FounderWiseDecisions, not feeds
← All articles FounderWise · Long-form

The supplier is cheap because of something

They are thirty percent below anyone else and they have missed three deliveries. Before you decide whether they are failing, work out what the price was buying and which of their failures can end you.

13 Nov 2026 13 min read By Joshua Pi’Rwot
Share X LinkedIn

Your main supplier is thirty percent below every alternative and has missed three of the last eight deliveries. Switching means a month of qualification and a higher unit cost you have not budgeted.

The debate in the room is whether three misses is a pattern. The prior question is what the thirty percent was buying, because a price that far below the market is a description of something that was left out.

Why these three models

The decision is whether to keep, fix or replace a supplier who is cheap and unreliable. The features that fire are a price that carries information about what is not included, a run of failures that is hard to distinguish from ordinary variance, and a set of consequences that differ enormously in severity.

Three lenses. Selection produces an equilibrium answer about what the price told you. Delay produces a complex answer about whether you can read the pattern yet and why your responses overshoot. Ruin sizing produces a random answer about which failure actually matters. The third should be settled first, because it decides how much the other two are worth spending on.

1. A price that low is a statement

Start with the discount, because it is the piece of evidence you already have and have never interpreted.

A supplier substantially below the market is not usually being generous or inefficiently priced. They are supplying something different, and the difference is what the discount is made of.1 The candidates are specific and checkable: less buffer stock, so a demand spike becomes your problem. No redundancy in their own supply, so one upstream failure passes straight through. Thinner quality control. No insurance. Informal labour arrangements. A currency or regulatory position they are carrying and you are not seeing.

Which means the three missed deliveries may not be a decline at all. They may be the thing you bought, arriving on schedule.

The productive move is to find out which. Ask them directly what they do differently to reach that price. Suppliers answer this more often than founders expect, because from their side it is a legitimate commercial explanation rather than an admission: we hold less stock, we use a single upstream source, we do not carry the certification the others do. Any of those is an answer you can act on.

You did not get a bargain. You bought a different product and the invoice did not itemise the difference.

And where the answer is that they simply run leaner, that is real and it is also fragile in a specific way: a firm with no margin has no capacity to absorb a bad quarter, which makes their reliability a function of their own conditions rather than of their intent.

Some will not answer, and the refusal is readable too. If they claim the price is simply their efficiency and cannot name one concrete thing they do differently, price the discount yourself. Ask them to quote a small penalty for late delivery, say two percent of order value per week. A supplier carrying real buffer stock will take that clause, because they expect never to pay it. A supplier running with nothing behind them will decline it, renegotiate what late means, or offer a credit on the next order instead. One clause prices the buffer without anybody having to describe it.

There is a version of the discount that is neither leanness nor omission, and it is common here. They are cheap because they get paid faster than their competitors do. A supplier taking mobile money on delivery instead of waiting sixty days for an invoice is financing far less working capital and can price accordingly. If that is the answer, you are paying for the discount in cash timing rather than in risk, and the right response is to protect the payment terms rather than to go hunting for what was left out. Test it in one question: what does your price become at thirty days. If the gap is roughly the discount, you have your explanation and nothing else is missing.

2. Three misses is not yet a pattern, and waiting is not free

The second lens is about the evidence, and it cuts in both directions.

A short run of failures in a variable process is consistent with a stable supplier having a bad stretch and with a supplier in genuine decline. Distinguishing them requires more observations than you have, and the situation is made worse by a lag: whatever caused the misses happened weeks or months before you saw them, and whatever you do about it will show up weeks or months later still.2

That structure produces the familiar oscillation. You escalate, they improve for two deliveries, you relax, it recurs. The improvement was not caused by the escalation and the recurrence was not caused by the relaxing, because the loop is longer than the interval between your interventions.

The remedy is the same as for any lagged loop, which is to measure closer to where the state changes rather than to react faster. Delivery performance is the lagging outcome. The leading indicators are things like whether they confirm orders within the usual window, whether their own upstream is on time, whether the person you deal with has changed, whether their invoices have started arriving earlier than usual. Those move first and several of them are visible without their cooperation.

How many observations you would actually need is worth being honest about, because it is more than anyone waits for. In a process with real variance, three failures in eight is consistent with a wide range of underlying reliabilities, and the run length required to separate them grows with how noisy the process is.4 You will act long before the delivery record can tell you anything, which is why the leading indicators are the whole game.

Make one of those indicators a standing question rather than an inference you have to construct later. On every order, ask when their own upstream input landed and confirm the name of the person handling it. Two lines in the order email, logged in one sheet. Six orders of that gives you a series that moves before delivery performance does, and the first month the answer turns vague tells you more than the next three delivery dates will.

And set the decision rule in advance rather than in the meeting after the next failure. If misses exceed a stated rate over a stated window, we qualify an alternative. Writing that down before the next incident is what stops the oscillation, because the loop is otherwise driven by whichever delivery most recently arrived.

That rule has a failure mode of its own. A threshold set in a calm week will fire in your peak season, which is the worst moment to qualify a new supplier and therefore the moment you will override it. Write the exception into the rule instead of leaving it to the day. State the rate, state the window, and state that if the threshold is crossed inside the peak the action is to buffer immediately and qualify the alternative once the season closes. A rule you will break is not a rule, and the one you break is always the expensive one.

3. Which of their failures can end you

The third lens ranks the consequences, and it should be run before either of the others because it sets the budget for the whole question.

Supplier failures are not one risk. A late delivery is a cost. A late delivery in the week of your largest customer’s commitment is a relationship. A quality failure that reaches an end customer is a liability. A supplier who ceases trading with your inventory or your tooling in their possession is a different category entirely, and in several African supply chains a supplier’s loss of an import licence or a bank relationship stops your business rather than theirs.

Where one branch removes you from the game, the average across branches is not the decision-relevant quantity and the size of the worst case governs.3

So write the branches, then check which are live. Do they hold anything of yours. Is any single customer commitment dependent on one delivery from them. Would a quality failure reach an end user before you saw it. Is their own licence or banking position something you have ever verified.

Retrieve before you assess, not after. A supplier who knows you are weighing alternatives and is holding your tooling now has a bargaining chip, and the form this takes is never a refusal. It is a delay. The mould is in use. The operator is away. Next month. Ask for the item as routine housekeeping, an asset audit, before any conversation about performance begins. If that conversation has already happened, ask instead for a signed schedule of what they hold on your behalf. It costs nothing, it is easy for them to sign because it changes nothing today, and it converts a possession problem into a documented claim.

The answers change what you should do, and they usually do not point at switching. If the fatal branch is that they hold your tooling, the action is to retrieve the tooling, which is cheaper and faster than qualifying a new supplier. If it is a single customer commitment, the action is a buffer for that commitment specifically rather than for everything. Ruin sizing tends to produce narrow, cheap interventions, because it identifies the one branch that matters instead of improving the whole relationship.

What the three say together

  • Ask what they do differently to reach that price. The answer is usually given and it is usually specific.
  • Write the branches: cost, relationship, liability, cessation. Fix the fatal one narrowly rather than improving everything.
  • Watch leading indicators rather than delivery performance. Order confirmation times, their upstream, staff changes.
  • Set the switching rule in advance, as a rate over a window, so the decision is not made by the most recent delivery.

Where they disagree

The selection argument and the ruin argument disagree about whether the price is the problem.

Selection says the discount is the root cause: you bought a thinner service and the misses are the specification working as sold, so the fix is to pay for what you actually need, either with this supplier or another. Managing the symptoms of a deliberate omission is expensive and endless.

The ruin argument says the price is mostly irrelevant. What matters is whether any branch can stop the company, and a cheap supplier with no fatal branch is a manageable irritation while an expensive one holding your tooling is an unpriced exposure. Optimising reliability in general is spending against the wrong risk.

The reconciliation is sequential and it saves money. Close the fatal branch first, narrowly, because it is usually cheap and specific. Then decide about the price, with the ruin question already off the table, which frequently reveals that the cheap supplier is perfectly acceptable once the one dangerous dependency has been removed. Founders usually run this in reverse, opening a supplier search because of missed deliveries while the tooling stays where it is.

What none of them contain

None of the three prices the relationship you have built. A supplier who has learned your specifications, tolerates your order patterns and answers the phone at short notice carries knowledge that a cheaper or more reliable alternative does not, and qualification costs are the visible part of a larger transfer.

None of them handles the supplier whose difficulty is temporary and known. A firm working through a currency shock or a port disruption may be the same firm in six months, and switching during their bad quarter forfeits a relationship for a condition that has an end date. Asking them what is happening is the cheapest way to find out and it is skipped remarkably often.

And one property the ensemble will not produce: you are also a customer they are assessing. A buyer who escalates on every miss, pays slowly and holds no forecast is one whose orders get filled after everyone else’s when capacity is short. Some portion of your reliability problem may be your position in their queue, and that is a variable you control.

The one action that survives the ignorance: this week, find out whether they hold anything of yours, tooling, inventory, data or a certification registered in their name. That is the branch that turns a supplier problem into a company problem, and retrieving it is almost always cheaper and faster than switching supplier.

Who has to move

The founder, because the fatal branch sits outside whoever manages the supplier day to day and is invisible from inside the delivery conversation. The instinct after three misses is to open a search for alternatives, which is a quarter of work aimed at the least dangerous version of the problem. The cheapest first test is two questions asked directly: what do you do differently to reach this price, and what of ours is in your possession. Both are answered in one call and they reorder everything that follows.

Sources and notes

  1. Jeffrey Carpenter and Andrea Robbett, Game Theory and Behavior, MIT Press. Adverse selection, and the general result that price carries information about unobserved quality when the seller knows more than the buyer, are developed in the chapters on asymmetric information. Used in section 1 for reading a substantial discount as a description of what has been omitted rather than as a bargain.
  2. John D. Sterman, Business Dynamics: Systems Thinking and Modeling for a Complex World, McGraw-Hill, and Donella H. Meadows, Thinking in Systems: A Primer, Chelsea Green Publishing. The behaviour of a balancing loop containing a delay, including the oscillation produced when interventions are spaced more closely than the loop’s response time, is developed in both. Used in section 2, with the standard remedy of measuring closer to where the state changes.
  3. Michael J. Mauboussin, The Success Equation: Untangling Skill and Luck in Business, Sports, and Investing, Harvard Business Review Press, with the standard treatment of absorbing barriers. The principle used in section 3 is that where one branch removes you from the game, expected value across branches ceases to be decision-relevant. No probability is assigned to supplier failure here; the argument requires only that the branches differ in kind rather than in degree.
  4. Michael J. Mauboussin, The Success Equation, Harvard Business Review Press. The difficulty of separating skill from luck in a short record supports the caution in section 2 that three misses in eight deliveries is consistent with both a stable supplier having a poor run and a genuine decline. The number of observations required to distinguish them depends on the underlying variance of the process, which is why the article recommends leading indicators rather than a longer wait.

A note on a number this article does not give. There is no miss rate at which a supplier should be replaced. It depends on the variance of their process, what a miss costs you, and whether a fatal branch is live. What transfers is the order: close the fatal branch, then read the leading indicators, then decide about the price.

Joshua Agonya Pi’Rwot, Founder.

Lock in your calls.

You’ve marked 0 of 5. Now choose how often you want the signals.

Step 1 · Pick your cadence

The DispatchWeekly · your Monday 5 callsFreealways

Step 2 · Where to send it

Personalize your BriefThe Brief

Tune every edition to the markets and industries you actually act on.

🔒 Unlock personalization — The Brief, $19.99/mo →
Free Dispatch forever · upgrade anytime · we never share your details.
Know where you stand?
The Dispatch tells you what changed. Knowing what to do about it is a different question, and it is the one FounderWise answers. Start with the free Traction Audit: 12 questions, about 3 minutes, scored out of 100.
Find out where you stand →

For teams, syndicates & programs

Recommended
Team
$15/seat · mo
Daily Brief for the whole team (min 3 seats).
  • Everyone on the same signal
  • Admin + shared watch-list
  • One invoice · ~25% off solo
Get Team →
Cohort License
$1,500/cohort
One cohort of up to 30 founders, for accelerators, funds & programs.
  • MVP Studio and GTM Studio for every founder
  • Your logo, your cohort
  • A funder report on what each founder did
Talk to us →
Pass the Dispatch on
Know a founder making these calls blind? Send them this week’s five — free, every Monday.

Decisions, not feeds. · Curated by Joshua Pi’Rwot · FounderWise · Free Audit · Store · parent of Business Growth Accelerator

Call committed. We’ll hold you to it.
Know where you stand →