# How to review project cost variance by role
Source: https://korrel.ai/blog/review-project-cost-variance-by-role
Published: 2026-10-07
Updated: 2026-09-30
Topic: how to review project cost variance by role
Tags: cost-tracking, estimating
---
## Key points

- Start with planned and actual hours and cost rates for each delivery role.
- Split total cost variance into hours, staff mix, and cost-rate effects in a fixed order.
- Change scope, staffing, or rates only when the matching effect explains the difference.

Project cost variance by role is the difference between planned and actual labour cost for each delivery role. For an architect, those roles could include director, architect, assistant, and technologist.

A role total shows where cost moved, but it does not explain whether the project used more hours, more senior staff, or higher cost rates. Each cause needs a different response.

This guide shows how to review project cost variance by role with a role variance bridge. The bridge moves from planned cost to actual cost, one effect at a time.

## What data does the review need?

Use one completed project, stage, or deliverable, rather than calling the result final while a live stage still has incomplete time records.

Collect four values for each role. This worked example uses invented hours and cost rates:

| Role | Planned hours | Planned cost rate | Actual hours | Actual cost rate |
|---|---:|---:|---:|---:|
| Architect | 60 | £50 | 90 | £55 |
| Architectural assistant | 140 | £25 | 150 | £27 |

Use internal cost rates, not client charge rates, to value delivery time at the hourly cost for each role. Use one rate basis across both columns.

Reconcile the role hours to the project record before you calculate variance. Missing time can make a cheap role look efficient and move its work into another role.

If timesheets are still changing, set a clear cut-off. The guide to [how often consultants should complete timesheets](/blog/how-often-consultants-fill-timesheets) explains how data timing affects staffing and margin decisions.

Map each actual time entry to a planned role label, keeping those labels as they appeared in the estimate. Record unmatched work separately rather than forcing it into the nearest role.

## How do you review project cost variance by role?

Calculate planned cost and actual cost for every role:

`Planned role cost = planned role hours × planned role cost rate`

`Actual role cost = actual role hours × actual role cost rate`

Then use one sign convention:

`Cost variance = actual cost − planned cost`

A positive result means the project cost more than planned; a negative result means it cost less. State this convention on the review because some finance reports use the opposite sign.

With planned cost of £6,500 and actual cost of £9,000, the example has a total cost variance of £2,500 above plan.

[OpenStax explains labour rate and time variance](https://openstax.org/books/principles-managerial-accounting/pages/8-3-compute-and-evaluate-labor-variances) through comparisons with standard costs. This project review adds one more split: staff mix. Use the [cost tracking guides](/blog/topics/cost-tracking) to keep the underlying records consistent.

## Build the original role variance bridge

Calculate the hours effect first, followed by staff mix and then cost rate. The order matters because each step keeps a different input fixed.

First calculate the planned average cost rate:

`Planned average rate = total planned cost ÷ total planned hours`

For the example, £6,500 divided by 200 hours gives £32.50 per hour.

### Step one: isolate the hours effect

The hours effect values the change in total hours at the planned average rate:

`Hours effect = (total actual hours − total planned hours) × planned average rate`

Actual hours were 240 against 200 planned. The hours effect is therefore `40 × £32.50 = £1,300`.

This step answers one question. What would the extra hours have cost if the planned staff mix and planned rates had stayed unchanged?

### Step two: isolate the staff-mix effect

To compare staff mix, which means each role's share of total hours, calculate the planned and actual shares for every role.

`Role share = role hours ÷ total hours`

Then calculate the staff-mix effect:

`Mix effect = total actual hours × sum of ((actual role share − planned role share) × planned role rate)`

The architect's share rose from 30% to 37.5%. The assistant's share fell from 70% to 62.5%.

The mix effect is `240 × ((7.5% × £50) + (-7.5% × £25)) = £450`.

This mix effect is adverse because the higher-cost role used more of the actual time, even before you consider the changed rates.

### Step three: isolate the cost-rate effect

The cost-rate effect values each rate change against actual role hours:

`Rate effect = sum of (actual role hours × (actual role rate − planned role rate))`

The calculation is `(90 × £5) + (150 × £2) = £750`.

The bridge now reconciles exactly:

| Bridge step | Cost movement | Running cost |
|---|---:|---:|
| Planned cost |  | £6,500 |
| Hours effect | £1,300 | £7,800 |
| Staff-mix effect | £450 | £8,250 |
| Cost-rate effect | £750 | £9,000 |
| Actual cost |  | £9,000 |

The three effects total £2,500, which equals actual cost minus planned cost. If your bridge does not reconcile, check the role mapping, hours, rates, and rounding.

## What does each effect tell you to inspect?

Review the stage output, coordination load, revision count, and missing tasks to investigate the amount of work shown by the hours effect. Do not assume the team worked slowly.

A stage total can hide its component work. [GAO's cost estimating guide](https://www.gao.gov/products/gao-20-195g) includes a work breakdown structure and updating estimates with actual costs. For this architectural review, compare the stage's tasks and role hours rather than relying on memory.

The staff-mix effect shows who did the work, so ask why senior input replaced planned junior time. The cause might be a difficult decision, weak delegation, late review, unavailable staff, or work that required a different skill.

Check salary changes, contractor use, project-specific overrides, and an outdated role-rate table to explain the value assigned to time in the cost-rate effect. Keep this separate from a changed client price. This bridge measures delivery cost.

The role with the largest cost variance is not always the cause. An architect can show an adverse cost variance because the estimate assigned work to an assistant that needed an architect's judgement.

For an active job, connect this role review to a [weekly project margin review](/blog/weekly-project-margin-review). The bridge explains the labour movement. The margin review decides whether current scope, forecast, or commercial action must change.

## How should you investigate the cause?

Start with the largest effect, not the largest role total, and use evidence from the same period to investigate it.

For an hours effect, compare planned and actual tasks to find the first task where hours changed. Read that task's instruction and approval, then its change record.

For a staff-mix effect, compare the planned responsibility with the person who performed it. Ask the project lead what blocked delegation. Record the cause in plain terms.

For a cost-rate effect, compare the estimate date with the effective date of each rate. Check whether the estimate used a standard role rate or a named person's rate.

Use one cause code for each material effect. Examples include extra client revision, planning response, coordination omitted, senior review repeated, staff unavailable, contractor substituted, or rate table outdated.

Do not use “adverse cost variance” as a cause: it repeats the result rather than explaining it. A useful cause must identify an estimate, scope, staffing, or rate assumption that you can test.

If the evidence shows extra work after sign-off, review the process for [charging for design changes after a RIBA stage](/blog/charging-for-design-changes-after-sign-off). Do not hide an unpriced scope change inside a future productivity target.

## What should you change next time?

Match the change to the effect.

When hours caused the variance, change the task quantity or stage allowance. Add the omitted activity to the estimate. Define included reviews and coordination points. Keep role shares and cost rates unchanged unless the evidence also supports those changes.

When staff mix caused the variance, change the staffing plan. Assign the work that needs decisions to the architect from the start. Give suitable production work to the assistant. Add a review point before junior work can create senior rework.

When cost rates caused the variance, update the role-rate table or name the expected contractor. State the date and source of the new rate. Do not add more hours to compensate for an old rate.

Test the change on comparable work. One project can contain an unusual client, site, or approval path. A repeated effect across similar stages gives stronger evidence for a standard estimate change.

Keep a short decision record:

1. Name the role and stage.
2. Record each effect: hours, staff mix and cost rate.
3. Attach the cause evidence.
4. State the estimate or staffing change.
5. Set the next review point.

This record makes the cost report useful for the next estimate and stops one unexplained adverse cost variance from becoming a general fee increase.

## Make role variance part of project close-out

Run the bridge after each completed stage when the next stage can still benefit. Run it again at project close-out before you estimate comparable work.

Keep the planned role structure, actual time, cost rates, and cause notes together. Use the same sign convention and calculation order on every review.

Korrel compares forecast and actual cost by role for active projects. [Start your trial](https://app.korrel.ai/signup) and use the role variance bridge to choose one supported change for your next estimate.

## Common questions

### What is project cost variance by role?

It is the difference between planned and actual labour cost, grouped by delivery role. A useful review then separates the effect of total hours, staff mix, and cost rates.

### Should a favourable role variance always reduce the next fee?

No. First check whether it came from a repeatable delivery change or one unusual project. Keep the fee unchanged until comparable completed work supports a new assumption.