The conversation goes like this. You quoted a fixed fee. Somewhere around week three the project stopped resembling the thing you priced — a new page here, a stakeholder with opinions there, a fourth round of feedback that reopened decisions you’d closed. You know you’re now working for roughly half your rate. And when you finally raise it, it comes out as “this has been a lot more work than I expected”, which lands on the client as a grumble rather than a business fact.
The difference between a grumble and an accepted change order is almost never how politely you ask. It’s whether you can point at something.
Why “it took longer” isn’t an argument
From the client’s side of the table, a fixed fee was the whole point. They accepted a number so they’d stop having to think about the number. When you come back asking for more, you’re not just asking for money — you’re asking them to reopen a decision they thought was closed, often with their own budget holder.
“It’s taken longer than I thought” gives them nothing to work with. It could mean the scope changed, or it could mean you underestimated, and from the outside those look identical. Most clients aren’t trying to get free work; they genuinely can’t see the difference between the project they commissioned and the one you ended up delivering, because they only ever saw the outputs.
Tracked time turns an assertion into a record. Not as a weapon — as the shared document that lets both of you look at the same thing instead of at each other.
Log against the shape of the quote
This is the part that has to happen before there’s a problem, and it takes almost no effort: structure your time entries to match the line items in your quote.
If your quote said discovery, design, build and two rounds of revisions, then those are your projects or phases in the tracker. When the conversation comes, you’re not producing a wall of undifferentiated hours — you’re producing a like-for-like comparison against the document the client already signed. “Design was estimated at 12 hours and ran to 31” is a specific, checkable claim about a line they approved.
Equally important: give out-of-scope work its own home from day one. The moment something arrives that wasn’t in the quote — the extra landing page, the emergency deck for a board meeting, the third stakeholder’s review cycle — it gets logged separately, under a name that describes it plainly. This is the entire mechanism. Work that sits in the same bucket as everything else becomes invisible; work logged apart from day one produces a list at the end that speaks for itself.
Do it even when you’re planning to absorb the cost. You may well decide to eat a small overrun for a good client — but you want that to be a decision you made knowingly, and you want the client to see the goodwill rather than never know it happened.
Have the conversation early, not at the end
The worst time to raise a scope conversation is when you’re delivering the final files. At that point it reads as an invoice dispute, the client has already mentally spent the budget, and you’ve lost all your leverage by finishing the work.
The right moment is the first time the data crosses a threshold you set in advance — say, when a phase passes 120% of its estimate, or when out-of-scope hours pass a day’s work. That’s early enough that the client still has choices, which is what makes the conversation collaborative rather than adversarial. Something like:
Quick heads-up before this goes further: the extra approval round and the two additional pages have added about nine hours beyond what we scoped. I can either fold them in and push delivery back a week, or keep the original date and bill the extra at my standard rate. Which works better for you?
Note what that does. It doesn’t accuse anyone. It attaches a number to a specific, recognisable event. And it hands the client two acceptable options rather than a bill they didn’t authorise.
Show a summary, never a raw log
When you do share the evidence, share the comparison, not the dump. Three or four lines — estimated versus actual, per phase, with the out-of-scope items named — is persuasive. A twelve-page export of every timer you started reads as defensiveness, and it invites the one conversation you really don’t want, which is a client auditing whether a particular Thursday afternoon was efficient enough.
The same restraint applies to the framing. The cause is “the scope grew”, not “you kept changing your mind”. You’re describing what happened to the project, not building a case against a person you’d like to work with again.
The bigger payoff is the next quote
Winning one change order is worth real money. Pricing the next ten projects properly is worth much more.
After a few completed jobs, your logs will show you the same overrun in the same place every time — the revision rounds, the content that arrives late, the phase you have systematically underestimated for years. That’s not a client problem, it’s an estimating problem, and it’s fixable: build the real number into the quote, cap the rounds explicitly, or price the volatile phase separately. Estimates built from your own history hold up in a way that estimates built from optimism never do. It’s the same evidence that tells you whether a retainer would suit the relationship better than another fixed fee.
Where Billable comes in
All of this depends on the tracking being effortless enough that you actually do it while the work is happening. Billable runs a live timer you can control from anywhere in the app, ties every entry to a project, and keeps each project’s tracked, unbilled and billed totals current — so the estimate-versus-actual picture is there when you need it, not something you reconstruct from memory.
When the scope conversation ends in a yes, the extra hours go straight onto an invoice, and quotes convert into invoices in a click so the paperwork matches what was agreed. Need to hand a client the underlying detail? Export the timesheet to CSV. It’s free for up to two active projects, then one flat $19 a month — every feature included either way. Start tracking what your projects really cost.