The Fixed-Price Ceiling: When to Switch a Client to T&M Mid-Project

Every agency owner has lived this project: the SOW was tight, the estimate was defensible, the client signed with a smile. Twelve weeks in, you're burning cash on a feature that nobody scoped, the PM is drafting change requests nobody wants to read, and the engineering lead is quietly updating their CV. Fixed-price didn't fail because you estimated badly. It failed because the shape of the work changed, and the contract didn't.

This is a playbook for spotting that moment early and moving the engagement to time and materials without blowing up the relationship.

Why fixed-price breaks (and why it's still worth selling)

Fixed-price contracts exist because non-technical buyers need a number they can defend to their board. That's a legitimate need. The problem is that fixed-price only works when three things are true at signing:

• The problem is well understood by both sides.
• The solution space is narrow — there's basically one reasonable architecture.
• The client's own organisation is stable (no reorg, no new CTO, no pivot mid-flight).

When all three hold, fixed-price is the cleanest commercial instrument you have. When even one cracks, you're subsidising the client's ambiguity with your margin. The trick isn't to stop selling fixed-price. It's to build an early-warning system for when the assumptions have quietly stopped being true.

The four signals that a fixed-price is about to underwater

We track these weekly on any fixed-scope engagement over about six weeks long. One signal is noise. Two is a conversation. Three is a renegotiation.

1. The change request rate

Count formal CRs plus informal "could you just..." asks in Slack. If the rate is climbing week over week rather than flattening as the product takes shape, the scope is not converging. It's diverging.

2. The decision-maker drift

The person who signed the SOW is no longer in the meetings. A new stakeholder — often more senior, often newer — is now shaping the product. Their priorities are legitimate but they weren't priced in.

3. The dependency surprise

Mid-project you discover the client's auth system, data warehouse, or third-party integration is not what was described. Every fixed-price contains an implicit assumption about the environment. When the environment turns out to be different, the price was for a different project.

4. The velocity-vs-burn gap

You're 60% through the budget and 40% through the backlog. This one is the least ambiguous. If your burn rate is consistently outpacing your delivery rate for three sprints, you will not finish inside the envelope. Continuing to pretend otherwise is a management failure, not a resilience win.

The renegotiation conversation

Most agency owners lose this conversation before they open their mouth because they frame it as their problem. It isn't. A scope that has drifted is the client's problem too — they just haven't been shown the maths yet.

Run the meeting in this order:

1. State the original deal in one sentence. "We agreed a fixed price of X for scope Y, delivered by Z."
2. Show what has actually changed. A single slide, or a single page. List the CRs, the new stakeholders, the environment surprises. No blame, just diffs.
3. Show the two forecasts. One assumes you push through on the original price. That forecast should honestly show which features get cut, which quality bars slip, and which risks compound. The second assumes a switch to T&M from a specific date forward, with a rate card and a not-to-exceed cap.
4. Recommend the second one, and explain why it's better for them. Fixed-price at this point means you optimise for finishing, not for their outcome. T&M means you optimise for their outcome, with a cap that protects their downside.
5. Ask for a decision by a specific date. Usually one week. Ambiguity here kills more projects than the price does.

Do not do this over email. Do not do this in a status meeting. Book a dedicated 45 minutes with the economic buyer.

The contract mechanics

The legal switch is simpler than most people think, but the sequencing matters. You want an amendment, not a new contract, so all the IP, liability, and confidentiality terms carry forward untouched.

A usable amendment structure:

Amendment 1 to SOW dated [date]

1. Effective from [date], the pricing model for work performed under
   this SOW converts from fixed price to time and materials.
2. Work completed and invoiced prior to [date] is unaffected.
3. Remaining fixed-price balance of [amount] is credited against
   T&M invoices until exhausted.
4. Rate card: [roles and hourly/daily rates].
5. Not-to-exceed cap: [amount] without written approval.
6. Weekly burn report delivered every [day] covering hours by role
   and remaining cap.
7. Either party may pause the engagement with 5 business days notice.
   All other terms of the original SOW remain in force.

The crediting mechanic in point 3 is the one clients respond to. It says: you're not being asked to pay twice. The money you've already committed still buys work — it just buys it at an honest rate now.

The weekly burn report in point 6 is non-negotiable. T&M without visibility is where trust dies. Send it every Friday, same format, same time, whether the client reads it or not.

What we tell the engineering team

A pricing change is not just a commercial event. It changes how the team should work.

Under fixed-price, engineers optimise for scope closure. "Is this in the SOW? No? Then no." Under T&M, that stance makes you look mercenary. The team needs to switch to optimising for client-visible progress per day — small, demoable increments, clear standups the client can drop into, honest flagging of dead ends before they eat a week.

We usually run a 30-minute internal reset when a project converts:

• The commercial model changed; the technical bar did not.
• Every hour is now visible. Log accurately. If you spent two hours on a spike that failed, that's fine, log it and say so.
• Say no to gold-plating. T&M does not mean unlimited scope; it means transparent scope.
• Escalate the moment a task looks 2x its estimate. Not at 3x, not at end of sprint.

This matters because a T&M engagement that feels loose to the client will not get renewed, no matter how clean the invoice is. Discipline is the product now.

When not to switch

Sometimes the right answer is to eat the loss and finish on the original terms. Specifically:

• The overrun is small (under ~15% of budget) and the relationship has long-term value.
• You made a genuine estimating mistake and the client's scope hasn't actually moved.
• The client is a reference account or case study you need more than the margin.

Be honest with yourself about which category you're in. "Strategic loss" is a legitimate call. "Hoping it'll be fine" is not.

Where we'd start

If you're reading this because a current project is wobbling, do one thing tomorrow: build the two-forecast slide. One page, original-price outcome versus T&M-conversion outcome, both honest. You don't have to send it yet. Just having it drafted will tell you whether you actually believe the fixed-price is salvageable, or whether you've been managing hope instead of the project.

If you want a second pair of eyes on a live commercial situation, our delivery team has run this conversation more times than we'd like to admit, and the pattern is almost always the same. Spot it early, frame it as their problem too, and put the amendment on the table before the trust runs out.