Why Projects That Looked Profitable Lose Money

A service business almost never discovers a loss-making project while it is running. It discovers it afterwards, if at all, and usually by accident — someone notices the same client keeps coming back with small requests, or that a quarter was busy and the bank balance did not move. The invoice was paid. The quote was accepted. The money still is not there.

The reason is simple and uncomfortable: revenue is recorded carefully and effort is not recorded at all. One side of the margin calculation is in your accounting system to the rupee; the other side exists only in people's heads.

The four places margin actually goes

Unbilled revisions. The quote covered two rounds. The project had five. Nobody re-quoted, because each individual request was small and refusing felt disproportionate. Five small yeses is the commonest way a healthy quote becomes a loss.

The long tail after delivery. The project is marked complete and invoiced, and then runs for another six weeks at a trickle — a change here, a question there. The work is real and the project is closed, so the hours land nowhere.

The wrong people doing the work. A job quoted on the assumption of a junior executing it gets done by a senior because the deadline slipped. Same output, materially different cost, and nothing in the system notices.

Coordination. Meetings, chasing approvals, re-explaining. It is genuinely part of the job and it is almost never in the estimate, which is why the gap between quoted and actual is usually widest on the projects with the most stakeholders rather than the most work.

Why the usual response does not help

The instinct after a bad project is to quote higher next time. That is a blunt correction, and it loses you the jobs you would have made money on while still mispricing the ones you would not. A twenty per cent margin added across the board mostly prices you out of your best work.

The sharper move is to find out which kind of project overruns. Almost always it is a type, not a random scatter — a client, a service line, a size of job, a particular stakeholder structure. Once you can name the type, you can price that type differently, scope it harder, or stop selling it. That is a decision you cannot make without data.

The honest prerequisite

None of this works if nobody logs time, and no software fixes that. A tool can make logging quick; it cannot make it happen. This is the point at which most "we should track profitability" initiatives quietly die, and it is worth facing before you buy anything.

Two things make it stick in small teams. Keep it coarse — hours against a project, not six-minute increments against tasks; precision nobody sustains is worse than a rough number everybody records. And make the output visible to the people doing the logging, not just to the owner. A team that sees which projects ran long starts scoping better on its own, which is worth more than the report.

Where Grovia takes a position here: when no time has been logged, it says the cost is unknown and why, rather than printing a confident zero. A profitability figure built on no data is worse than an empty one, because it gets believed.

Who can skip this entirely

If you bill hourly against logged time, you already have this and the project either made money or you did not bill it. If you sell a product with fixed delivery effort, variation is small enough not to matter. This is a fixed-price-service problem specifically — agencies, consultancies, interior and build firms, anyone who quotes a number before knowing exactly what the work will be.

A test you can run this week

Take three finished fixed-price projects: one you remember as good, one you remember as painful, and one you barely remember. For each, write down what you invoiced, then ask whoever did the work for an honest estimate of days spent — not a reconstruction from timesheets, just their best recollection. Multiply by a rough daily cost. The painful one will lose money and you knew that. What you are looking for is the third project, the forgettable one. If that is thin too, your pricing is structurally off rather than occasionally unlucky, and no amount of working harder will fix it.

Grovia is a business platform built for Indian service companies — invoicing, GST reporting, projects, collections and field work in one place. Start a free trial and see whether it fits the way you already work.

Frequently Asked Questions

How do I know if a project is losing money before it finishes?

Compare hours logged so far against the share of work completed, not against the calendar. A project seventy per cent through its budgeted hours and thirty per cent through its scope will not recover on its own.

What is a realistic buffer to add to an estimate?

Rather than a flat percentage, look at your own completed projects and measure how far estimates ran over by type of work. Most teams find the overrun is concentrated in one or two kinds of job, which is a far more useful correction than raising every quote.

My team will not fill timesheets. What now?

Make it coarser. Hours against a project once a day is sustainable; detailed task-level tracking in a small team usually is not. A rough number everybody records beats a precise one half the team abandons.

How should I handle scope creep without souring the relationship?

Price the change at the moment it is asked for, not at the end. A small, immediate "that is an extra, here is the cost" is a normal business conversation. The same amount raised as a surprise on the final invoice is a dispute.

Is project profitability worth tracking for a team of five?

Yes, but only at the project level. At that size you need to know which jobs are worth repeating; you do not need per-task costing, and attempting it usually ends with nobody recording anything.