Software Project Stuck? How We Approach Project Rescue

You’ve been told “almost done” for three months, whatever ships keeps breaking, and you’ve stopped believing anything the team says. If that’s where your project is, this is how we approach it. The video above walks through it; this is the written version.

When It Becomes a Rescue

We call it a rescue when we step in to solve a serious problem, and there’s usually some urgency tied to it.

If your product works and makes money, and a new feature could make it more, that’s an opportunity and a standard development job. If you’ve spent the funding, you’re three months behind and things still don’t look good, that’s a problem, and it’s exactly where a rescue starts. It might be a missed deadline, a milestone promised to investors, or an app full of bugs that no one has fixed for a year.

What Can Still Be Saved

Not everything. If you’re three days away from the deadline on a two-month project where nothing has been done, it’s unlikely anyone can join and ship it on time.

The good news is that AI has made timelines much shorter, and there’s always something you can do to move things forward. Maybe that’s delivering enough to justify extending the deadline, or building only the part your investors or clients care about most.

Two Starting Points

You usually end up here in one of two ways: you have a team and you no longer trust them, or you have no team at all.

In our experience, having no team is the easier of the two. There’s nobody to convince and nobody’s feelings to manage, so you can focus on the problem in front of you instead of on someone in the middle who doesn’t want to be helped.

A Purely Technical Rescue

A client called us because their system was down. We had it back in under a day, spent the next two weeks making sure it wouldn’t go down again, and that kept the business running for a few more years, until they were ready to rebuild it properly.

The system was buggy, but the problem was purely technical. We became their dev team, everyone cooperated, and we all wanted the same thing. The full story is in our first Project Rescue case study.

Hundreds of Thousands Spent

At the other end, a client had spent hundreds of thousands of dollars on an app that kept breaking during events, and nobody could fix it for months.

The old CTO got into a fight with the owner. An agency came in and put a junior developer in charge, and after hundreds of hours of debugging, the app was still breaking. Then a new CTO said Python was the problem and the whole thing should be rewritten in another language, and the board actually considered it.

The bug was in the code, but by then there was distrust in the process, the investors were tired and the clients were unhappy. It was no longer a quick coding job. It was about rebuilding the company’s trust in its own technology.

It’s Rarely the Stack

From my experience, when someone opens with “it’s the framework” or “it’s the language”, there’s a high chance they don’t know what they’re talking about. The tech stack is rarely the cause; how you build on it usually is.

The Second Opinion

If you have a team, bringing in a second opinion can make them feel you don’t believe in them. After months of under-delivering, that distrust may well be justified, but you still worry it could make things even worse. That’s a real risk, especially if the team isn’t acting in good faith, so we plan for it like any other.

It’s Never Just the Code

Every problem shows up on several layers, and everyone involved sees a different one. The CEO sees the symptom: it’s not done yet. Tech blames sales for changing the scope all the time. Sales blames tech, because you can’t sell something that breaks in every demo.

So it could be a code problem, a people problem, or both, and looking only at the code is too shallow. Technical problems are the easiest to solve, and I wish they were the only thing founders had to worry about. Usually, though, it’s a mix of four things:

• Skill: the team doesn’t have enough of it to build what’s expected.
• Process: handoffs, documentation and quality assurance aren’t good enough.
• People: they don’t get along, so even skilled people don’t talk, and what gets built drifts away from what’s needed.
• Politics: conflicting interests and fights over budgets.

The First Week

We start with a free assessment, and it covers all four.

It begins with a call about what’s stuck, what it’s holding up, and who has the code. If the stack or the problem is outside what we know, we tell you right there. Then we get access to the code, the infrastructure, and any specs or documents you have, and we talk to the people involved: you, your team, sales, whoever else matters.

By the end of the week, you know how deep the problem goes, what the risks are, and your options, with what each one costs. You don’t have to hire us after that.

What Goes Live Next

Before you commit, we agree what goes live in under 60 days. As I said, there’s always something we can ship, and it’s usually the thing you’ve been stuck on. The first delivery typically lands within a month.

Everything else goes into a roadmap set to meet as many of your business goals as possible, and for every risk we found, it says how we’ll handle it.

Help or Take Over

Our goal is to rescue the project, and that looks different each time. Sometimes it means putting the plan together and supporting your team while they build. Other times, it means taking over and getting things built ourselves. We agree on the approach together in the first week, and the way we charge follows from it: a fixed price, a team, or a mix.

The plan also covers the way out: what we hand over, to whom, and when. At the end, you choose whether to keep working with the same people, or take the code, the accounts and the documents, and go.

To Sum Up

When a project is stuck, it’s rarely only the code. It’s usually skill, process, people and politics together. That’s why we look at all of them first, and only then agree what goes live.

If this sounds like where you are, tell us what’s stuck and book a 30-minute call. We’ll tell you the honest next step, even when it isn’t working with us.