The Kill Fee Clause: How Agencies Get Paid When Clients Ghost Mid-Project

Every agency owner has the same scar. A signed statement of work, a team spun up, two sprints of momentum — and then a Slack message that starts with "Hey, so we need to pause things for a bit." Six weeks later the client is gone, and you're eating the bench cost.

The fix isn't better clients. It's a kill fee clause that makes cancellation expensive enough to be a real decision, not a casual one. Done right, it protects your margins without souring the deal at signing. Done wrong, it either gets negotiated out during redlines or triggers a legal fight you don't want. This is how we've come to write ours after enough painful cycles.

Why "just bill for work done" isn't enough

The default agency posture is: if a client cancels, we invoice for hours delivered and move on. That sounds fair. It is not.

When a client pauses mid-project, you lose three things, not one:

1. Bench time — engineers you already committed to the project can't be redeployed instantly. Two to four weeks of salary bleed is typical.
2. Pipeline damage — you turned away other work to take theirs. That opportunity cost doesn't show up on any invoice.
3. Context tax — if they come back in three months, you'll spend a week reloading context, re-onboarding, and reconciling stale code. Nobody pays for that.

A time-and-materials invoice covers none of this. A milestone-based fixed price covers even less, because you may have burned 60% of the effort on a milestone that was 30% due to be invoiced.

The kill fee closes that gap. It's not a penalty. It's the price of the optionality the client is exercising when they walk away.

What a kill fee actually is

A kill fee is a contractually agreed payment triggered when the client terminates the engagement for convenience — meaning not because you breached, but because they changed their mind, ran out of money, got acquired, or decided to build in-house.

The two knobs you control are trigger and quantum.

Trigger conditions

The cleanest trigger is: any termination not caused by uncured material breach by the agency. That covers pauses that stretch beyond a defined window (we use 30 days), unilateral cancellation, and "strategic pivots" that quietly drop your scope.

Watch out for the escape hatches clients' lawyers add:

• "Termination for changed circumstances" — vague and clients love it. Push back hard or narrow it to acquisition/insolvency only.
• "Termination for convenience with 30 days notice" with no fee — this is the default in a lot of MSAs. It's the clause you're specifically overriding.
• "Pause" vs "terminate" — define a pause. Ours: any suspension over 30 calendar days converts to termination automatically.

How much to charge

The number depends on when the cancellation happens. A flat fee is easier to negotiate but rarely fair. We use a tapered structure that mirrors our actual exposure:

Phase	Kill fee
Before kickoff (contract signed, team not started)	15% of remaining SOW value
Weeks 1–4	25% of remaining SOW value + all work delivered
Weeks 5 through midpoint	20% of remaining SOW value + all work delivered
Past midpoint	15% of remaining SOW value + all work delivered

The curve dips after the early weeks because by then you've been paid for real deliverables and your bench-reallocation risk is lower. The early-phase fee is higher because that's when your exposure is worst — you've committed people and haven't invoiced much yet.

Equity-heavy or discounted engagements should have proportionally higher kill fees, because your effective hourly rate is already subsidised. Losing the back half of a discounted build is a double hit.

How to write the clause without spooking the client

The biggest mistake is presenting the kill fee as a punishment. Frame it as team commitment cost, because that's what it actually is.

Here's the plain-English version we open with in conversations, before the legal draft:

"To hold a team of four engineers for your project, we turn down other work. If you cancel, we can't instantly refill that capacity. The kill fee covers the ramp-down cost so we can keep our team employed while we find replacement work."

Nobody argues with payroll. They argue with "penalty."

In the contract itself, keep it surgical:

8.3 Termination for Convenience.
Client may terminate this SOW for convenience upon fifteen (15) days
written notice. Upon such termination, Client shall pay:
(a) all fees for Services performed and Deliverables accepted through
    the effective termination date; plus
(b) all non-cancellable third-party costs incurred by Agency; plus
(c) a Team Commitment Fee calculated as a percentage of the unbilled
    remaining SOW value, per the schedule in Exhibit B.

A suspension of Services exceeding thirty (30) calendar days shall be
deemed a termination for convenience under this Section 8.3.

Put the tapered schedule in an exhibit, not the main body. Exhibits get less scrutiny during redlines.

Redline survival tactics

Expect three pushbacks:

• "Cap it at fees paid to date." No. That defeats the entire point. Counter with a hard cap at 25% of total SOW value if they need a ceiling.
• "Remove the auto-conversion of pauses." Compromise: extend the pause window to 60 days but keep the conversion.
• "We need mutual termination rights." Fine, but your termination-for-convenience right shouldn't trigger a fee against you — you're the one delivering. Make it asymmetric and defend it.

If a client refuses any kill fee at all, that's data. It usually means they're already unsure they'll finish the project. Price that risk in elsewhere or walk.

Enforcement is a separate problem

A clause you can't collect on is a rhetorical device, not protection. Two things matter:

Invoice on trigger, immediately

The day termination is confirmed in writing, send the kill fee invoice with a clear reference to the clause number and the calculation. Don't wait. Don't "let them come back to us with a proposal." Delay signals negotiability.

Deposits and retainers as collateral

This is where a lot of agencies underprotect themselves. We now require an upfront deposit equal to roughly one sprint of work on any engagement over a certain size, held against the final invoice. If the client cancels, the deposit is applied against the kill fee first. You're not fighting to collect — you're just not refunding.

For larger contracts, a monthly retainer paid in advance does the same job. You're always one month ahead, so a cancellation means you keep the current month plus whatever the kill fee schedule dictates.

When not to enforce it

Contracts are levers, not laws. There are cases where enforcing the full kill fee is technically correct and strategically stupid:

• The client's founder just got fired and the new CEO is pausing everything. They may come back in six months as a bigger contract. Waive part of the fee in exchange for right of first refusal on the resumed work.
• The project failed partly because of your team's performance. Even if you'd win the argument, don't have it. Bill for delivered work, drop the fee, and preserve the reference.
• The client is a strategic logo you'd chase for free. Sometimes the marketing value of not being the agency that sued them is worth more than the fee.

The point of the clause isn't to always collect. It's to give you the right to collect, which changes every conversation that follows.

Where we'd start

If you don't have a kill fee clause today, don't try to retrofit it into active contracts. Start with the next SOW. Draft the tapered schedule this week, walk your sales lead through how to frame it as team commitment cost, and put a 15% deposit requirement on any new engagement above whatever threshold matches your cash cycle.

Then — and this is the part most agencies skip — track cancellations for the next four quarters. Note when they happened, what phase, and what you actually collected versus what the clause entitled you to. That data tells you whether your tapered percentages are calibrated to your real exposure, or whether you're still subsidising clients who change their minds. For more on structuring engagements around risk, our services page has a breakdown of how we scope different contract types.