A services company with healthy overall margin can be running a third of its projects at a loss. The average conceals it, and the average is what most companies look at.
- ›Why averages mislead
- ›The four numbers per project
- ›Why this is hard to measure
- ›Acting on what you find
Why averages mislead
Portfolio margin is a weighted blend. One large profitable retainer can mask several small loss-making projects, and the loss-makers usually share characteristics: a specific client, a type of work, or a delivery lead who under-scopes.
Those patterns are the actionable finding, and they are invisible at portfolio level.
The four numbers per project
Contract value. What you sold, including change orders. The last part matters: unbilled scope changes are the most common margin leak.
Delivered cost. Hours consumed times loaded cost, plus direct expenses. Loaded cost, not salary; the difference is significant.
Margin, in currency and percentage. Both, because a 40% margin on a small project and 15% on a large one need different responses.
Budget consumption against completion. The forward-looking one: 70% of budget spent on 40% of scope is a problem you can still fix.
Why it is hard to measure
Each number lives in a different system. The contract value is in the CRM, the hours are in a time tracker, the costs are in payroll, the expenses are in accounting, and the completion is in the project tool.
Producing project margin therefore becomes a quarterly reconciliation exercise, which means it happens quarterly at best, which means loss-making projects finish before anyone notices.
This is the clearest example of consolidation paying for itself operationally rather than in licensing. Where these live in one system, margin is a view. See business software for agencies.
Acting on what you find
Under-scoped work: fix estimation for that work type, and record change orders as billable events rather than absorbing them.
A specific client consuming more than they pay: this is a pricing conversation, not a delivery problem. Have it with data.
A delivery pattern: if the same lead's projects consistently overrun, that is a coaching conversation, and it requires the data to be about the projects rather than about the person.
Work types that never make money: the hardest and most valuable finding. Some service lines are structurally unprofitable, and the answer is to stop selling them.
Tip: Show budget consumption to the people delivering the work, in real time. Most overruns are preventable by the person causing them, who currently has no visibility into it.
Measuring realization
Alongside margin, track realization: what share of billable hours actually gets invoiced. Write-offs and unbilled scope live here, and it is often a bigger leak than cost overrun.
FAQ
Do we need time tracking for this?
For hourly work, yes. For fixed-fee work you still need effort recorded somehow, or cost is a guess.
What margin should we target?
Varies widely by service type and market. Your own distribution is more useful than a benchmark: look at the spread across projects, not the middle.
How often should we review?
Live for budget consumption, monthly for completed-project margin, quarterly for patterns by work type and client.
Brainis holds deals, projects, time, costs, and invoices in one system, so project margin is a view rather than a reconstruction. See it for agencies.
Sharing insights on business operations, AI, and modern team management.
Run your company on Brainis
All 11 Operating Systems on every plan, from $29 a month. No per-seat pricing — you pay for AI capacity, not headcount.
See pricing