EXAMTOPICS / TECHNOLOGY INSIGHTS

Why Projects Struggle: Four Causes and Practical Fixes

Project management becomes difficult when a plan looks precise but rests on assumptions that were not tested. A schedule cannot resolve conflicting objectives, create unavailable specialist capacity or make an external supplier deliver on time. Likewise, a dashboard can show activity while the real outcome remains unclear. Projects need decisions about people, resources, risk and value—not just a set of tasks.

Four recurring problems help explain why projects run into difficulty: unclear goals, unrealistic planning, weak communication and governance, and unmanaged changes or risk. These are diagnostic themes, not a universal ranking or proof that every troubled project has one root cause. Often several interact, and the appropriate response depends on the evidence available for the particular project.

Reason 1: The project starts without a shared definition of success

A project charter might call for “a better customer portal,” but stakeholders can understand the phrase differently. A technical team may prioritize speed, a sales team may expect new functionality and operations may want fewer support requests. These expectations are not necessarily incompatible, but failing to reconcile them leaves room for substantial work that nobody accepts.

The Association for Project Management’s conditions-for-success research identifies planning and review, clear goals, governance, capable teams and commitment as important factors associated with traditional project outcomes. It does not establish that following a single checklist guarantees success. The useful lesson is to clarify what value a project should create before deciding how to schedule it.

Write down the problem, intended users, scope, explicit exclusions, expected deliverables and acceptance evidence. For example, “provide authenticated access” is a technical output; “reduce avoidable support contacts” is a possible business benefit. The output can be checked immediately, while the benefit needs an agreed observation period and can be influenced by other factors.

Invite the people who carry out the affected business processes to review realistic examples. A sponsor who authorizes funding may not know every exception a support worker handles daily. Early feedback can reveal missing requirements before implementation, but it does not eliminate later uncertainty. Record unresolved assumptions and how they will be tested.

A warning sign is that teams disagree about what “done” means. If one group means software deployed and another requires users to complete a task safely, define the acceptance test and authorized approver. Without that agreement, completion percentages can create false confidence.

Reason 2: Scheduling ignores real capacity and dependencies

Many dates come from optimistic assumptions rather than work estimates informed by the people doing the tasks. A developer booked for five days may also have incident response duties. A data migration might depend on vendor access, regulatory review or an unchanged upstream schema. When these dependencies are not recorded, a polished Gantt chart conceals the uncertainty rather than managing it.

The APM’s project-controls guidance

Ask specialists to describe their capacity, technical assumptions, quality checks and plausible obstacles. Estimate work in appropriately sized pieces, then map prerequisites between teams. A task cannot begin simply because its calendar start date arrives if the required access or data is unavailable.

For a dependency outside the team’s authority, name an owner, a review date and an escalation or contingency. Avoid presenting an unverified supplier date as a guaranteed milestone. If a dependency slips, examine the actual effect on the critical work rather than assuming that every task in the schedule must move by the same number of days.

Distinguish a desired target from a current forecast. Stakeholders may need a target date for coordination, but the uncertainty and assumptions should remain visible. When reliable information changes, adjust the forecast and discuss any trade-off in scope, resources, quality or delivery timing.

Make planning a continuous review, not a one-time ceremony

A plan should be maintained as evidence develops. Review dependencies, capacity and material assumptions at intervals appropriate to the work. During early discovery, uncertainty may be substantial; after successful integration and testing, some estimates may become firmer. Precision should grow from observed progress rather than from pressure to provide an exact number.

Do not treat a change to an estimate as proof that the team has failed. A new requirement, unexpected defect or supplier change may legitimately affect the forecast. What matters is whether the project recognized the new information promptly, explained its consequences and obtained appropriate decisions.

Equally, avoid optimistic replanning that repeatedly pushes unfinished tasks into future weeks while reporting that the overall project remains healthy. If the same dependency blocks several review cycles, escalation or a different delivery approach may be needed. A useful schedule supports decisions, not merely attractive reporting.

Reason 3: Communication and governance are unclear

A technically capable team can remain blocked if nobody has authority to choose between competing demands. A security specialist may identify a serious access risk while a commercial sponsor wants to launch immediately. A project manager can coordinate the information, but may not have authority to waive security requirements or accept business risk independently.

The APM stakeholder-engagement principle

Create a decision map: who approves scope, budgets, quality exceptions and release readiness? Who provides technical advice? Who must be informed when an approved change affects them? Clarify these boundaries with the sponsor and relevant functional leaders. An escalation route is useful only if someone responds before the issue becomes irreversible.

Keep a short decision log for meaningful trade-offs. Record the issue, options, evidence, assumptions, decision maker and date. When an earlier decision is revisited, explain what new information justified the change. This reduces arguments based on conflicting memories and helps new staff understand the reasoning.

Status information should distinguish confirmed facts, forecasts and unresolved risks. A red indicator is useful only when accompanied by the impact and needed action, but a forced green status can be much more harmful. Reporting should enable better decisions, not reward the appearance of certainty.

Use collaboration to resolve competing objectives

Conflict may reveal different legitimate priorities rather than uncooperative personalities. A design team may value usability while operations needs lower support burden. Ask each group for the evidence behind its position, then make the actual constraints visible. A small prototype or test may resolve a disagreement more reliably than a long meeting.

Recognize that stakeholders have different information needs. A sponsor often needs decisions and business impact; developers need clear acceptance criteria and technical constraints; users need to know how a change affects their daily work. Sending identical detail to every audience can produce both information overload and unanswered questions.

For an example of practical management behaviors, a related discussion of communication, collaborative leadership and delivery control shows how those disciplines support one another. The purpose is not to increase the number of status messages but to keep responsibility and decisions understandable.

Reason 4: Changes and risk are managed after damage occurs

Requirements can change for legitimate reasons. A regulation may require an additional control, user research may show the initial design is ineffective, or a supplier may remove an interface. Problems arise when new work enters without evaluating its consequences or when emerging risks remain hidden until they become issues.

A proportionate change process records the proposal, objective, expected impact and authorized decision. Small wording fixes need less review than major architectural changes or commitments involving customer data. The Project Management Institute’s discussion of project failure factors

Distinguish a risk from an issue. A risk is an uncertain event that may affect objectives; an issue has occurred and needs action. Record the potential impact, owner, response and escalation trigger. Avoid invented numerical probabilities simply because a software tool asks for a percentage.

Quality is part of change control. A release can satisfy its planned date while exposing users to security or data-integrity defects. Define tests, acceptance rules and who can authorize residual risk. Closing tasks without completing required verification is not equivalent to delivering a usable result.

Where a change must be accepted, show the alternatives: adjust scope, allocate resources, change timing or decline the request. Document the choice and update the working plan or approved baseline as appropriate. Quietly adding scope while insisting that time and budget remain fixed creates a contradiction that later reporting cannot repair.

Diagnose the four problems through one scenario

Suppose a company is replacing an internal purchasing system. Departments disagree about approval steps, the external supplier has not delivered a required integration and a sponsor requests more reporting while keeping the original launch date. Developers continue to close tasks, but end-to-end purchase tests remain unreliable.

The first problem is the unclear definition of the acceptable purchasing journey. The second is the unverified supplier dependency and unrealistic capacity for integration testing. The third is a missing decision route for competing requirements. The fourth is an expanding scope without an assessed trade-off.

Address these jointly. Process owners can agree a minimum acceptable workflow, technical leads can identify the evidence needed for integration, and the sponsor can decide whether to defer optional reporting or revise the date. The team should then update the forecast using real test results rather than task counts alone.

Document decisions, assumptions and next review dates. If critical issues remain unresolved, report them clearly. Possible outcomes include a smaller authorized release, a later milestone or different resource allocation; no option can be promised without examining feasibility and risk.

Build early warnings and meaningful project reviews

At a reasonable interval, ask four questions. Do stakeholders still agree on the outcome and acceptance criteria? Does the schedule reflect realistic capacity and dependencies? Are material decisions reaching authorized people in time? Are changes, quality defects and risks being assessed transparently?

Warning signals can include repeated redefinition of “done,” several blocked dependencies, unanswered decisions, growing rework and test failures discovered too late to influence the plan. None is a universal prediction of failure, but each warrants investigation. A review should produce corrective actions with owners rather than another report that repeats the same unexplained status.

Good project management is adaptive without being arbitrary. Sometimes a change of scope or date is the responsible response to new information, provided the impact is understood and the decision is authorized. The four causes in this guide offer a way to diagnose uncertainty and strengthen decision-making, not a promise that any method can eliminate project risk or guarantee a successful delivery.