Job cost variance

What it is

Every construction job starts with a budget, broken down by cost code (concrete, framing, electrical, and so on). As work happens, costs post against those codes and the project manager updates an estimate of what is left to spend.

Job cost variance compares the two views: the budget, adjusted for approved change orders, and the projected final cost. It is usually reviewed for each open job every month, then rolled up by project manager, division, and company.

Why operators track it

• It catches overruns while there is still time to act. An overrun found early can be managed through scope, crew, or change orders. Found late, it is much harder to recover.
• It drives projected margin. If a job will cost more than budgeted and the contract value has not changed, the expected margin shrinks by the same amount.
• It feeds the WIP schedule. Projected cost at completion is also the denominator for percent complete, which drives earned revenue and over/under billing.
• It improves the next bid. Variance by cost code shows estimators which assumptions were wrong.

How it is calculated

Job cost variance = revised budget − estimate at completion

• Revised budget = original budget + approved change order budget
• Estimate at completion (EAC) = cost to date + estimated cost to complete
• Variance % = job cost variance ÷ revised budget

The estimated cost to complete comes from the project manager's forecast, not a formula. That is deliberate: it captures what the PM knows about the work remaining.

Worked example

Job 2417 has an original budget of $1,200,000. Approved change orders add $80,000, so the revised budget is $1,280,000. Cost to date is $910,000 and the PM expects $420,000 more, so EAC is $1,330,000.

How to read it

• Positive means under budget, negative means over under the convention used here (budget minus cost). Some teams flip the sign. Pick one and label every chart.
• The trend matters more than the level. A job sitting at −2% for six months is a different situation from one that moved from 0% to −2% last month.
• Look at cost codes before conclusions. A small total variance can hide a large overrun in one code offset by savings in another.
• We do not publish a "good" range. CFMA's Construction Financial Benchmarker covers related profitability ratios; we will cite it once we confirm citation terms.

Definitional choices and common pitfalls

• Variance at completion vs cost-to-date variance. This page uses budget vs projected final cost. Some teams compare budget × percent complete to cost to date instead, when they trust physical percent complete more than PM forecasts.
• Pending change orders. Including unapproved change orders makes the budget look healthier than it is. Excluding them can make a job look over budget for work the owner will likely pay for. Many teams show pending change orders as a separate column.
• Committed cost. Purchase orders and subcontracts that are signed but not invoiced are not cost to date. They are useful as a check on the cost-to-complete estimate.
• Stale forecasts. If PMs do not update cost to complete, EAC stays flat and overruns stay hidden.

How different roles use it

Questions it answers

• Which open jobs are projected to finish over budget?
• Which cost codes are driving the overrun?
• Did any job move significantly since last month?
• Which PMs' forecasts tend to hold up?

How to present it in a report

• Start with a table of open jobs: revised budget, EAC, variance in dollars, variance as a percent. Sort by largest negative variance.
• Let the reader drill from a job into its cost codes.
• Show the change since last month next to the current value.
• State the sign convention and whether pending change orders are included at the top of the report.

Data you need

Fields

What the raw data looks like

Example data A few rows as they might come out of the source system.

Where this data usually lives

• Construction ERP and accounting holds budgets, job cost, change orders, billing, and retainage. For example: Sage 300 Construction and Real Estate, Viewpoint Vista (Trimble).
• Project management often holds change orders and their status, and sometimes budgets and forecasts. For example: Procore.
• Estimating produces the original budget by cost code before the job starts.
• Payroll and time tracking feed labor cost and hours to job cost.
• Spreadsheets often hold the PM's cost-to-complete forecast or the WIP schedule itself.

The same field can live in more than one system. Agree with the customer which one is the source of truth.

Common data gaps and traps

• Change orders not entered. Field work on a change that has not been written up makes the job look over budget.
• Forecasts not updated. If cost to complete is stale, the estimate at completion stays flat and overruns stay hidden.
• Forecasts kept outside the system. Some PMs keep cost to complete in a spreadsheet. Ask where it lives.
• Cost codes that do not match. Actual cost posted to a code with no budget line drops out of a simple join.
• Committed cost mixed into cost to date. Purchase orders and subcontracts not yet invoiced are not cost to date.

Related metrics

• WIP over/under billing: uses EAC to compute percent complete.
• Gross margin by job: variance flows straight into projected margin.
• Labor productivity by cost code: often the cause of labor variance. Coming soon.
• Change order cycle time: slow approvals keep revised budgets out of date. Coming soon.

FAQ

What is a good job cost variance?

Zero or positive. The more useful question is whether the variance is changing month over month. Projected margin that keeps fading is a warning sign.

Is job cost variance the same as cost variance in earned value?

They are related. Earned value cost variance is earned value minus actual cost to date. Job cost variance at completion compares budget to projected final cost.

Should committed costs be included?

Not in cost to date. Committed but uninvoiced costs are better used to sanity check the cost-to-complete estimate.

How often should job cost variance be reviewed?

Monthly at minimum, aligned with the WIP schedule. Weekly on large or troubled jobs.

revised_budget          = original_budget + approved_change_order_budget
estimate_at_completion  = cost_to_date + estimated_cost_to_complete
job_cost_variance ($)   = revised_budget - estimate_at_completion
job_cost_variance (%)   = job_cost_variance / revised_budget * 100

WITH budget AS (
  SELECT b.job_id, b.cost_code,
         SUM(b.original_budget) AS original_budget
  FROM job_budget b GROUP BY 1, 2
),
cos AS (
  SELECT job_id, cost_code, SUM(budget_amount) AS approved_co_budget
  FROM change_orders
  WHERE status = 'approved' AND approved_at <= :as_of
  GROUP BY 1, 2
),
actuals AS (
  SELECT job_id, cost_code, SUM(amount) AS cost_to_date
  FROM job_costs WHERE posted_at <= :as_of GROUP BY 1, 2
),
ctc AS (
  SELECT job_id, cost_code, estimated_cost_to_complete
  FROM cost_forecasts WHERE period = :period
)
SELECT
  j.job_id, j.pm_id, j.division, b.cost_code,
  b.original_budget + COALESCE(c.approved_co_budget, 0)                AS revised_budget,
  COALESCE(a.cost_to_date, 0) + COALESCE(f.estimated_cost_to_complete, 0) AS estimate_at_completion,
  (b.original_budget + COALESCE(c.approved_co_budget, 0))
    - (COALESCE(a.cost_to_date, 0) + COALESCE(f.estimated_cost_to_complete, 0)) AS job_cost_variance
FROM jobs j
JOIN budget b       ON b.job_id = j.job_id
LEFT JOIN cos c     ON c.job_id = b.job_id AND c.cost_code = b.cost_code
LEFT JOIN actuals a ON a.job_id = b.job_id AND a.cost_code = b.cost_code
LEFT JOIN ctc f     ON f.job_id = b.job_id AND f.cost_code = b.cost_code
WHERE j.status = 'open';
-- Note: cost codes with actuals but no budget line need a FULL OUTER JOIN in production.

Version 1.0 · Last reviewed 2026-10-20 · Machine-readable definition