Most projects that lose money looked fine, until they didn’t. Those are the ones that are the most frustrating. It’s one thing to have a snake-bit project where you deal with a series of unfortunate events that are beyond your control. It is quite another thing to book profit that suddenly erodes in the last 10% of the project.
Those projects have a familiar pattern: The project manager is confident, and the forecast holds at the margin the estimate promised for the majority of the project. Then, somewhere past 80% complete, the numbers start to move. A $12M job that carried a $600K fee forecast in month eight closes with a $400K write-down, and leadership spends closeout asking how nobody saw it coming.
The good news is, good projects don’t have to end like this.
In my experience, the answer is rarely one bad decision. It is a series of small ones that never got surfaced:
- No baseline. The estimate was loaded loosely or overwritten, so there is no baseline to measure against.
- Paperwork buyout. Buyout was treated as paperwork, and scope gaps were bought into subcontracts nobody read closely.
- Hidden variances. Variances piled up in odd amounts across dozens of codes, and contingency and fee quietly absorbed costs that should have been priced as change orders.
- One bucket. Buyout savings and contingency sat in the same bucket, so no one could say how much cushion was there to start with.
- Bad news held back. And in many cases, project managers learned that bringing bad news to leadership earned them a hard conversation instead of help, so they held it until they couldn’t.
Margin fade does not have to be a surprise. Profit loss can be managed or even eliminated when the team applies a handful of basic principles with discipline. This article covers a few of those best practices that can mean the difference between making the bid-day fee and writing the project down.
Key Takeaways
- Margin fade starts early. Most write-downs begin at buyout and grow in variances that nobody has to explain, then show up past 80% complete.
- Lock the estimate as the original budget. Construction budget management starts with a locked baseline and formal revisions that keep an audit trail from estimate to closeout.
- Give buyout to the people who will build the job. Project managers and superintendents have the strongest incentive to close scope gaps before contracts are written.
- Pick a variance method and document every change. Letting variances ride or leveling the budget both work as long as any change to the budget or cost is explained and “paid for” when it appears.
- Leadership sets the tone. Monthly reviews catch problems early only when project managers trust that raising one starts a conversation about solutions.
Load the estimate as the original budget and lock it
Once the contract is signed, the estimate should be loaded exactly as built into your ERP and/or project management system, designated as the original budget, and locked. This simple action preserves an audit trail from the estimate through every revision, change order, and forecast. If costs start to outpace budget, you now at least can identify which codes were underestimated, poorly bought, or eroded by scope gaps.
A common complaint among project managers is that the estimate is built differently than how they buy a project. Provided every budget change is accomplished through a formal budget revision process to create an audit trail, the team can still recast the budget when needed. An example of this is when five self-performed concrete lines collapse into one lump-sum subcontract.
Additionally, keeping non-revenue revisions separate from revenue-generating change orders will provide a coherent story of where gaps occurred or business decisions were made.
Put buyout in the hands of the person who will build the job
Many firms have a centralized purchasing process where the estimator buys out the job. This is a workable model, although as a former project manager I prefer that the operations and field team own buyout. My thinking here is the PM and superintendent are the people who will live with each scope. They have the strongest incentive to scrutinize and own it.
All too often a young PM will be tempted to make the mistake of pencil-whipping the buyout. They will write scopes and contract values straight from subcontractor quotes. This is a sure way to find scope gaps later in the project that will cost money. Another common mistake is to use the stated budget line in the estimate without considering the potential of scope hold included.
A good buyout requires careful review of inclusions and exclusions against the documents, finding scope gaps between trades and closing them through the subcontract scopes before contracts are written.
None of the above should be done in a vacuum. A formal handoff from estimating to operations can surface potential scope gaps or unique estimate structures. Senior leaders should be providing appropriate oversight throughout the buyout process to ensure the project is on track financially as well as if the pace of the buyout is supporting the project schedule. After all, the buyout plan and execution is budget management by another name.
Is your budget telling you the truth about your jobs?
Margin loss on a job usually builds slowly between the estimate and closeout, through small decisions that never come up in a cost review. Ascent reviews how your budgets are set, tracked, and reviewed each month, and shows you where profit is slipping while there is still time to recover it.
Two Legitimate Approaches to Managing Variances
As the project progresses and cost begins accumulating against the budget codes, variances will begin to show up. Those variances could be positive or negative and occur for any number of reasons: Positive buy, negative buy, or cost exceeding the commitment or budget due to a change to name three.
There are two defensible ways to handle a budget and the variances:
“Letting it ride” leaves the budget untouched and allows pluses and minuses to accumulate in the variance column per cost code to keep the original budget intact regardless of any estimate errors or scope gaps found. The benefits of this method are to maintain a history to highlight estimating errors. One would manage the project profit using the net of all the cost codes and adjusting fee (or contingency if you’re lucky enough to have one) to absorb net losses or gains.
“Budget leveling,” as I preferred, involves leveling the budget to the commitment so the variance column reads zero and contingency is balanced monthly like a checkbook, for two reasons: any nonzero variance then becomes an immediate flag for a miscoded invoice, an errant timesheet, or labor running long, and the project manager must decide how each problem gets paid for when it arises, which keeps jobs that looked healthy all year from closing at breakeven.
I prefer the budget leveling method because it keeps the cost report clean and variances show up visually as red flags to track down. I found that letting the cost ‘be what is going to be’ made it visually difficult to see changes month to month in the variance.
But more importantly, I believe that ‘letting it ride’ can have an unintended negative psychological impact on a PM that allows them to take less ownership of managing the cost.
Either way, the best practice is to document each change to the budget or any cost greater than planned to understand why there was a variance and how to avoid mistakes in the future. I felt that as long as I had the discipline to document any changes, I would rather handle that documentation as I went. I also felt it forced me into a place mentally where I had to know how I was going to pay for problems. I effectively had to balance the checkbook every month. My method forced me to face project challenges in lieu of sweeping them under the rug.
Reconcile in stages and keep buyout savings apart from contingency
A middle ground to the two above budget management methods would be to reconcile the budget right after buyout to reconcile the budgets and the commitments (subcontracts and POs) written.
As a practical matter I suggest reconciling in two or maybe three passes through buyout. The first after early release packages such as structure, MEP, steel, and elevators. These are your larger dollar trades and usually where you can find the estimating errors. The second pass after the finishes trades, and occasionally a third for late scopes like landscaping and hardscape.
Move the savings from each pass into a dedicated unallocated or buyout savings code, and keep it separate from contingency, which on GMP and cost-plus work is often disclosed, contractually governed, and subject to shared savings.
When an unrecoverable cost appears, spend buyout savings before contingency, much as a developer spends equity before drawing on the loan. Reconcile allowances and owner change orders on their own tracks, since neither should touch these funds.
After buyout, review monthly and adjust rarely
Every project’s budget (and cost) should be reviewed monthly by project and company leadership. A well-bought project should rarely need monthly revisions outside of codes that have labor cost them. General Conditions or General Requirements are obvious examples for a general contractor.
If you find a scope gap or make a business decision to eat cost on an owner driven scope change, process any non-revenue change orders and the corresponding budget revision together so neither is forgotten.
Leadership should treat widespread monthly revisions as a signal that scopes were not bought thoroughly, and subcontractors have found the gaps.
Reporting: What should leadership see each month?
Leaders sometimes worry that a column of zeros hides what they want to discuss. That is a fair and reasonable concern. We are trained through experience to start to look for what is not there on projects in lieu of what is obviously there.
Snapshots taken before and after each monthly reconciliation address that directly, as do the audit trail of approved budget changes and the draw history on contingency and buyout savings.
Company-level reports in some platforms calculate earned revenue differently from project budgets, so a portfolio view built in an analytics layer such as Power BI is often a good tool to use.
Leadership visibility without discouraging candor
None of this works unless project managers trust that surfacing a problem early starts a conversation about solutions, and leadership establishes that trust through the tone of every cost review. Problems surfaced early allow for time to solve them.
Leadership should focus on creating processes to surface problems as early as possible and then creating an environment of solutions in lieu of casting blame. Review of the buyout progress and monthly review of the PCO log, budget revisions, and variance log provide plenty of conversation to understand that the project manager has a handle on things.
Also, project managers need to understand that questions do not always mean distrust. As project managers advance in their career there will be new exotic contract structures to administer, owner challenges, and scope of work never seen before. There is always room for learning.
Bringing the discipline together
Remember that $12M job that carried a $600K fee forecast in month eight and closed with a $400K write-down. That loss did not happen in the last 10% of the project. That’s just where it showed up. The loss happened at buyout, when a scope gap went into a subcontract nobody read closely, and it grew every month a variance sat in a column nobody had to explain. By the time the math finally shook out, there was no time left to recover it.
Every practice in this article is aimed at moving that moment of discovery earlier. A locked original budget gives you a baseline to measure against. A buyout owned by the people who will build the job closes gaps before they turn into cost. A consistent method for managing variances, whether you let it ride or level the budget, forces each problem to be explained and paid for when it appears. Staged reconciliation and a separate buyout savings code tell you exactly how much cushion you have left, and monthly reviews backed by good reporting give leadership visibility without turning the cost review into an interrogation.
None of this requires new software or a bigger staff. It requires an agreed method, systems configured to support it, and the same accountability on every project, along with a culture where the project manager who raises a problem early is recognized for doing the job well. Profit isn’t an accident. It is the result of a system applied consistently from estimate to closeout.
If you would like a second set of eyes on where your own process may be leaking margin, we’re here to help.
Frequently asked questions
1. What causes profit fade on construction projects?
Profit fade rarely comes from one decision. It builds from small ones that never get surfaced: an estimate loaded loosely or overwritten, buyout treated as paperwork, scope gaps bought into subcontracts, and changes absorbed by fee instead of priced as change orders. The loss usually starts early and only shows up late in the project when there is no time left to recover.
2. Should the estimate be loaded as the original budget?
Yes. Once the contract is signed, load the estimate exactly as built into your ERP or project management system, mark it as the original budget, and lock it. That keeps an audit trail from the estimate through every revision, change order, and forecast. If the team needs to recast the budget to match how the job is bought, do it through a formal budget revision so the history stays intact.
3. Who should own buyout on a construction project?
The project manager and superintendent who will build the job have the strongest incentive to scrutinize each scope. If you have a central purchasing or Estimator buyout model, whoever owns it, a formal handoff from estimating should surface scope gaps and unusual estimate structures, and senior leaders should review buyout progress for both cost and schedule. Buyout is budget management by another name.
4. What is the difference between letting variances ride and budget leveling?
‘Letting it ride’ leaves the original budget untouched, so pluses and minuses build in the variance column. ‘Budget leveling’ resets the budget to the commitment so variances read zero, and any new variance becomes an immediate flag. Either method works when every change is documented with why it happened and how it will be paid for. However, letting the budget remain untouched can hide problems as variances change over time and memories fade.
5. Should buyout savings be kept separate from contingency?
Yes. Move the savings from each buyout pass into a dedicated unallocated or buyout savings code and keep it apart from contingency, which on GMP and cost-plus work is often disclosed, contractually governed, and subject to shared savings. Separate codes show exactly how much cushion is left. When an unrecoverable cost appears, spend buyout savings before contingency.
6. How often should construction project budgets be reviewed?
Review every project’s budget and cost monthly with project and company leadership. A well-bought job should rarely need monthly revisions outside labor-driven codes such as self perform scopes of work or general conditions. When revisions show up across many codes, treat it as a sign that scopes were not bought thoroughly. Process any non-revenue change orders together with their budget revisions so neither is forgotten.
Services mentioned in this article
- Fractional Executive Advisors — part-time executive leadership across finance, operations, and growth.
- Fractional CFO — senior finance leadership for construction companies that need CFO-level oversight without a full-time hire.
- Project Management — training, tools, and accountability for project managers who run the job like a business.
Related reading
- What Are Cost Codes and Cost Types in Construction Accounting
- How WIP Statements Improve Project Financial Reporting
- 4 Things the CFO Thinks the PM Understands
- Change Orders
About the author
Jeff Robertson is a Senior Consultant and Principal at Ascent Consulting. Jeff has more than 30 years in commercial construction, from field laborer to project manager on projects from $3M to $100M to operations manager of a business unit with more than $100M in annual revenue, plus time on the owner side as a development manager. He now helps contractors build the processes that protect margin.
About Ascent Consulting
Ascent Consulting is a construction-business consulting firm that is software-agnostic and operations-first. We work inside general contractors, specialty trade contractors, and private equity-backed construction companies, and our consultants have run the accounting, operations, and project teams we now advise. From budget discipline and project management to fractional CFO leadership, our work starts with the same question this article asks: is your budget telling you the truth about your jobs? Talk to our team.
