The Project Financial Management Process: A PMO-Finance Workflow from Budget to Forecast

The project financial management process is the governed cycle in which the PMO and Finance plan, approve, record, reconcile, forecast and act on the financial position of projects. It connects delivery assumptions and a feasible schedule with budgets, actual costs, contract rules and the accounting system, so the people with decision rights can act on a variance while it is still a forecast. It runs in eight steps: define the case, estimate, approve baselines, configure contract rules, reconcile actuals, compare plan with forecast, test interventions, then decide, close and learn.

A typical scenario for a monthly review: the ledger has closed, Finance brings actual costs, the PMO brings status. Project A has $300,000 of budget left, so it looks safe. Two months later it misses its contract date, because the verification team it needed was booked on another project all along. The penalty was never in the cost report.

The cost report was correct. It answered a different question: what was spent, not whether the remaining work can finish on time with the people available.

Most processes are built around the budget, and the budget is only one of three decisions. The other two, which work is still worth the money and which work starts next, are left to the calendar.

A budget decides how much may be spent. It does not decide which work is still worth the money, or what starts next.

What is project financial management?

Project financial management, or financial management in project management, is the forward-looking control of a project's money: what it may spend, what it has spent, what it will earn, and how changes in scope, dates and people move those numbers. Its output is a decision, not a report: continue, change, park or stop.

Four terms are often used as synonyms. They are not:

Term	Question it answers	Main owner	Main system	Time horizon
Project financial management	Will this project deliver its financial case, and what should we do now?	PMO with Finance	PPM, fed by ERP	Forward
Project accounting	What was spent, committed, invoiced and recognized?	Finance, project controller	ERP, accounting	Backward, by period
Project finance	How is a large capital asset funded?	Treasury, lenders	Financial models	Life of the asset
Portfolio financial control	Which projects, in what order, earn most from the money and people available?	Portfolio board, PMO	PPM	Rolling

The process in this article sits across the first and last rows. It uses the ledger without replacing it.

Inside it, three decisions run at three different rhythms:

1. How much may be spent. The spend envelope: an annual or funding-cycle approval by the executive team and Finance. It should stay stable.
2. Which work is worth doing, and in what order. The investment ranking. It changes when value, cost or dates change, and the forum that sets it must be able to remove work, not only add it.
3. What starts now. Work release: ready work goes to a team when that team has room. This happens continuously, week by week.

Collapse the three into one annual event and the approved list becomes a commitment nobody revisits. A project funded in January is still running in September after half its assumptions have changed, because the only decision anyone took was the budget. Flow Economics calls this pattern the frozen portfolio.

Project management financials, the individual fields such as Approved Budget, Remaining Budget and predicted profit, are defined in our guide to project financials. This article is about the process that uses them: who decides what, when, and with which numbers.

The systems and roles behind the process

Managing project finances means working across several systems. Each owns one kind of number, and the process works when everyone knows which system is authoritative for which figure.

System	Owns	Does not own
ERP and accounting	Ledger postings, period close, revenue recognition, invoicing, collections	Feasible delivery dates
Time tracking	Hours booked per person and task	Rates, budgets
Procurement	Purchase orders and commitments	Labor cost, schedule
CRM and contracts	Contract value, terms, approved change orders	Delivery forecast
PPM decision layer	Resource-constrained forecast, predicted dates, commercial exposure from those dates, ranking, scenarios	Statutory records, cash received
BI	Consolidated reporting across systems	Source data

Figure 2. Systems of record keep the numbers; the decision layer turns them into dates and options.

In an enterprise project portfolio management software setup, the PPM layer reads cost and contract inputs from the systems of record and adds what none of them has: the dates the portfolio can reach with the people it has.

Roles follow the same split. Finance should not be accountable for operational data it cannot see, and the PMO should not interpret contracts. A model to adapt, with one accountable owner per row:

Activity	Portfolio board	PMO	Finance	Resource owner	Project manager	System owner
Approve the spend envelope	A	C	R	C	I	I
Set Revenue and the contract rule	I	C	A/R	I	C	I
Maintain baseline, dates and remaining effort	I	A	I	C	R	I
Record costs and maintain rates	I	I	A	I	I	R
Run the forecast and publish exposure	I	A/R	C	C	I	I
Rank, park or stop investments	A	R	C	C	I	I
Release ready work to a team	I	A	I	R	C	I
Approve interventions above threshold	A	R	C	C	I	I
Execute and communicate the decision	I	C	I	C	A/R	I
Configure fields, permissions, integrations	I	C	C	I	I	A/R

R: responsible, A: accountable, C: consulted, I: informed.

The eight-step project financial management process

The eight steps form a loop, not a sequence you run once. They group into six management moments: set up, validate, monitor, predict, decide and learn. The first three steps are project financial planning; the rest is control.

Step	Moment	Accountable	Decision it serves	Rhythm
1. Define the business case and assumptions	Set up	Sponsor	Is it worth funding?	At intake
2. Estimate effort, cost and revenue	Set up	PMO	What will it need, and from whom?	At intake, at stage gates
3. Approve work-hour and monetary baselines	Set up	Portfolio board	How much may be spent?	Annual or funding cycle
4. Configure contract and change rules	Validate	Finance	What is each date worth?	At contract, at change
5. Collect and reconcile actuals	Monitor	Finance	What has been spent?	After each period close
6. Compare plan, actual and forecast	Predict	PMO	Where is exposure still preventable?	Monthly and after major changes
7. Investigate and test interventions	Decide	PMO	Which option costs the portfolio least?	On exception
8. Approve, communicate, close and learn	Decide, learn	Portfolio board	Continue, change, park or stop?	Periodic ranking, at closeout

Figure 1. The eight steps as a loop: the budget is approved once a year, the ranking is refreshed every quarter, and work is released every week.

1. Define the business case and assumptions

Start with what the project is supposed to return. For contract work, that is revenue and the contract terms. For internal work, it is a benefit with a named owner: lower rework, a new capability, a compliance date met. Internal projects never send an invoice, but they still have a price of time. Every week they slip, the benefit arrives a week later.

Write the assumptions next to the numbers: scope, rates, the start window and the teams the work depends on. Without them, nobody can later tell whether the estimate was wrong or the world changed.

2. Estimate effort, cost and revenue

Estimate effort per team before you estimate money. The figure that matters most is the hours each project needs from the bottleneck: a resource constraint of people and skills, not a step in a process. In most engineering portfolios it is one or two specialist teams that every project needs at the same time.

Cost follows from hours, rates and additional costs such as materials. Finance owns the rates, resource owners the effort estimates, the contract owner the revenue. Keeping the three sources separate makes the plan auditable.

3. Approve the work-hour and monetary baselines

Approve two baselines, because they answer two questions. The work-hour budget says how much effort the project may use. The approved monetary budget says how much money it may consume. A project can stay inside its money while it overruns its hours on a cheaper team, and the reverse.

Approval sets the spend envelope, not start dates. A project can be worth funding without starting in January; released when its team has room, it is likely to finish sooner. Treat the approved list as an envelope, not an irrevocable twelve-month schedule. See project budgeting and cost control for how the two baselines sit next to the delivery plan.

4. Configure contract and change rules

Turn the contract into rules the forecast can price. Enter Revenue, then select the bonus and penalty rule that applies: grace period, rate per day or week late, cap. Finance or the contract owner interprets the contract. The PMO maintains the dates the rule is measured against.

In Epicflow, schemes can be kept per client or contract type, with one active per project, and a simulator tests a scheme against a project's revenue and an end-date variance before it is assigned.

A change order updates Revenue or the budget through a recorded decision, never a silent edit. Watch two exceptions here: a contract project with no Revenue shows no commercial exposure, and a project with no applicable scheme shows its dates but not what they are worth.

5. Collect and reconcile actuals

Hours come from time tracking, cost postings from the ledger, commitments from procurement. Reconcile them after each period close, so everyone compares the same snapshot.

Expect differences, and explain them. Labor cost in a PPM tool is hours multiplied by configured rates: a management figure, not a ledger entry. Accruals and close timing will make it differ from the ledger. Print the authoritative system next to each figure on the project financial report, and the argument about whose number is right becomes a reconciliation line.

Remaining Budget is Approved Budget minus Actual Costs. It shows unconsumed funds, not the cost to complete. A negative Remaining Budget is an exception: it goes to change control, with the sponsor deciding on a re-approval. It is never absorbed by a quiet rebaseline.

6. Compare plan, actual and forecast

This is the step most processes skip, and the one that turns a cost review into project management financial analysis. Rerun the resource-constrained forecast after each period close and after any major change: a new project, a lost specialist, a moved milestone. It predicts when each project will finish, given the work remaining and the people available across the whole portfolio.

Three dates now sit side by side: baseline end, current due date and predicted end. The contract rule prices the due date as the actual bonus or penalty and the predicted end as the predicted one. The gap is exposure that is still preventable. With project management forecasting software, you can act on a predicted penalty before it becomes an actual one.

Two cautions. Moving the due date changes the actual adjustment, not the feasibility of the predicted date. And a predicted profit that subtracts only recorded costs is not an estimate at completion; add the estimate to complete. Then ask the forward question: what does the remaining spend still buy?

7. Investigate and test interventions

Trace each variance to its cause first: overspend, delay or scope. Delay is the one cost reports miss, and in a shared-resource environment it usually comes from a queue in front of one team. Resource capacity planning shows which team that is, and which projects are waiting for it.

Then build options: keep the plan, add capacity at the bottleneck, change the sequence, reshape scope, park a project or stop one. Test each against the whole portfolio. On a single project, a week gained looks created. Across projects that share people, it is usually borrowed from another one. What-if analysis shows what each option does to every project's dates and money before the live plan changes. The PMO and resource owners build the options; Finance judges whether the amounts are material.

8. Approve, communicate, close and learn

The authorized forum takes the decision with the options and their prices on one page. It can remove work as well as add it. The project manager executes and tells the teams and sponsors affected, including the sponsor whose project now moves later.

Record the decision, its assumptions and its expected effect, then check the effect at the next review. Refresh the ranked list. Adjust schemes and escalation thresholds that produced noise or missed something.

At closeout, reconcile the final figures with the ledger and compare estimated with actual hours per team. That difference is how the next project financial plan gets better than this one.

Project financial review cadence and controls

A single monthly meeting cannot carry all three decisions. Give each its own rhythm:

Rhythm	Decision	Inputs	Forum
Weekly	Release ready work; watch delivery signals	Predicted dates, load on the bottleneck team	PMO and resource owners
Monthly, after period close	Reconcile actuals; rerun the forecast; review exposure	Ledger actuals, hours, predicted adjustments	PMO and Finance
Quarterly or at a stage gate	Rank, park or stop investments	Remaining value against remaining investment	Portfolio board
Annual or funding cycle	Approve the spend envelope	Strategy, capacity, the ranked list	Executive team, Finance
On exception	Act on a trigger below	The exception and its options	Per threshold

Exceptions should have a defined path, not a debate about who owns them:

Exception	What it means	Who acts	Default response
No Revenue on a contract project	Commercial exposure cannot be calculated	Finance	Enter Revenue before the next review
No applicable contract rule	Dates are visible, their price is not	Finance, contract owner	Select or create a rule; test it first
Negative Remaining Budget	Spend exceeds approval	Controller, sponsor	Change control; no silent rebaseline
Predicted penalty beyond grace	Preventable exposure	PMO, resource owner	Model options; Finance judges materiality
Due date and predicted finish disagree	The commitment is not feasible	PMO	Fix the cause, not the date
Remaining case has deteriorated	Finishing returns less than before	Portfolio board	Continue, reshape, park or stop
Higher-return work arrives mid-year	The ranking is out of date	Portfolio board	Test in a scenario; re-rank; envelope may stay

Two controls keep the numbers honest. Segregation of duties: whoever records costs does not approve budget changes, and whoever maintains dates does not set Revenue. An audit trail: every change to a baseline, date, Revenue or rule is logged with its author and reason. Calculated fields stay read-only, so nobody overwrites a result instead of its input. Exceptions with a contract or safety impact also belong in project risk management.

Worked example: forecast delay changes the financial decision

Illustrative example. All figures are teaching assumptions, not a customer result or software output.

An engineering company runs two projects through one verification team. The monthly review falls in week 26.

	Project A (customer contract)	Project B (internal platform)
Approved Budget	$1,200,000	$600,000
Actual Costs	$900,000	$350,000
Remaining Budget	$300,000	$250,000
Estimate to complete	$350,000	$400,000
Forecast total cost	$1,250,000 ($50,000 over)	$750,000 ($150,000 over)
Value	$2,000,000 revenue	$700,000 benefit, cut from $1,500,000 after a product line was descoped
Baseline end and due date	Week 30	Week 40
Predicted end	Week 36	Week 40

Contract rule for A: two weeks' grace, then 1% of revenue per full week, capped at 8%. B's benefit is spread over three years from completion, so a week of delay costs about $4,500 ($700,000 / 156 weeks).

Why is A predicted for week 36? The verification team is booked on B's test phase in weeks 29 to 32. A's verification waits behind it, runs weeks 33 to 36, and finishes six weeks late: four weeks beyond grace, so a predicted penalty of 4%, or $80,000.

The actual fields show nothing wrong: A's due date is still week 30, so the actual penalty is zero. The predicted state shows Predicted Revenue of $1,920,000. Including the estimate to complete, the margin is $670,000, an ROI of 53.6% on total investment. On what is left to spend, A returns 5.5x ($1,920,000 / $350,000).

B now loses money on total investment: $700,000 of benefit against $750,000 of cost, an ROI of -6.7%. On what is left to spend, it still returns 1.75x ($700,000 / $400,000). Stopping B would save $400,000 and give up $700,000. The $350,000 already spent does not count.

Three options go to the review:

Option	Effect on A	Effect on B	Cost to the portfolio
1. Keep the sequence	Week 36, penalty $80,000	On plan	$80,000
2. Hire a contractor for A's verification ($70,000, available from week 31)	Week 33, penalty $20,000	On plan	$90,000
3. Move B's test phase behind A's verificationLowest cost	Week 32, inside grace, no penalty	Four weeks later, about $18,000 of delayed benefit	$18,000

Option 3 costs the portfolio $62,000 less than option 1 and needs no new money. Option 2, the instinctive answer, costs more than doing nothing. The spend envelope does not change in option 3. The ranking does: B moves behind A for four weeks.

The decision follows the RACI. The PMO and the verification team's resource owner built the options. Finance confirmed the contract reading and that $80,000 is material. The portfolio board approved the new order, because it changes the ranking, and B's sponsor heard the price: about $18,000. A's project manager executes. Canceling B was rejected, because finishing it still returns more than it costs. In a larger portfolio, the scenario test in step 7 also checks that the move creates no new queue for a third project.

Common process failures

Failure	What it causes	Fix
One number for the budget	Hour overruns go unseen	Separate work-hour and monetary baselines
Late actuals	Decisions use last quarter's position	Reconcile after each period close
Silent rebaselining	Variance vanishes without a decision	Rebaseline only through change control
No accountable owner	Exceptions wait for a meeting	One A per RACI row
False precision	Debates over cents	Round; show assumptions and data date
Mismatched calendars	Ledger month and delivery week disagree	Agree one cut-off per review
Stale rates	Labor cost drifts	Update rates on a fixed cycle
Missing assumptions	Nobody knows why estimates moved	Store assumptions with estimates
PPM and ERP confusion	Two versions of revenue	Name the authoritative system per figure

Behind most of them sits one failure: the annual budget is treated as the whole process. The spend envelope is controlled; the ranking and the release happen by default.

How Epicflow fits the PMO-Finance workflow

Epicflow is the decision layer in this process, not the ledger. It builds a resource-constrained prediction of the whole portfolio, a plan built around the people you actually have, and puts the money on top of it. When the plan moves, the money moves with it.

• Set up and validate (steps 1 to 4). Approved Budget, Revenue and the bonus and penalty scheme are editable in the Pipeline. One scheme is active per project, and a simulator tests it before assignment. Business Value, Custom Fields and Attributes carry the case and its context.
• Monitor (step 5). Actual Costs and Remaining Budget are calculated, read-only fields next to the delivery dates.
• Predict (step 6). A Prediction run produces predicted end dates and the predicted bonus or penalty, revenue and profit for every project. Epicflow also shows whether each project is still worth finishing: what it returns per dollar still to spend, as a multiple, next to ROI. The technical name is RRI, return on remaining investment, built on Stephen Devaux's work (Devaux, 2014). Invoice Forecast and Cash Flow show when milestones become invoiceable and when payment is expected, derived from the schedule, not from a bank balance.
• Decide and learn (steps 7 and 8). What-If Analysis tests each option before the live plan changes. For larger portfolios, the Epicflow Portfolio Optimizer searches for the sequence that earns most; your people decide. Each role configures its own Pipeline view on the same numbers.

Figure 3. Budget, recorded costs and predicted financial exposure in Epicflow Pipeline. Enlarged details from the supplied product demo.

What stays in accounting and ERP: ledger postings, period close, invoicing, collections, revenue recognition, procurement, payroll and statutory reporting. Epicflow forecasts; the ERP records.

What changes when the review prices time

A PMO that reviews project financials asks: are we spending what we approved? A Value Management Office asks: is the work we are doing now still the best use of the money and people we have? A good project financial management process makes the second question routine: the budget stays approved, the ranking stays alive, and work starts when its team has room. Flow Economics sets out what a value management office does in more depth.

Bring your project list and one real decision, such as which project gets your busiest team next month. We run it on your own data and show how the dates and the money change. Schedule a Live Demo

Further reading: Flow Economics on how to know when a project is no longer worth finishing.

Frequently Asked Questions

References

Devaux, S. A. (2014). Managing projects as investments: Earned value to business value. CRC Press.