Recovering a Troubled Project: A Rescue Playbook for Red-Status Projects
Almost every experienced project manager eventually inherits a project that is late, over budget, short of people or simply lost. The earlier lesson on why IT projects fail introduces recovery briefly. This lesson goes further and gives a practical rescue playbook: how to stabilise, diagnose, reset and rebuild confidence, and how to recognise when stopping is the right answer.
Is the Project Really in Trouble?
Before launching a rescue, check the facts. Typical signs of a troubled project include:
- Repeated missed milestones, with each new date as uncertain as the last.
- Budget forecasts that keep rising or reserves already consumed.
- Growing numbers of open defects and failed tests with no clear trend.
- Stakeholders who no longer trust the status reports.
- Unclear scope: people give different answers to “what are we delivering?”
- Key people leaving or burning out.
- A sponsor who is disengaged or avoids decisions.
Measure it using the indicators in Project Health Checks, KPIs and Dashboards. Distinguish a project that is slightly behind from one where the original plan is no longer credible. The first needs adjustment; the second needs a reset.
The Rescue Sequence
- Stop and stabilise. Pause non-essential work and new commitments for a short period, such as one to two weeks. Keep the critical team together and protect them from external pressure.
- Diagnose honestly. Run an independent health check: interview team members, sponsor and users, review scope, plan, risks, finances and quality evidence. Ask what is really going on, not what the status report says.
- Reconfirm the objective. Meet the sponsor and key stakeholders. Is the business case still valid? What outcome matters most and what is the real deadline?
- Decide the direction. Choose among continue with adjustments, re-plan and reset, reduce scope, change the approach or team, pause, or stop (see the next section).
- Re-baseline scope. Define a minimum viable scope that delivers the essential value, and move the rest to a later release.
- Re-estimate bottom-up with the people who will do the work, using actual performance data from the project so far (see Estimating IT Projects).
- Fix people, roles and governance. Clarify who decides what, fill critical gaps, remove blockers and shorten the decision loop.
- Establish a tight rhythm. Short planning horizons, weekly reviews, a visible task board and a daily check on blockers.
- Deliver an early win. A visible, working increment within weeks rebuilds credibility faster than any presentation.
- Communicate openly. Share the diagnosis, the new plan and the trade-offs. Trust returns when stakeholders see honest numbers.
Options at the Decision Point
| Option | When it fits | Cost or risk |
|---|---|---|
| Continue with adjustments | Plan is basically sound; problems are specific and fixable | May understate deeper issues |
| Reset and re-plan | Original estimates or scope were unrealistic | Requires new baselines and sponsor approval |
| Reduce scope (MVP) | Core value can be delivered sooner with fewer features | Some stakeholders will be disappointed |
| Change delivery approach | The method does not suit the work, for example a big-bang plan for uncertain requirements | Short-term disruption and learning curve |
| Replace or strengthen the team or vendor | Capability or capacity is the root cause | Handover delay and relationship cost |
| Pause | A dependency, funding or business priority blocks progress | Loss of momentum and knowledge |
| Stop | Business case no longer holds, or cost to complete exceeds value | Write-off and reputational discomfort |
Beware the Sunk Cost Trap
Money and time already spent are gone whatever you decide. The right question is: from today, does the cost to complete justify the expected benefits? If not, stopping is a good decision, and often the best outcome of a rescue review is a clear and early “stop”. Honest termination, with lessons captured (see Project Closure and Lessons Learned), is far better than a slow failure.
Sample 30-60-90 Day Recovery Plan
| Period | Focus | Outputs |
|---|---|---|
| Days 1 to 30 | Stabilise and diagnose | Health check report, sponsor decision on direction, reduced scope proposal, interim governance |
| Days 31 to 60 | Reset and rebuild | New baseline, re-estimated plan, clear roles, first working increment, refreshed risk register |
| Days 61 to 90 | Deliver and sustain | Regular delivery rhythm, improved quality metrics, restored status reporting, benefits tracking |
Protecting the Team
Troubled projects often run on overtime and blame. Recovery fails if the team collapses. Set realistic hours, acknowledge past strain, separate technical problems from personal criticism and celebrate early progress. Watch for key-person dependencies and spread knowledge. See Leadership, Conflict Resolution and Negotiation Skills for working with people under pressure.
Worked Example: Delayed CRM Implementation
A CRM implementation is 5 months late and 40 per cent over budget. The new project manager pauses new feature work for ten days. Interviews reveal that sales leaders and the vendor have different understandings of scope, requirements keep arriving through informal chats and testing has barely begun. She presents three options to the sponsor and recommends a reduced first release covering lead and opportunity management for two regions. The sponsor agrees. The team re-estimates, sets up weekly demos to sales users and fixes decision rights through a small product group. A first working release goes live six weeks later. The reporting module and the remaining regions are moved to a second release with its own business case.
Common Mistakes
- Adding more people late and expecting a quick fix.
- Keeping the same plan and simply compressing the schedule.
- Reporting a hopeful date without new evidence.
- Skipping the sponsor conversation about objectives and trade-offs.
- Failing to consider stopping the project.
- Confirm the project is truly in trouble using data, then stop and diagnose before acting.
- Reconfirm the objective with the sponsor and re-baseline a realistic scope and estimate.
- Deliver an early win and keep communication honest.
- Ignore sunk costs: continue only if the future value justifies the cost to complete.