Being the only supplier does not give you pricing power. It gives you a ceiling set by something most manufacturers never calculate, and the ceiling is usually far lower than the fear of the customer’s reaction suggests and far higher than cost-plus will ever produce.
That is the whole argument and it contradicts the two positions people usually hold. The first is that sole source means you can charge what you like, which is wrong and produces the pricing action that triggers a requalification programme. The second is that you cannot risk moving price at all because the relationship is too important, which is wrong in the opposite direction and is the more expensive of the two errors.
The method-selection question of when to price this way at all is covered in the cost-plus comparison. This is what to do once you have decided a product belongs in the value bucket and you are the only one making it.
Why does sole source not mean unlimited pricing?
Because your customer always has an alternative, and it is almost never a competing supplier. It is redesigning the product so they no longer need yours, insourcing it, qualifying a second source at cost, or accepting the failure mode your part prevents. Every one of those has a price, and the highest of those prices is your ceiling.
The word monopoly makes people think of unlimited power, and that is a consumer intuition rather than an industrial one. In a business-to-business relationship your customer has engineers, a supply chain function and a standing incentive to remove single points of failure from their bill of materials. What sole source actually buys you is time and switching cost, not immunity.
The practical consequence is that pricing here is bounded arithmetic rather than an open question, and the arithmetic is available to you if you are willing to go and find it.
There is a second reason the intuition fails, and it is about who you are actually negotiating with. Your counterpart in procurement is not deciding whether to pay your price. They are deciding whether to escalate, and an escalation about a sole-source supplier lands on an engineering director who has been asked for years to reduce single-source risk. A price move that is large enough to prompt that escalation has done something far more consequential than raise a price. It has put your part on a list.
Watch a related video: Contribution Margin vs P&L: The $4M Mistake.
What is the customer’s actual alternative?
One of four things, and you need to know which one before you name any number. The four are redesign, insource, qualify an alternative, and do without. Each has a different cost and a different timeline, and the cheapest one that is genuinely available to them is the one that sets your ceiling.
Redesign means they change their product so your part is no longer required. This is the most common alternative for engineered components and it is expensive for them: engineering hours, validation, possibly customer approval of their own. It is also permanent, which is why triggering it is the worst outcome available to you.
Insourcing means they make it themselves. Ask whether they have the equipment, the process knowledge and the volume to justify it. For most low-volume engineered items the answer is no, and that answer is worth knowing precisely.
Qualifying an alternative means finding somebody else who can make it. The relevant number is not whether such a supplier exists but what qualification costs them in time, testing, tooling and risk.
Doing without means accepting the consequence your part prevents, which applies mainly to aftermarket and spares. Here the alternative cost is the downtime, and it is usually enormous, which is why aftermarket is chronically underpriced.
Rank the four by cost before you do anything else, because only the cheapest one matters. A customer who could redesign for a hundred thousand dollars and requalify for four hundred thousand will redesign, and the four hundred thousand figure is irrelevant to your pricing. Businesses routinely estimate the most expensive alternative, feel reassured, and price into a ceiling that does not exist.
Note also that the four alternatives are not equally permanent. A requalified second source can be reversed if their quality disappoints. A redesign cannot. That asymmetry should make you far more cautious about triggering the redesign path than the requalification one, even when the requalification is cheaper for them.
Two of the four are frequently available and never considered by the supplier. Redesign is the one manufacturers underestimate most, because it feels remote from a purchasing conversation, and it is precisely what a procurement escalation triggers. Doing without is the one they overestimate, assuming the customer cannot possibly accept the consequence, when in aftermarket the customer has frequently already built a workaround you do not know about.
How do you find the ceiling?
Cost out the cheapest genuine alternative from their side, then discount it for the risk and delay they would carry in switching. The result is the maximum price at which staying with you remains the rational choice, and it is a number you can estimate to within a useful range in an afternoon.
Do this from their perspective rather than yours. If qualifying a second source would take them nine months, consume three hundred engineering hours and carry a risk of a field failure during transition, that is the cost of leaving. Price below it, visibly, and you have removed their business case for switching without giving away the difference.
Your salespeople know more of this than you expect. They have heard the customer’s engineers talk about validation timelines, they know which programmes are frozen and which are open for change, and they know whether anybody has mentioned dual sourcing. Ask them before you commission anything.
Where you genuinely cannot see the customer’s economics, do not guess generously. An unverifiable ceiling estimated optimistically is how a pricing action becomes a requalification programme, and the requalification is permanent while the price gain lasts a year.
Two sanity checks on any ceiling you produce. First, express it as a percentage increase over your current price and ask whether a buyer would treat a move of that size as routine or as an event. Anything that reads as an event needs to be phased, whatever the arithmetic says, because the reaction is triggered by the size of the step rather than by the resulting level. Second, ask what happens to their unit economics at the new price. If your component moves from two percent of their cost to five, somebody in their business now has a reason to care about it who previously did not.
How do you find the floor?
Fully loaded cost plus the return the product owes you for the capital, capacity and engineering it consumes, which is a different and higher number than cost plus standard margin. For a low-volume engineered item the loaded cost includes things standard costing usually misses.
Count the tooling, the qualification effort you already spent, the documentation you maintain, the inventory you carry, the setups the item forces, and the engineering time it consumes annually in changes and support. An item produced twice a year carries a year of overhead per production run and standard costing rarely reflects that.
The floor matters even when the ceiling is far above it, because the floor is what you defend when the customer pushes, and defending a number you have not calculated is how the conversation ends at cost-plus.
Be honest about which costs are genuinely attributable and which are already sunk. Tooling paid for eight years ago is not a cost of the next unit, and including it inflates a floor you will then defend unnecessarily. What belongs in the floor is what you will spend again: the setups, the inventory, the documentation maintenance, the annual engineering support, and the capacity the item occupies that something else could use.
That last item is the one that changes the answer in a constrained plant. If the line is full, the floor is not your cost, it is your cost plus the contribution you gave up on whatever you could not make instead. In an unconstrained plant that term is zero. Know which situation you are in before you set the number.
Where the customer’s economics are visible, the ceiling calculation is unusually reliable, because it rests on their arithmetic rather than on a judgement about elasticity. That makes this one of the few pricing situations where you can name a number and defend it with something other than an assertion about value.
Where in between should you land?
Closer to the ceiling than instinct suggests, and explicitly below it by a margin you can point to. The space between floor and ceiling is frequently enormous on sole-source items, which is why this category is the single largest pocket of recoverable margin in most industrial businesses.
Two considerations set the position. The first is how long your position lasts. A component protected by a qualification barrier that expires when their product is redesigned in three years should be priced to capture value inside that window rather than to be sustainable past it. A component protected by intellectual property or a genuine process capability can be priced for the long term.
The second is what the price does to their behaviour. Price at the ceiling and you have given their engineering team a business case for redesign, funded, with a payback they can defend internally. Price meaningfully below it and no such case exists. The gap you leave is not generosity, it is the mechanism that keeps you sole source.
Name the gap out loud in the conversation if the relationship supports it. Telling a customer that you are aware of what an alternative would cost them and have deliberately priced below it is an unusual thing to say and it lands better than any cost justification.
Move in steps rather than in one jump wherever the gap is large. A price that needs to double over three years is a very different proposition from one that doubles in a quarter, and the three-year version is frequently achievable where the immediate one would trigger the redesign you were trying to avoid. The arithmetic on the total recovery is close to identical and the risk profile is not.
The exception is a position with a known expiry. If the programme ends in eighteen months, phasing across three years captures nothing, and the right move is a single correction with a full explanation of why the current price was set where it was.
Watch for one signal above all others in the months after a move: any request for drawings, specifications, material certifications or test data beyond what the customer normally receives. That is what the early stage of a requalification looks like from your side, and it usually arrives as a routine administrative request from somebody you have never dealt with. Treat it as information rather than as paperwork, and route it to whoever owns the relationship the same day.
How do you hold it over time?
By writing down four things at the moment you set the price, because the erosion here is slow and nobody will remember the reasoning in three years. This is the part that fails in practice, long after the pricing decision was made correctly.
Record the ceiling estimate and its basis. The alternative you costed, the number you reached, and the date. When somebody proposes a concession in 2029 the first question should be whether the alternative has got cheaper, and without the record nobody can answer it.
Record the expiry of your position. Name the event that would end it: a redesign cycle, a patent expiry, a competitor’s capability investment, the end of a programme. Put a review date just before it.
Set a floor and an approver. Sole-source items attract concession requests precisely because the margin is visible, and a price with no floor beneath it drifts toward cost-plus one reasonable exception at a time.
Review annually against their economics, not yours. The right annual question is not whether your cost moved. It is whether their alternative got cheaper, which is what actually changes the ceiling.
One limitation worth stating plainly. Everything above assumes you can see enough of the customer’s operation to estimate what leaving you would cost them. For products sold through distribution, or into applications you never visit, you cannot, and the honest position is that you are estimating a ceiling you cannot verify. In that case price conservatively, move in smaller increments, and treat each increment as a test rather than as a conclusion.
About the Stagnation Assassin
Todd Hagopian is a Fortune 500 transformation executive who has generated $3B+ shareholder value across five turnarounds at Berkshire Hathaway, Illinois Tool Works, and Whirlpool. He is VP of Product Strategy and Marketing. Known as The Stagnation Assassin, he is the author of two published books: The Unfair Advantage: Weaponizing the Hypomanic Toolbox and Stagnation Assassin: The Anti-Consultant Manifesto. His blog is published in 15+ languages and read by operators worldwide. Bring him to your stage via his speaking page or connect with him on LinkedIn. His newest book is 10-Minute Transformation (February 23, 2027).