Why a Failing Software Project Needs a Reset

Knowing how to rescue a failing software project starts with understanding what has actually gone wrong. Perhaps the delivery date keeps moving, development costs are increasing, or your supplier insists the software is nearly finished but cannot demonstrate a working system. You may be questioning whether to invest more money, replace the development team or abandon the project altogether.

As a former CTO and Agile Coach, I would resist making that decision until the evidence is clear. A troubled project may still be worth completing, but recovery requires more than another deadline or a promise to work harder. You need a reliable picture of progress, a realistic plan and the confidence to decide whether continuing makes commercial sense.

Takeaways

  • Verify the current position. Working, tested software provides better evidence than optimistic progress reports.
  • Identify the causes before changing the team. Scope, governance, technical problems and supplier performance require different responses.
  • Reassess the business case. Further investment should be justified by the value still achievable.
  • Reset delivery with clear accountability. A recovery needs realistic priorities, forecasts, responsibilities and checkpoints.
  • Use independent oversight when appropriate. An objective assessment can help leadership make difficult decisions and regain control.

Table Of Content

Is Your Software Project Really Failing?

A software project is not necessarily failing because it misses a deadline or exceeds its original estimate.

Software development involves uncertainty. Requirements evolve, unexpected technical problems emerge and business priorities change.

What matters is whether the people responsible for delivery can explain the situation, assess the consequences and establish a credible way forward.

A project becomes seriously troubled when leadership can no longer rely on the information being provided or reasonably predict what additional investment will achieve.

Warning Signs Worth Investigating

Some warning signs are particularly concerning when they occur together:

  • Deadlines repeatedly move. Each new completion date replaces an earlier commitment without a convincing explanation.
  • Costs increase without visible results. Expenditure continues, but little additional usable functionality is demonstrated.
  • Progress is difficult to verify. Reports describe completed activities without showing what actually works.
  • Requirements keep expanding. New features are continually added without reconsidering timeframes, budgets or resources.
  • Quality problems are accumulating. Testing is delayed, defects keep returning or releases create new operational problems.
  • Nobody owns the overall outcome. The business, developers and suppliers have different understandings of what must be delivered.

One warning sign does not establish that a project is failing.

For example, a revised estimate may be entirely reasonable after an important requirement changes. The concern is when changes accumulate without proper assessment or accountability.

If your immediate problem is repeated schedule slippage, my guidance on a software project running late explains why a new deadline alone rarely addresses the underlying cause.

The Difference Between Delay and Failure

I would distinguish a delayed project from a failing project by asking three questions.

First, does the original business objective still justify the investment?

Second, is there a realistic technical and commercial route to achieving that objective?

Third, can the organisation demonstrate sufficient progress and maintain effective control over the remaining work?

A project can be late and still be commercially worthwhile.

Equally, a project can meet its revised schedule while delivering software that the business no longer needs.

Recovery should focus on business value, not simply finishing what was originally planned.

Regain Control Before Spending More Money

The first step in software project recovery is to create a controlled opportunity to assess the situation.

This does not necessarily mean stopping all development.

Some work may be essential to maintain existing systems, protect data or complete functionality needed for business continuity.

However, continuing to approve new features and expenditure without understanding the current position can make the situation harder to resolve.

Pause Unnecessary Scope Changes

I would temporarily suspend approval of discretionary new requirements until the review is complete.

That gives the project team a stable set of priorities and prevents additional commitments being introduced while existing problems remain unresolved.

Any exceptions should be agreed by the person accountable for the business outcome.

The aim is to stabilise decision-making, not to punish the developers or shut down useful work.

Identify the Decision-Maker

A recovery needs an accountable business sponsor.

That person must have sufficient authority to approve priorities, resolve competing requirements, make investment decisions and escalate material risks.

The project manager or CTO may coordinate recovery, but major commercial decisions must sit with the appropriate authorised executive.

If nobody can make a final decision about scope or funding, the recovery is likely to encounter the same obstacles as the original project.

Protect What You Already Own

Before considering a major supplier change, confirm the organisation’s rights and practical access to the work completed.

Depending on the project, this may include source code, repositories, cloud accounts, databases, documentation, designs, deployment processes and third-party licences.

Check who controls production access, where backups are stored and whether another competent team could reasonably maintain or continue the software.

Review intellectual property ownership and relevant contractual provisions before assuming that everything developed or hosted by a supplier can be transferred.

Do not revoke access or stop contractual payments impulsively. Changes may create security, legal and operational consequences.

This is about protecting business continuity while you establish your options.

Business executives reviewing delivery reports for a troubled software project
Assessing a Failing Software Project

Establish What Has Actually Been Delivered

You cannot create a reliable recovery plan without knowing the genuine state of the software.

This is often more difficult than it first appears.

Project reports may show substantial progress because development tasks have been completed. However, those tasks may not be integrated, adequately tested or usable by the people who need the system.

A feature described as developed is not necessarily ready for production.

Ask to See Working Software

I would ask the team to demonstrate the most important functionality using realistic business scenarios.

For example, if the project involves replacing an order management system, demonstrate how an order moves through the complete process.

Can an employee create the order, confirm inventory, issue an invoice, update the customer and produce the necessary operational information?

A demonstration of individual screens may look impressive while important integrations remain unfinished.

The distinction matters because the business is paying for functioning outcomes, not simply a collection of completed development tasks.

The US Government Accountability Office’s Agile Assessment Guide describes practices such as continuous testing, regular backlog refinement and acceptance criteria. Although written for government programmes, these concepts are useful when assessing whether software development progress is genuine.

Request Evidence, Not Percentages

A statement that the project is 90% complete tells you very little unless the remaining work has been assessed properly.

The final stages may involve complex integration, data migration, security testing and resolving defects. These can require substantial effort.

Instead, review evidence that can be verified.

QuestionUseful evidence
What functionality works?Demonstrations of completed business processes
What has been tested?Test results, known defects and acceptance records
What remains unfinished?Prioritised backlog and outstanding dependencies
What has been released?Release history and deployment records
What has been spent?Invoices, actual expenditure and commitments
What will completion require?Remaining work estimates, assumptions and risks
Who controls the software?Repositories, hosting arrangements and access records

You do not need to read source code to understand whether the project is producing useful outcomes.

However, technical expertise may be necessary to determine whether the underlying software can be maintained, extended and deployed safely.

Speak to the People Doing the Work

I would not rely entirely on reports prepared for senior management.

Speak with the developers, testers, project manager, business stakeholders and relevant suppliers.

Ask what they believe is preventing progress.

A CEO may see repeated missed deadlines. Developers may see requirements changing every week. The supplier may be waiting for decisions or information the business has not provided.

Each perspective matters.

Compare what people say with the project evidence rather than deciding in advance who is responsible.

Identify the Root Causes of Project Failure

Once the current position is clear, investigate why the project has reached that point.

The most visible problem is not always the underlying cause.

A late project might appear to have insufficient developers when the real issue is that requirements are continually changing or important decisions remain unresolved.

Likewise, weak project reporting may conceal deeper technical problems that are making delivery increasingly difficult.

The Australian Government’s Major Digital Projects Report 2026 identifies recurring challenges involving project management, governance, scope, budgets, resources, schedules and technical issues.

These findings relate to government projects, but the underlying issues are useful areas for private businesses to examine.

Scope and Requirements

The project may have begun with an achievable set of requirements.

Over time, additional features, integrations, reporting functions and business rules were introduced without revising the delivery plan.

Every request may have seemed reasonable in isolation.

Together, they can change the size and complexity of the project.

Ask which requirements were part of the original agreement, which were added later and who approved the consequences.

Unrealistic Estimates

The original schedule may have been created before the team understood important technical constraints.

Perhaps integrations were poorly documented, data quality was worse than expected or the application required more extensive testing.

Some uncertainty is normal.

Repeatedly issuing new estimates without identifying what was learned is a different problem.

A credible recovery needs estimates based on the work now understood, including the remaining uncertainty.

Technical Constraints

Software architecture, accumulated technical debt and weak development practices can affect delivery.

Examples include unreliable automated tests, difficult deployments, tightly coupled systems, poor documentation and recurring defects.

These problems can make each new feature more expensive to implement.

However, technical debt should not become a general explanation for every delay.

Ask which technical issues are materially affecting delivery, what evidence demonstrates their impact and what improvement would be gained by addressing them.

People and Supplier Problems

The project may lack critical skills or sufficient development capacity.

An external supplier may be underperforming, or responsibilities between the supplier and internal team may be unclear.

Sometimes the team is capable but has insufficient authority to challenge unrealistic commitments.

I believe people before technology is especially important during recovery.

Blaming individuals before understanding their working conditions can damage trust and cause capable people to leave just when their knowledge is most valuable.

Weak Governance

A project may have developers, a project manager and a regular reporting process without having effective leadership.

Important decisions might take weeks. Business priorities may conflict. Nobody may have clear authority to approve scope changes.

The recovery plan must address these decision-making problems rather than simply introduce more frequent status meetings.

Decide Whether the Project Is Worth Saving

Not every failing software project should be rescued.

One of the most important responsibilities of a business sponsor is deciding whether the remaining investment can still produce sufficient value.

The money already spent should inform your understanding of what exists, but it should not be the reason to continue.

The relevant question is what the business can reasonably achieve from this point forward.

Compare the Available Choices

OptionWhen it may be appropriateMain consideration
Continue with correctionsThe objective is valid and delivery problems are manageableCorrect specific weaknesses without unnecessary disruption
Reset the projectThe objective remains worthwhile but scope, cost or schedule is unrealisticAgree on a new, achievable delivery baseline
Change the supplierThe existing supplier cannot credibly address material problemsConsider transition costs, contractual rights and knowledge transfer
Pause the projectEssential information, funding or business decisions are unavailablePreserve useful assets and minimise continuing exposure
Stop the projectThe business case no longer justifies further investmentProtect the organisation and recover whatever value remains

These are not always mutually exclusive.

A business may reduce scope, change governance and negotiate revised supplier commitments as part of one recovery.

Reconsider the Business Case

Imagine a hypothetical wholesaler that has invested in a custom order management platform.

The project is significantly delayed, and the remaining development appears more expensive than originally anticipated.

Before approving further funding, the CEO should compare the expected value of completing the system against the remaining cost and risk.

Could an existing commercial platform meet the essential requirements?

Could part of the custom software be retained?

Would a smaller first release deliver enough value to justify continuing?

Might stopping now preserve more capital for other priorities?

There may be no perfect answer.

The purpose of a recovery assessment is to give leadership a defensible basis for choosing.

Understand the Cost of Changing Suppliers

Replacing developers may seem attractive when trust has deteriorated.

However, a new supplier needs time to understand the system, assess code quality, resolve access issues and learn the business requirements.

The organisation may also need to negotiate contract termination, intellectual property rights, documentation and ongoing support.

Changing suppliers can be the correct decision, but it should follow an assessment of what can realistically be transferred and what work may need to be repeated.

Independent technical or legal advice may be needed for material contractual concerns.

Build a Realistic Software Recovery Plan

If the decision is to recover the project, the next step is to establish a new plan based on evidence rather than the original expectations.

A recovery plan is not simply the old schedule with different dates.

It needs to address the factors that caused the project to lose control.

The Project Management Institute’s guidance on recovering troubled projects describes structured assessment, recovery decisions and corrective actions. That supports treating recovery as a managed intervention rather than an informal request for the team to work harder.

Reconfirm the Business Outcome

Start by agreeing why the project still matters.

What business process must improve? What customer problem must be solved? What operational risk must be reduced?

Write down the outcomes in language the business sponsor and delivery team both understand.

The recovery should not attempt to deliver everything originally imagined if only part of it remains essential.

Reduce Scope Where Practical

Identify the smallest useful release that can deliver meaningful business value.

Separate essential functionality from improvements that can wait.

For example, a system replacement may need accurate transactions, core integrations, appropriate access controls and reliable reporting before employees can use it safely.

Additional dashboards, convenience features or optional automation may be suitable for later releases.

Scope reduction must not remove necessary security, quality, regulatory or operational requirements.

It should reduce unnecessary complexity rather than defer problems that will prevent the system from operating.

Rebuild the Delivery Forecast

Ask the team to estimate the remaining work using the requirements and technical knowledge now available.

The forecast should include development, testing, integration, data migration, deployment, training and any transition activities that are relevant.

Identify assumptions, dependencies and areas of uncertainty.

Where uncertainty is substantial, provide a range or conditional forecast rather than a precise date that cannot reasonably be supported.

The objective is to create a forecast leadership can use for decisions.

Assign Clear Accountability

Every important workstream should have an accountable owner.

The business should know who approves priorities, who manages the delivery plan, who resolves technical decisions and who accepts completed functionality.

Where suppliers are involved, their responsibilities should be agreed explicitly.

Avoid creating a recovery structure that requires every minor decision to reach the CEO.

The executive sponsor should focus on business outcomes and material decisions while allowing capable team members to manage the work within agreed authority.

Establish Delivery Checkpoints

Break the recovery into manageable stages with clear acceptance criteria.

At each checkpoint, ask what has been completed, what remains, what risks have changed and whether the business case still supports continuing.

A checkpoint should demonstrate useful progress, not simply confirm that another reporting period has passed.

If the first stage cannot produce the agreed evidence, investigate before approving substantial additional commitments.

Business and technology leaders agreeing on a staged software project recovery plan
Planning Software Project Recovery

A Practical 30-Day Project Recovery Plan

A troubled project may require months to complete, but leadership should not need months to establish its current position and agree on a recovery approach.

I would consider a focused 30-day process for an established business with a substantial project already in progress.

This is an illustrative framework, not a promise that every assessment or recovery can be completed in 30 days.

Week 1: Stabilise the Project

Agree on the immediate decisions required.

Pause discretionary scope changes, establish an accountable sponsor and identify commitments that cannot safely be interrupted.

Collect the current plan, expenditure, requirements, technical information, supplier agreements and delivery reports.

Confirm access to essential software assets and operational systems.

The outcome should be a clear mandate for the assessment and an understanding of immediate risks.

Week 2: Verify the Delivery Position

Review working software and the most important business processes.

Compare completed functionality against the requirements and acceptance criteria.

Examine outstanding development, testing, defects, dependencies and remaining commitments.

Speak with the people delivering and using the software.

Identify the main causes of delay and where the existing information cannot yet be relied upon.

The outcome should be an evidence-based assessment rather than another optimistic project status report.

Week 3: Evaluate Recovery Options

Reconfirm the business objective and identify the functionality that remains essential.

Review the realistic options for continuing, reducing scope, changing suppliers, pausing or stopping.

Assess expected costs, delivery timeframes, technical risks, business consequences and supplier obligations.

Prepare an initial recommendation with the assumptions and uncertainties made explicit.

The outcome should be enough information for the authorised decision-maker to choose a direction.

Week 4: Approve the New Direction

Agree on the recovery scope, priorities, budget boundaries, responsibilities and reporting arrangements.

Establish staged delivery checkpoints, quality expectations and escalation rules.

Where required, renegotiate supplier commitments or arrange transition planning.

Communicate the decision to the delivery team and relevant business stakeholders.

The outcome should be a credible plan and a clear understanding of what the project is now expected to deliver.

Recovering an Outsourced Development Project

Outsourced software development creates an additional challenge.

The business may depend on the same supplier for delivery, technical information and explanations of project performance.

This does not mean the supplier is unreliable. It does mean that the business needs enough independent visibility to make informed decisions.

Review Commercial Commitments

Examine the relevant agreement, statement of work and approved variations.

Clarify which deliverables have been accepted, which remain outstanding and how changes to scope or milestones are handled.

Consider payment triggers, warranties, support arrangements, intellectual property, termination provisions and the supplier’s obligations to provide documentation or assist with transition.

For material disputes, obtain appropriate legal advice. A technical recovery assessment should not substitute for contractual interpretation.

Evaluate the Supplier’s Response

A supplier willing to acknowledge problems, demonstrate completed work and present a credible corrective plan may be worth retaining.

The response should include realistic commitments, identified dependencies and a practical explanation of how the causes of failure will be addressed.

Be cautious about accepting another deadline supported only by assurances that the team will work harder.

Likewise, a supplier’s reluctance to disclose important technical or delivery information deserves investigation.

Protect the Working Relationship

The business and supplier may have different interpretations of the original agreement.

Try to separate those commercial disagreements from the immediate work of protecting systems, data and project continuity.

Avoid public blame, threatening technical staff or abruptly changing access arrangements without understanding the consequences.

A recovery may still require these people to work together.

Preserving a professional relationship can make it easier to correct the project or manage an orderly transition.

How to Tell Whether Recovery Is Working

Once the revised plan is underway, leadership needs to know whether the changes are producing results.

More detailed reporting does not necessarily mean better control.

I would focus on a small set of measures that help determine whether the project is becoming more predictable and whether useful software is being delivered.

Measure Usable Progress

Track functionality that has been completed, tested and accepted against agreed requirements.

The exact measures will depend on the delivery approach.

For Agile development, completed increments of usable software provide valuable evidence. The Scrum Guide describes the importance of transparency, inspection and a shared Definition of Done.

However, simply increasing the number of completed tasks or story points does not prove that the business is receiving more value.

Ask whether the delivered functionality supports the objectives agreed in the recovery plan.

Watch Costs and Quality

Compare actual expenditure with the revised forecast.

Review important defects, unresolved technical risks and whether testing is progressing alongside development.

An apparent increase in delivery speed is not necessarily positive if quality is deteriorating or unresolved defects are accumulating.

Watch for problems that could make the eventual release unsafe, unstable or difficult to support.

Report Decisions, Not Just Activities

Executive reporting should answer five straightforward questions:

  1. What useful functionality has been completed?
  2. What is the current forecast for the next agreed checkpoint?
  3. Which risks or dependencies threaten delivery?
  4. How is spending tracking against the revised plan?
  5. Which decisions need executive attention?

The report should highlight meaningful changes rather than repeat the same information every week.

It should also make clear when an issue is unresolved and who is responsible for addressing it.

Common Software Project Rescue Mistakes

Even after recognising that a project is in trouble, organisations can make decisions that increase the cost of recovery.

Replacing Developers Immediately

Changing the development team before understanding the causes of failure may introduce new delays, additional cost and loss of technical knowledge.

First establish whether the problems involve capability, capacity, leadership, requirements, architecture or supplier performance.

Replace people or suppliers when the evidence supports that decision, not simply because confidence is low.

Adding More Developers Without a Plan

Additional capacity may help when work can be divided effectively and the existing team has the leadership and processes needed to support more people.

It is less useful when development is blocked by unclear requirements, unresolved decisions or complex technical dependencies.

More developers will not automatically make a poorly defined project easier to deliver.

Demanding an Earlier Deadline

Pressure can encourage the team to focus, but it cannot eliminate genuine technical complexity.

An unrealistic deadline may encourage optimistic reporting, reduced testing or unfinished work being presented as complete.

The better approach is to agree on a realistic scope and forecast, then hold people accountable for communicating changes promptly.

Approving a Complete Rebuild Too Quickly

A struggling system may have serious architectural problems.

That does not automatically mean rebuilding it is the most economical option.

A rewrite introduces its own delivery risks and may require the organisation to recreate functionality that already works.

Compare repair, targeted replacement, staged modernisation and a complete rebuild against the business objectives and remaining investment.

Ignoring Organisational Causes

A project can have capable developers and sound technology while still failing because the business cannot make decisions or agree on priorities.

If those conditions remain unchanged, another supplier or project manager may encounter the same difficulties.

Successful recovery requires leadership to address its own contribution to the problem.

When Independent Project Delivery Helps

Independent involvement is most useful when the business cannot confidently determine the project’s true position or needs experienced leadership to implement the recovery.

You may have received conflicting explanations from internal teams and suppliers.

You may be considering additional funding but lack confidence in the revised estimates.

Or your organisation may not have someone with sufficient software delivery experience to challenge assumptions, resolve priorities and oversee the recovery.

Start With the Right Assessment

If the immediate concern is software quality, development practices, architecture or whether the existing development team can reasonably deliver, an independent Software Development Review may be a suitable first step.

It can help clarify the issues before the business makes major decisions about developers, funding or rebuilding the software.

A review is not always necessary.

If the main problems are already understood and the team has the capability and authority to correct them, internal action may be sufficient.

When Recovery Needs Ongoing Oversight

Some projects need more than a one-off assessment.

The business may need support establishing a credible recovery plan, managing supplier commitments, coordinating internal stakeholders or maintaining executive visibility during delivery.

White Internet Consulting provides independent Project Delivery consulting for businesses facing these challenges.

The work can involve understanding the current delivery position, clarifying priorities, assessing risks, strengthening supplier accountability and supporting the practical steps needed to regain control.

My approach is to work with the existing people and suppliers wherever that makes sense.

I do not begin with the assumption that the project needs a replacement development team, a particular software platform or an entirely new methodology.

The appropriate involvement depends on what is wrong and what the business needs to achieve.

Supporting the Team Through Recovery

Recovery can be difficult for developers, project managers and business stakeholders.

People may be frustrated, concerned about their jobs or reluctant to report problems after months of missed commitments.

Introducing a more demanding reporting process will not necessarily improve that situation.

I would begin by creating clear expectations, listening to the people involved and making responsibilities manageable.

People need to know what matters most, what decisions they can make and how to raise concerns without being blamed for issues outside their control.

That is the practical meaning of people before technology.

When leadership gets those conditions right, the team has a better opportunity to focus on delivering useful outcomes.

Independent technology consultant discussing software project recovery options with business leaders
Independent Software Project Recovery Advice

Frequently Asked Questions

Can a Failing Software Project Be Saved?

Yes, some failing software projects can be recovered through clearer priorities, realistic scope, improved governance and changes to delivery practices. Recovery depends on whether the remaining work is technically achievable and commercially worthwhile.

How Do You Rescue a Failing Software Project?

Start by confirming the business objective and verifying what software has actually been completed. Then identify the root causes, compare recovery options and establish a realistic plan with clear responsibilities, costs and delivery checkpoints.

Should We Replace Our Software Developers?

Not automatically. Missed deadlines may result from changing requirements, weak governance, unrealistic estimates or technical constraints rather than poor developer capability. Review the evidence before deciding whether changing developers or suppliers is justified.

How Long Does Software Project Recovery Take?

The assessment and initial reset may take days or weeks, depending on complexity and access to information. Completing the recovered project can take considerably longer, so distinguish the time needed to create a recovery plan from the time needed to deliver the remaining software.

When Should We Stop a Software Project?

Consider stopping when the original business objective is no longer worthwhile, the remaining costs and risks outweigh the expected benefits, or no credible delivery approach is available. The decision should also consider contractual commitments, operational dependencies and the value of work already completed.

Conclusion: Regain Control Before You Commit

A troubled software project is not necessarily beyond recovery, but continuing without understanding the problems can expose your business to further cost and risk. Establish what has actually been delivered, decide whether the remaining investment makes sense and put a credible plan in place before making new commitments. If you need an experienced independent perspective, explore Project Delivery support to discuss how to rescue a failing software project.

Share This Post

Need help getting a technology project delivered?

Technology projects can lose momentum quickly. Unclear scope, changing priorities, supplier problems and missed deadlines can all put delivery at risk.

If a project is struggling, running late or simply needs stronger direction, take a look at my Project Delivery service or Contact Us to start the conversation.

Iain White Project Delivery Consultant

Technology projects rarely go exactly to plan. What matters is how quickly problems are recognised and brought under control.

Iain White helps businesses bring clarity, structure and momentum to complex technology projects.

He focuses on clear scope, realistic priorities, delivery risks and honest communication, helping business leaders understand what is happening, what is getting in the way and what needs to happen next.

Iain has worked across projects ranging from software and cloud initiatives to major technology transformations. In one case, he helped recover a project that was already six months late by re-scoping the work, resetting expectations and establishing a more realistic path to delivery.

His experience across technology strategy, software delivery, governance, risk and leadership helps him identify problems early and work with teams, suppliers and stakeholders to get delivery moving again.

Through White Internet Consulting, Iain helps businesses regain control of technology projects and turn plans into practical results.