How to Negotiate a Software Agency Contract Mid-Project When the Scope Has Already Blown Out

By the time a founder realises the scope has blown out, they’re usually three months in, $40-80k deep, and staring at a project tracker that shows 60% of features “in progress” simultaneously. The agency isn’t lying, exactly. Nobody is. But the original brief was loose, the discovery was too short, and now you’re in that painful middle zone where stopping feels like waste and continuing feels like throwing good money after bad. This is one of the most common situations we see – and almost nobody talks about how to actually handle it.

The instinct is to either blow up the relationship or go quiet and hope things stabilise. Both are wrong. What you actually need is a structured mid-project reset – and knowing how to run one without torching the engagement or the product you’ve already half-built.

Why Scope Blows Out (And Why It’s Rarely One Person’s Fault)

Most scope blowouts follow the same pattern. The initial brief was written before anyone had done real technical design. Features that sounded like one thing turned out to be three things. The client added requests in Slack that never made it to the formal scope document. The agency said “we’ll figure that out when we get there” one too many times.

Sometimes there’s genuine negligence on the agency side – an inexperienced PM, poor change control, a team that under-quoted to win the work. Sometimes the founder changed their mind about what they were building mid-flight and didn’t think through the downstream cost. Usually it’s both, layered.

The reason this matters is that how you negotiate the reset depends entirely on where the fault sits. If the agency misquoted or built the wrong thing without flagging it, you have genuine leverage. If you added features and didn’t formalise the scope change, you’re in a more collaborative conversation. Walking in with the wrong posture will kill the negotiation before it starts.

Do This Before You Say Anything

Before you send a single email or jump on a call, you need to do your homework. This takes a day, maybe two, but it makes every subsequent conversation cleaner.

• Pull the original contract and scope document. What was explicitly included? What was marked as out-of-scope? What was ambiguous?
• List every change request since signing. Emails, Slack threads, meeting notes. Even informal ones. Especially informal ones.
• Map what’s been built versus what was scoped. Get a specific, feature-by-feature read. Not vibes – a list.
• Estimate what’s genuinely left. Ask your technical lead, or get a second opinion from an independent developer. This doesn’t need to be precise, but you need a ballpark in hours and dollars.
• Clarify your IP position right now. Does your contract give you ownership of code already delivered? Can you take a partial handover? This is your leverage if things go badly.

The mistake we see most often is founders going into a difficult commercial conversation without this prep. They rely on feelings and memory. The agency has the invoice history. You need comparable clarity.

The Conversation You Actually Need to Have

Set up a formal reset meeting – not a casual catch-up. Frame it as a “project health review” if you want to keep the temperature low. The goal of this conversation is to get to shared agreement on three things: where you are now, what it will realistically cost to finish, and whether the original commercial structure still makes sense.

Come in with your mapping exercise. Present it factually. “Here’s what was in scope. Here’s what we’ve built. Here’s what I can see is left. I want us to align on whether the original budget and timeline is still realistic, and if it’s not, I want to restructure before we go further.”

That framing is important. You’re not accusing. You’re not threatening. You’re treating this like two parties who both want a working product and need to agree on how to get there.

If the agency is professional, they’ll come to this conversation ready. A good agency knows exactly when a project is drifting – they’re often dreading this conversation as much as you are. Give them a structured way to be honest and most of them will take it.

What Renegotiation Actually Looks Like

There are three realistic outcomes from a mid-project reset, and you should decide before the meeting which one you’re optimising for.

Option one: Scope reduction. You cut the remaining feature list to what’s essential and finish on something close to the original budget. This is usually the right call. Most founders discover, when forced to, that 30-40% of their remaining scope was “nice to have” dressed up as “must have.” Ship the core. Add the rest in a second phase.

Option two: Budget increase with a locked scope. You agree on a revised budget to deliver a defined, frozen list. This only works if you formalise it. Get a new statement of work signed. No more Slack requests. No more informal additions. If the agency can’t commit to a fixed scope at this point, that tells you something important.

Option three: Managed exit. Sometimes the relationship is genuinely broken – trust is gone, the codebase is a mess, or the agency is in over their heads. In this case, you negotiate a clean handover: all code, all credentials, all documentation, in exchange for settling the outstanding invoice. A good contract makes this straightforward. A bad contract makes it a fight. Either way, get a lawyer involved if the numbers are significant – anything over $30-40k AUD in dispute is worth an hour of legal time.

Protecting Yourself for the Rest of the Build

Whatever you agree to in the reset, change how the remaining project is run. The structural issues that caused the blowout won’t fix themselves.

At minimum, you want:

1. A written change control process. Any request that affects scope, timeline, or budget gets documented and signed off before work starts. Non-negotiable.
2. Weekly delivery reporting against scope. Not status updates – actual feature completion against the agreed list, with any drift flagged immediately.
3. Staged payments tied to delivery milestones. If you’re paying monthly retainers regardless of delivery, you have no commercial lever. Tie at least some payments to defined outputs.
4. A defined IP transfer schedule. Don’t wait until the end. Get code deployed to your own infrastructure or repository incrementally. If the relationship breaks down again, you’re not starting from zero.

None of this is adversarial. A professional agency will welcome this structure – it protects them too. If they push back on basic change control or delivery reporting, pay close attention to that signal.

The Part Nobody Warns You About: Your Own Behaviour

Here’s the uncomfortable bit. A lot of the founders we work with who’ve come out of a failed agency engagement had a hand in the blowout that they don’t fully own. They added features because a competitor launched something shiny. They changed the core user flow after seeing a demo of a different product. They approved a direction in a meeting and then reversed it by email two weeks later.

That’s not a character flaw – it’s what happens when you’re building something new and learning as you go. But it has a commercial cost, and pretending it doesn’t will make every future agency conversation harder.

If you’re mid-blowout and you know you’ve contributed to it, say so in the reset meeting. Briefly, factually, without excessive self-flagellation. It changes the entire dynamic. Agencies are used to defensive founders. A founder who says “I know I’ve added scope without formalising it – let’s work out what that cost and how we move forward” will almost always get a better commercial outcome than one who comes in swinging.

The goal isn’t to win the negotiation. It’s to get a working product out the other side of a difficult situation. Keep that in front of you and most mid-project resets are survivable.

If you’re heading into a build and want the scope and commercial structure set up so this conversation never happens, talk to Amora about your build before you sign anything.