source case — the tool says the residual probability you entered does not reduce risk, and prints no break-even.
• The break-even probability exceeds 100%. A probability cannot go above 100%, so instead of printing “104.9%” the tool says that no outage probability can justify this cost.
• The amortisation period is 0. The set-up cost cannot be spread over years, so nothing is calculated and the reason is stated.
• Outage days or daily cost is 0. The denominator of the break-even is zero, so the tool says the break-even cannot be calculated rather than showing a meaningless figure.
• The recurring cost eats the saving. Payback on the one-off set-up is the set-up cost divided by the annual risk saving minus the recurring annual cost. If that difference is zero or negative there is no payback, and the tool says so instead of returning a negative number of years.

How we handle this

On our side the same calculation runs in the other direction: if we can only supply an item from a single source, we say so in the quote, because the buyer’s risk is our risk too. On critical lines we take written stock and lead-time confirmation before quoting, and where an equivalent is acceptable we price a second channel in advance.

If your enquiry contains an item that depends on a single source, say so when you request a quote — we will separate the lines that have an alternative from the ones that do not, and put them on separate lead times.

Frequently Asked Questions