Why Project Forecasts Become Unreliable — and How to Fix Them
Project forecasts rarely become unreliable because someone cannot add up a spreadsheet.
They become unreliable because the forecast is disconnected from the things that actually drive delivery.
Resources move. Priorities change. Actual costs arrive late. Plans are updated without the forecast changing. Forecasts are changed without checking the delivery plan. Previous periods stay open too long. Assumptions remain hidden in emails and spreadsheets.
By the time the portfolio review takes place, the numbers may still look tidy — but they no longer represent what is really happening.
A good forecast is not simply a prediction. It is a controlled view of what the organisation currently expects to spend, when it expects to spend it, and why.
The first problem: forecasting without resource control
One of the most common causes of unreliable forecasts is allowing project costs to be forecast without first confirming whether the required resources are actually available.
A project manager may plan three people for the next four months and build a cost forecast around them. But if those people are already committed elsewhere, the forecast is based on capacity the organisation does not have.
The financial number may look reasonable, but the delivery assumption underneath it is weak.
This is why resource capacity and financial forecasting should not be treated as separate processes.
A stronger control chain is:
Capacity → Allocation → Project assignment → Forecast
If the organisation cannot confirm who or what is available to the project, it should be very cautious about committing the associated cost forecast.
The second problem: plans and forecasts drift apart
In many organisations, the project plan and the financial forecast live in different places.
The delivery plan may be in one system, resource information in another, and the forecast in Excel.
This creates a simple but serious problem.
The delivery date moves, but the forecast does not.
A resource assignment changes, but the monthly cost profile remains the same.
A piece of work is removed from scope, but its forecast remains buried in the numbers.
Over time, the plan describes one future while the forecast describes another.
The two should move together.
When the remaining work changes, the forecast should reflect the cost impact. When the forecast changes materially, there should be a clear delivery reason behind it.
The third problem: actuals are uploaded but not reconciled
Uploading actual costs is not the same as controlling them.
A monthly actual file can be loaded into a system and still leave major questions unanswered.
Why was a cost higher than forecast?
Was a cost expected this month but posted early?
Has an invoice slipped into next month?
Was time booked to the wrong project or cost category?
Is the remaining forecast still valid after the actual spend?
Without reconciliation, actuals simply become historical information.
The important control is what happens next.
A strong monthly process compares actual performance with the previous forecast, identifies the reason for differences and then corrects the remaining forecast.
The objective is not merely to explain the variance.
It is to improve the future number.
The fourth problem: teams spend too much time explaining the past
Monthly reporting often becomes an exercise in writing commentary about what already happened.
“Spend was £40,000 higher than planned because…”
That explanation may be necessary, but it is not enough.
The more important question is:
What does this mean for the rest of the project?
If a project has spent more than expected this month, does the total forecast increase?
Was the spend simply brought forward?
Has the project consumed contingency?
Will fewer resources now be required later?
Does the approved baseline still remain realistic?
Strong project control uses variance information to correct the future.
Most teams explain the variance.
Strong project control corrects the future.
The fifth problem: forecast ownership is unclear
A forecast can involve several people.
The project manager understands delivery.
The resource manager understands capacity.
Finance understands actual costs and accounting treatment.
The PMO understands portfolio governance.
The sponsor understands the business commitment.
Problems appear when everyone contributes to the number but nobody clearly owns the final commitment.
A reliable monthly process should make it clear:
who prepares the forecast;
who confirms resource assumptions;
who reconciles actuals;
who reviews material changes;
who approves the final number; and
when the period is considered closed.
The forecast should eventually become a committed management position, not an endlessly changing working file.
The sixth problem: prior periods remain editable
If historical periods can continue changing, trust becomes difficult.
A number reported last month may no longer match the number being shown today.
The team then spends time trying to work out which version was correct and why the history changed.
Once the monthly review is complete, prior periods should normally be locked.
Corrections should be controlled and traceable.
That creates a stable history and a clear audit trail.
It also makes month-on-month forecasting much more meaningful because people can see what changed, when it changed and who approved it.
What a reliable forecast process looks like
A dependable forecast is built through a repeated control cycle.
A practical monthly flow is:
Capacity → Allocation → Plan → Forecast → Actuals → Reconcile → Reforecast → Review → Approve → Close
Each stage strengthens the next.
Capacity confirms what is available.
Allocation confirms what has been committed.
The project plan defines the remaining work.
The forecast translates that work into expected cost.
Actuals provide evidence of what has happened.
Reconciliation explains differences.
Reforecasting corrects the remaining future.
Review challenges the assumptions.
Approval creates accountability.
Close protects the integrity of the period.
This is what turns forecasting from a spreadsheet exercise into a governance process.
The forecast should be explainable
A good forecast should survive a simple question:
Why is this number here?
For any material forecast value, the organisation should be able to trace it back to something tangible.
A resource.
A rate.
A quantity.
A planned activity.
A supplier commitment.
A delivery assumption.
An approved change.
If a forecast cannot be explained, challenged and traced, confidence in it will always be limited.
This is particularly important at portfolio level.
Senior leaders do not need every line of project detail, but they do need confidence that the aggregated number is built from controlled information underneath.
Baselines matter, but they should not become a prison
A baseline is essential because it provides the reference point against which performance is measured.
But projects change.
The answer is not to continuously overwrite the baseline until it matches reality.
That destroys the value of comparison.
Instead, organisations should preserve the original approved position, track changes against it and create a new approved baseline when governance determines that a genuine rebaseline is required.
That provides two things at the same time:
control of history and realism about the future.
Resource placeholders can make forecasts more honest
Sometimes the organisation knows work is required but does not yet know exactly who will perform it.
That does not mean the requirement should disappear from the forecast.
A placeholder resource can represent a required role, skill or resource type until a named person is assigned.
This is often more honest than pretending the capacity is already secured.
It also exposes future capability gaps early enough for management to act.
The important distinction is between:
we know the work requires this resource
and
we have actually secured this resource.
Good forecasting should make that difference visible.
Forecast confidence comes from discipline, not optimism
There is always uncertainty in project delivery.
No system can remove that.
But organisations can control how uncertainty is represented.
Risks should influence forecasts where there is a realistic expected financial effect.
Known commitments should be reflected.
Remaining work should be reassessed regularly.
Resource assumptions should be challenged.
Actual performance should continuously inform the future.
The aim is not to create a forecast that never changes.
In fact, a forecast that never changes may be a warning sign.
The goal is to create a forecast that changes for understandable reasons and through a controlled process.
A simple test for your own portfolio
Ask these questions at your next portfolio review:
Can every project explain how its remaining forecast was calculated?
Is the forecast supported by an up-to-date delivery plan?
Are the required resources actually available and allocated?
Have this month's actuals been reconciled against what was expected?
Have material variances changed the remaining forecast where necessary?
Can you see what changed since the previous approved forecast?
Is there a clear approval point for the current forecast?
Are prior reporting periods protected from uncontrolled changes?
If several answers are “no”, the problem may not be forecasting accuracy.
The problem may be the control environment around the forecast.
From forecast reporting to forecast governance
Project forecasting becomes more dependable when organisations stop treating it as a monthly number-entry exercise.
The forecast should sit inside the operating rhythm of the portfolio.
Resource decisions affect delivery.
Delivery affects cost.
Actual performance affects remaining expectations.
Governance determines what becomes committed.
When those elements are connected, the forecast becomes much more than a spreadsheet number.
It becomes a management tool.
And that is ultimately what senior decision-makers need: a number that can be understood, challenged and trusted.
Put stronger forecast controls into practice
ProjectFiles connects resource capacity, allocation, project planning, financial forecasting, actuals reconciliation, approvals and monthly close in one governed environment.
Clarity. Control. Confidence.
Numbers that hold up when challenged.
