Project management that resolves to margin

Why a task tracker cannot tell you what a project earned

Nearly every service firm we work with runs a competent project tool and still cannot answer, without a spreadsheet, which of last quarter’s engagements were profitable. That is not a failure of discipline. It is arithmetic that requires four things the task tracker does not hold.

• Loaded cost, not billed rate. Margin needs what the hour cost you — salary, employer taxes, benefits, and an overhead allocation — which lives in payroll and the ledger.
• Pass-through costs. Subcontractors, travel, software licences, and hosting billed to the client, which arrive as vendor bills in accounts payable.
• What was actually invoiced. Not what was planned. Write-offs, courtesy discounts, and disputed lines are the difference, and they happen in AR.
• Revenue recognition. On a fixed-fee engagement, revenue earned is a function of progress, not of what you billed — and that schedule belongs to accounting.

So the honest answer is that project profitability is an accounting calculation that happens to need a schedule as an input. Running it in a tool that does not hold the ledger means exporting four datasets and joining them by hand, monthly, forever. Most firms do this quarterly at best, which means they find out about a bad engagement roughly ten weeks after they could have done something about it.

What continuous margin changes

When forecast-at-complete recalculates on every posted timesheet and vendor bill, the conversation moves from post-mortem to intervention. A project trending fourteen percent over at week five is a scope conversation you can still have with the client. The same project discovered at close-out is a write-off and an awkward renewal.

It also changes how you price. After two or three quarters of clean margin data by project type and by client, you can see which engagements are systematically under-scoped — and that pattern is nearly always specific and fixable rather than general. It is usually one phase, one client, or one pricing assumption.

Where the agents help

The Project Coordinator agent drafts timesheets from calendar entries and activity, so consultants confirm rather than reconstruct their week on a Friday afternoon — which is where most time data goes wrong. It flags allocations that exceed capacity, tasks with no owner, and projects whose burn rate has diverged from plan. All of it is Level 1 by default: it proposes, a person commits.

If your delivery team lives in Jira or Asana

Keep them. Engineering teams should not be moved off their tracker to satisfy a finance requirement, and asking them to is how a rollout fails in month two.

We read tasks, assignments, and completion state from Jira, Asana, Linear, Monday, or ClickUp into the business graph, and add the layer they do not have — cost, billing, utilisation, and margin. Delivery keeps its tool. Finance gets its number. This is the more common arrangement among our software and agency customers, and it is the one we usually recommend first.