Photo: Google Gemini · AI-generated
Asking for a fixed price is reasonable. You want to know what this costs before you commit.
The problem is what a fixed price actually is. It is not a price for the work. It is a price for the work plus whatever the supplier thinks might go wrong, because they are the ones carrying that risk now.
On some projects that premium is small and worth paying. On others it is large, invisible, and paid whether or not anything goes wrong.
What you are really buying
A fixed bid has two numbers inside it. The honest estimate, and the padding.
You never see the split. A supplier who quotes without padding and then hits a surprise has three options: absorb the loss, cut quality quietly, or come back with a change request. All three are worse for you than having paid a fair premium in the first place.
So the padding is not greed. It is the rational response to being asked to price uncertainty. The real question is whether the uncertainty is large enough that the premium costs more than it protects.
Integration work is discovery work
Here is where fixed-scope reliably goes wrong.
If two systems already agree with each other, connecting them is buildable work and you can price it. But most integration exists precisely because two systems disagree — and the shape of that disagreement is not knowable until you are inside both.
We built a supply-chain intelligence platform for a garment manufacturer selling on Amazon, sitting between Zoho Inventory and Sellerboard and producing one reconciled view of stock, sales, cost and replenishment. The work was resolving SKU and variant mismatches between two systems that each considered itself correct.
Nobody could have specified that up front. Not the client, not us. You find out how two systems disagree by putting real data through them, and every answer changes the next question. That is discovery, and it ran on Time & Materials because that is what the work honestly was.
Had it been fixed-bid, one of two things happens. Either the estimate carries enough padding to cover a disagreement nobody has seen yet, or the first genuine surprise turns a delivery relationship into a contractual one.
What fixed-bid does to the working relationship
That second failure is the expensive one, and it is not really about money.
Under a fixed bid, every discovery becomes a negotiation about whether it was in scope. Both sides start reading the specification like a contract, because it is one. The supplier gets cautious about suggesting improvements, since anything not written down is a loss. The client gets cautious about asking questions, since every question sounds like a change request.
You end up with two teams optimising against the agreement instead of against the outcome. Nobody planned that. The commercial model produced it.
When fixed-scope is the right call
Often, and it deserves saying clearly.
Fixed-scope works when the work is genuinely known: a defined build against a settled specification, a technology the supplier has used many times, a well-drawn boundary with few external dependencies. It works when you need budget certainty more than you need the best possible outcome — a real and legitimate constraint. It works for a second or third build of something the supplier has already built once.
The signal is simple. If you can describe the finished thing precisely enough that two suppliers would build roughly the same system, fixed-scope is fine. If you cannot, you are asking someone to price a guess.
Making Time & Materials safe
The honest objection to T&M is that it looks like an open cheque, with the risk moved entirely onto the buyer. That objection is fair, and “trust us” is not an answer.
A supplier proposing T&M should be proposing structure with it: a cap you agree not to cross without a conversation, short cycles that end in something you can look at, a running view of spend against remaining work, and genuine stopping points — moments where cancelling costs you nothing further and you keep what has been built.
That last one matters most. Fixed-scope buys certainty by fixing the deliverable. Well-structured T&M buys it by making the decision reversible at frequent intervals. Both are legitimate ways to control risk. Only one survives contact with work nobody can fully specify.
If you are being asked for a fixed price on something genuinely unknown, we should talk.