Project Cost Management: Budget, Baseline and Contingency

A restaurant owner once told me his kitchen renovation was "on budget until the last week." I asked what happened in the last week. "The bills arrived," he said.

That is the oldest joke in project management, and it is not a joke. Money moves in four stages: you plan it, you commit it (a signed quote or an order), you receive the work and you pay the invoice. Most people only watch the last stage. By the time it shows up on the bank statement, the decision that caused it was made weeks ago.

This guide explains how to manage project cost like a project manager: an estimate that is honest, a budget that is approved, a baseline you can measure against and a rhythm of control. It is written for owners, managers and project leads who do not have a finance team behind them. The numbers in the examples are illustrative, and nothing here is accounting, tax or financial advice. Ask a qualified accountant about your specific situation.

The three jobs

The PMBOK Guide groups cost work into three activities, and the names are worth remembering because they point at three different questions.

1. Estimate costs. What do we think each piece of work will cost?
2. Determine the budget. Having added the pieces and the cushions, what is the total we approve, and when will the money be needed?
3. Control costs. As the project runs, how does the actual compare with the plan, and what do we do about it?

Many small projects do only the first, on the back of an envelope, and then act surprised in the third. The jobs depend on each other: a control system needs a baseline, and a baseline needs an estimate that was built from the work.

Estimating: a range that tightens

A common mistake is to treat an early number as a promise. Imagine you are asked in the first week, "How much will the renovation cost?" The honest answer has a range, for example "between $90,000 and $140,000 until we have drawings." That is not weakness. It is information.

As the project becomes clearer, the range narrows:

• Rough order of magnitude. Used at the idea stage, with little detail. The range is wide. Its job is to say whether the project is in the right neighborhood.
• Budget estimate. Used when the scope is defined and quotes or unit rates exist. The range is narrower. Its job is to approve a budget.
• Definitive estimate. Used when drawings are complete and the major packages are quoted. The range is tight. Its job is to commit the money.

Industry practice assigns typical accuracy ranges to these stages, and they vary by industry, so I will not quote a table. The principle is what matters: say what stage the number is at and how wide the range is. The business case guide shows how to test a case against a cost that comes in higher than hoped.

How to build an estimate

The most reliable method is bottom-up, using the work breakdown structure from the scope post: estimate each work package, then add them. A package estimate usually has four parts:

• Labor: hours times rate, for your people and for contractors.
• Materials and equipment: quantities times prices, with a note of when the price was quoted and for how long it holds.
• Fees and permits: the costs that are fixed by someone else.
• Indirect costs: the share of management, insurance, site costs or overhead that the project causes. They are easy to forget and they are real.

Two supporting methods help you check yourself. Analogous estimating looks at a similar past project and adjusts it. Parametric estimating uses a rate, such as a cost per square foot or per unit. Neither replaces a bottom-up number, but both tell you if it is in the right range. If your detailed estimate says $60 per square foot and your last three projects cost $95, there is a conversation to have.

Always write the assumptions and exclusions next to the number. "Excludes furniture. Assumes standard soil conditions. Prices valid until December 15." Those lines are where the arguments later come from, and they are where the answers are too.

Contingency and reserve: two different cushions

Every estimate is uncertain, so every budget needs a cushion. There are two kinds, and mixing them up causes trouble.

• Contingency is money set aside for risks you can identify: the price of steel might rise, the inspection might fail, the supplier might be late. It is tied to the risk register, it is part of the cost baseline and the project manager decides how to use it.
• Management reserve is money held for risks you cannot identify. It sits above the baseline and needs the sponsor's approval to spend.

A practical way to set contingency is to look at the risks and give each a rough cost and probability. If you do not have the time, a percentage of the estimate is a common starting point, and it should be larger for earlier, rougher estimates than for later ones. The project risk management guide explains how to score risks in plain English, and the contingency is simply the cost side of that exercise.

One rule keeps contingency honest: when it is spent, say on what. "Used $4,000 of contingency for the delayed delivery of the glass" is a statement. "Contingency is down" is a mystery.

The cost baseline: the budget over time

Once the estimate and the cushions are agreed, you spread them over the weeks or months in which the money will be spent. The result is a curve that rises slowly, faster in the middle and slowly again at the end, often called an S-curve. The approved curve is the cost baseline. It answers two questions: how much in total, and how much by when.

In the PMBOK series the invented clinic spreads its $250,000 base cost over nine months, with about $45,000 spent by month three. That is a baseline in miniature: a total, a timing and a line you can compare with reality.

Illustrative example: a storefront renovation

Here is a made-up project to follow through. A small shop owner plans a three-month renovation of her storefront.

• Estimate by package (illustrative): demolition $6,000, framing and drywall $14,000, electrical $9,000, flooring $11,000, fixtures $13,000, signage and permits $5,000. Total $58,000.
• Contingency: the owner lists three risks: hidden wiring problems behind the walls, a delayed fixture order and a failed inspection. She sets aside $6,000, about 10 percent, and writes down which risks it covers.
• Management reserve: her lender asks her to keep $3,000 of her own cash aside for anything else. That is outside the baseline.
• Cost baseline: $64,000 spread over twelve weeks: about $12,000 by week three, $38,000 by week eight and the rest at the end.

At week five she looks at the sheet. She has paid $21,000, signed commitments (orders and signed quotes) of $41,000 in total and has seen $22,000 worth of work. The baseline says she should have done about $26,000 worth. So she is behind on progress, and her committed cost is already higher than the work done. Two numbers, one signal: she should find out why before the next payment, not after.

Controlling costs: three habits

1. Track commitments, not only invoices

An invoice is the last stage of spending. A commitment, such as a signed quote or a purchase order, is the moment the money is spent in every way that matters. Keep a simple list: item, supplier, committed amount, invoiced so far and what remains. The sum of commitments tells you where you are headed. This is the same reasoning as in change orders in construction: the cost is decided at the change, not at the bill.

2. Forecast the final cost

"We have spent $40,000" tells you where you are. "We expect to finish at $68,000" tells you where you are going. A simple forecast is: cost so far plus the estimate for the work that remains, updated with what you have learned. The tool that makes this systematic is earned value, which compares the value of the work done with the money spent and the money planned. The earned value guide explains it with a worked example. Two numbers worth knowing: the cost performance index (value of work done divided by actual cost) and the schedule performance index (value of work done divided by planned value). An index below 1 means you are getting less than you paid for, or less than you planned.

3. Approve changes before you spend

Most overruns are not one big mistake. They are fifty small "just do it" decisions. A light change process helps: a one-line request that says what changes, what it costs, what it does to the date and who approved it. The scope creep guide has a template. Without it, your baseline is a decoration.

Reporting cost without drowning in it

A weekly one-page status report is enough for most small projects. For cost, put four lines on it:

1. Budget (baseline): the approved total and the planned spend to date.
2. Actual: invoices paid and commitments signed.
3. Forecast at completion: your best estimate of the final cost.
4. Contingency left: the balance, and what it still protects.

If the forecast is higher than the baseline, say so, say why and say what you will do: reduce scope, find savings or ask for more money. A report that says "on budget" every week until it says "over by 30 percent" is a report nobody trusts.

Cost and benefit: do not forget the why

Cost management is about spending well, and spending well is judged against the benefit. A project that costs 10 percent more than planned but delivers twice the benefit is not a failure. A project that comes in on budget and delivers nothing is. The benefits realization guide explains how to keep the benefit in view while you manage the cost, and the business case is where the two first meet.

A word on small projects

Everything above scales down. For a project of a few thousand dollars, the estimate can be a short list, the contingency a single line and the baseline a two-column table with a date and an amount. What cannot be skipped is the habit: write the estimate with its assumptions, name what the cushion is for, keep a running list of commitments and look at the forecast every week. Ten minutes of that habit protects more money than an hour of arguing about an invoice after the fact, and it gives you the numbers you need the next time someone asks, "How much will it cost?"

Mistakes to avoid

• A single-number estimate. Always give a range at the early stages and name the stage.
• No assumptions written down. The argument you avoid today comes back in month three.
• Contingency without a reason. If nobody can name the risk, it becomes a slush fund.
• Watching only invoices. By the time the invoice arrives, the decision is old.
• Ignoring indirect costs. Your own time, insurance and overhead are real.
• Quiet changes. A change without an approval is a leak.
• Forgetting cash flow. A project can be on budget and still run out of cash in month four if the payments arrive before the income. Plan when the money moves, not only how much.

Check where you stand

The tool below helps you judge whether your budget is a baseline or a hope. Tick what is already true.

Interactive tool

Is your budget a baseline or a hope?

Loading the tool…

What to do this week

1. Take a project you are running and write its cost estimate by work package, with assumptions and exclusions.
2. List the three biggest risks to the cost and put a rough amount next to each. That is your contingency.
3. Decide who can approve a change and above what amount.
4. Start a commitments list and update it every week.
5. Write the forecast at completion, even if it is a guess, and put it on your weekly page.
6. If you need to test a decision, use the cost lines in a project management plan, where budget and schedule sit together.

If you would like help putting a simple cost control system in place, see my services, take the diagnostic or browse the free templates.

Frequently asked questions

What is the difference between a budget and a cost baseline?

A budget is the approved total amount for the project. A cost baseline is that budget, minus management reserve, spread over time, so you know how much should be spent by each date. You compare actual spending and work done with the baseline to see whether you are on track.

How much contingency should a project have?

There is no single right percentage. It depends on how uncertain the estimate is and on the risks you have identified. Early, rough estimates need more, and detailed estimates with firm quotes need less. A good practice is to tie contingency to named risks in a register and to record what it is spent on.

What is the difference between contingency and management reserve?

Contingency covers identified risks and is part of the cost baseline, usually controlled by the project manager. Management reserve covers unidentified risks, sits above the baseline and needs approval from the sponsor or management to spend.

How do I track cost on a small project without special software?

A spreadsheet is enough: one line per work package with the budget, commitments, invoices paid and the forecast to complete, reviewed weekly. Add a list of approved changes. The important part is the weekly habit and the forecast, not the tool.

Join the conversation

Did this help? A quick reaction tells me what to write next.

Useful Made me smile Go deeper on this

Be the first to react.

Comments

Loading comments…

Leave a comment

Name (optional) Which bill surprised you most on your last project, and when was the decision behind it made? Post comment

No email and no sign-up. Every comment is reviewed before it appears. Please keep it respectful and leave links out.