01 · The questions
What should you ask a team that wants to take over your project?
Ask the nine questions below of every team and write down the answers, including ours.
| # | Question | A good answer sounds like |
|---|---|---|
| 1 | What do you do before you quote? | Read the code, or run a paid audit, then quote the next phase |
| 2 | What will the audit report contain? | Findings ranked by risk, what to fix first, and a fix or rebuild view with reasons |
| 3 | Will you say if a rebuild is the better answer? | Yes, with the evidence, even though it changes the scope |
| 4 | How do you stop things getting worse while you learn the code? | Freeze risky changes, add tests, then change |
| 5 | Who will work on it? | Named engineers, with location and time zone stated |
| 6 | Who owns the code, accounts and keys during and after? | You do, from day one; you hold the repository and cloud accounts |
| 7 | How is the work priced? | A fixed price for a first phase; later phases agreed after it |
| 8 | What happens at the end? | A handover you can run without them, with documents and a deploy anyone can repeat |
| 9 | Who runs it after launch? | A plan agreed before launch, or a clean handover to your team |
02 · Order of work
Why should a rescue start with an audit?
Because nobody can price a rescue honestly without reading the code.
- 01
A team that quotes a full fix from a demo and a conversation is guessing, and you carry the cost of the guess.
- 02
A code audit turns the unknown into a ranked list, and it is a cheap way to test a team before you commit to a larger phase. If you cannot get an audit first, ask for a first phase small enough to stop after. Ask also for the report in a form your next team can read, so the work is not trapped with the firm that wrote it.
03 · Warning signs
What are the red flags in a rescue proposal?
04 · Types of team
What kinds of team take over failing projects?
Each type suits a different job, and none is best for all.
| Type | Tends to suit | Watch for |
|---|---|---|
| The original vendor | Small fixes where the team still exists and knows the code | An incentive to defend earlier decisions |
| Freelancers and small studios | Small apps and tight budgets | One person's availability; handover depth |
| Specialist rescue and engineering firms | Audits and takeovers with senior review | Smaller capacity; fit with your stack |
| Large integrators | Very large programs with many workstreams | Cost and junior staffing |
| Your own engineers | Long-term ownership once the code is stable | Time to recruit and onboard while the problem keeps growing |
05 · Proposals
How do you compare rescue proposals fairly?
Give each team the same access and the same question, ideally a read-only copy of the repository and one page describing the problem, then compare scope lines, not totals.
- 01
Two proposals at the same price can cover different work, so check what each includes: tests, deploy pipeline, documents, handover.
- 02
Check what happens at each phase boundary and who carries the risk if a phase runs over. Ask whether a phase can be stopped without a penalty, and what you receive if you stop. For price context, see software rescue cost. Before you start, collect the items on the vendor handover checklist.
06 · Disclosure
How would Ace8 answer these questions?
Ace8 is the modernization practice of AceMQ, and we sell this service, so treat this as one set of answers among others.
- 01
We audit first. The rescue audit has a fixed price agreed after a scoping call, and later phases are agreed once we have read the code. We tell you when the finding is a rebuild. You own the repository and accounts throughout.
- 02
AceMQ engineers do the work, and AceMQ supports the result 24/7 after go-live if you want that; see supported after migration. Ace8 has no published rescue case studies of its own yet, so ask us for the audit sample rather than a client reference.
07 · Next step
What is the first step to take?
Send the repository and a short description of what went wrong.
- 01
We scope an audit on a call, with no obligation to continue. Start at software project rescue.