Project Profitability Analysis: How to Measure It Properly and What to Do When the Numbers Are Ugly
A proper project profitability analysis compares what each project earns against the fully loaded cost of the people who delivered it, measured at completion rather than at quote, so that you can see which projects and clients actually make money. A first review across live projects needs three things, namely honest timesheets, a loaded cost rate for every person and a margin figure tracked against the original estimate, and most firms can assemble those within a month without buying new software.
The suspicion that certain clients lose money is common in technology services firms of 20 to 60 people, largely because the finance system reports revenue by client but cost only by department, which means the loss is real but invisible. What follows is a way to make it visible, and then a way to decide what to do about it.
Why the numbers you have cannot prove what you suspect
Most growing firms can tell you revenue per project to the penny, while cost per project is either missing or calculated from a blended day rate that someone set two years ago and nobody has revisited since. The result is a report in which every project looks moderately healthy, because the cost side has been averaged into meaninglessness, and the founder's instinct that one particular account is a drain has nothing to stand on.
The gap usually comes from three places: time that is never recorded because it feels like overhead, a cost rate that ignores the difference between a senior architect and a graduate, and fixed-price work that is judged on whether it was invoiced rather than on what it took to deliver.
Allocating people costs to projects honestly
People are almost always the largest cost in a services business, so how you allocate them decides whether the analysis means anything.
Build a loaded cost rate for each person
Start with salary, then add employer National Insurance, pension contributions, benefits and any recurring tooling that is specific to that person, and divide the total by the hours they are realistically available to work in a year once holiday, bank holidays, training and sickness are taken out. That gives you a cost per available hour, which is the figure you multiply by the hours booked to each project. Grouping people into four or five bands by seniority is acceptable if individual rates feel too sensitive, provided the bands reflect real differences in cost.
Decide where non-billable time goes
Overhead such as sales, management and internal meetings should sit outside project cost, because loading it on to projects makes every project look worse and hides which ones are genuinely weak. Time that was spent on a client but not billed, however, such as rework, extra workshops or the account manager's weekly calls, belongs to that project, since it is precisely the cost the analysis exists to expose.
Rule of thumb: if the hours were spent because this client exists, they belong to this client's projects, whether or not anyone could invoice them.
Get timesheets that are honest rather than tidy
Timesheets filled in on Friday afternoon tend to match the budget rather than the reality, which is the opposite of what you need. Make it clear that over-budget hours will not count against the person recording them, because the purpose is to find mispriced work, not to blame the people who absorbed it. Where accurate time recording is patchy, the related point about measuring utilisation properly applies here too, since both numbers depend on the same discipline.
Margin at quote versus margin at completion
Margin at quote is the gross margin you expected when the work was sold, calculated from the estimated hours, the planned team mix and the agreed price. Margin at completion is what the project actually returned once every hour, every substitution of a more senior person and every concession has been counted.
The difference between the two, which is often called margin erosion or slippage, is more useful than either figure on its own, because it tells you where the problem originates. A project that was quoted at a thin margin and delivered as planned has a pricing problem, while a project that was quoted well and finished badly has a delivery or scope problem, and the two call for entirely different conversations.
| Pattern | What it usually means | Where to look first |
|---|---|---|
| Low at quote, low at completion | Work sold too cheaply | Pricing and discounting habits |
| Healthy at quote, low at completion | Estimates or scope control failing | Change control and estimating |
| Healthy at both, client still feels painful | Cost sitting outside the project | Unrecorded account management time |
For live projects, track a third figure as well, the forecast margin at completion, which is hours spent so far plus an honest estimate of hours remaining. This is the number that gives you time to act before the project closes.
Running a first review across live projects within a month
A first pass does not need to be perfect, and a review that is roughly right across every live project is worth more than a precise one covering three.
- Week one: list every live project with its contracted value, commercial model and quoted hours, and agree loaded cost rates with finance.
- Week two: pull actual hours to date, ask each project lead for an honest estimate to complete, and record any known unbilled work.
- Week three: calculate quoted margin, forecast margin at completion and the gap between them for each project, then roll the results up by client.
- Week four: rank projects by the size of the forecast loss in pounds rather than percentage, and agree an action for the worst five.
Ranking by pounds matters because a small project losing a large share of its value is irritating, whereas a large account losing a modest share can quietly consume a quarter's profit. Once the review is done, the headline figures belong on the monthly scorecard described in our piece on professional services KPIs, so that the exercise becomes a habit rather than an event.
Triage for projects that are underwater: reprice, restructure or resign
When a project or client turns out to be losing money, there are three honest options, and the analysis should tell you which one fits.
Reprice
Repricing suits clients who value the work and whose loss stems from a rate that was set too low or never increased. It is usually the first conversation to have, because it is the least disruptive and the one most clients half expect, and our guide to raising prices without losing the clients that matter covers how to handle it.
Restructure
Restructuring suits projects where the price is reasonable but the shape of the work is wrong, for instance a fixed-price contract with open-ended scope, a team weighted too heavily towards senior people, or change requests absorbed without charge. Moving to a different commercial model, tightening change control or changing the team mix can bring margin back without changing the headline rate.
Resign
Resigning the work, by completing current obligations and declining renewal, is right when a client will neither pay more nor accept a different structure, and when the account consumes senior attention that profitable clients need. It is the conversation founders avoid longest, often because the client was an early supporter, which is understandable but rarely a sound commercial reason.
The hardest conversation is usually the one with the client who has been with you longest, which is also why it tends to be the most expensive one to postpone.
Which conversation first: start with the largest loss in pounds where the cause is clearly pricing, because repricing is quick, low risk and builds confidence for the harder restructuring and resignation decisions that follow.
Where Vitori fits
Plenty of firms can run this review themselves, and a capable finance director with a cooperative delivery team will get a long way with the steps above, whether with Vitori or anyone else. Where outside help earns its place is in the part that follows the spreadsheet, namely having the repricing conversations, redesigning estimating and change control, and making sure margin discipline holds once the attention moves on. Vitori's Operator model embeds in exactly that work, using the Operational Scale Framework to connect delivery and operations so that profitability is visible every month and decisions about it no longer depend on the founder, which is ultimately the aim: a business that runs, and scales, without the founder in every decision. If that sounds relevant, you can get in touch with Vitori to talk it through.
Common questions about project profitability analysis
Should project profitability include overheads like rent and management?
Project profitability analysis is most useful at gross margin level, meaning direct delivery costs only. Rent, sales and general management sit below the line, because spreading them across projects makes every project look weaker and obscures which ones are genuinely mispriced or poorly delivered. Overhead coverage is better judged against total gross margin for the whole business.
Do I need special software for project profitability analysis?
A first project profitability analysis can be run in a spreadsheet using exports from your timesheet and accounting systems. Dedicated professional services automation tools make the review faster and repeatable, but they only help once timesheets are honest and cost rates are agreed, so it is worth proving the method manually before committing to new software.
How often should project margins be reviewed?
Forecast margin at completion for live projects is worth reviewing monthly, ideally as part of the regular management meeting, so that problems surface while there is still time to act. A fuller review of completed projects and client-level profitability every quarter helps identify patterns in estimating, pricing and scope control that a single project would not reveal.