Financial Visibility · Project margin
I don’t know whether a project made money until months after it finishes
You find out too late whether a project made money when costs, time and change requests are only reconciled after delivery. The fix is not better accounting. It is a weekly margin view per project, owned by a named person, so overruns show while there is still scope, price or staffing to change.
This is written for founder-led firms between £500k and £15M that deliver work as projects, contracts or matters — agencies, construction businesses and professional practices. It is written from inside that work: an embedded operator installs the view, develops the person who owns it, and hands it over.
Why the answer arrives months late
The margin arrives late because three things are recorded late. Time is booked after the week it was worked, supplier and subcontractor costs land at month end, and change requests are agreed in conversation and priced afterwards or never. By the time the accounts close, the project has shipped and every decision that could have protected the margin has already been taken.
Your team books time in arrears
People record hours on Friday, at month end, or when someone chases. A week of senior time on a fixed-price job stays invisible until you have spent it three times over. Your team is not careless. Nobody set a deadline for booking time, and no number moves when they miss it.
Finance codes the cost at month end
Subcontractor invoices, freelancers, materials and licences reach the ledger weeks after someone committed to them. The commitment is the moment your margin changed. Coding it later records history. It governs nothing.
Someone agrees a change, nobody prices it
A client asks for one more round, a different finish, a second opinion. Your account lead says yes to protect the relationship. No rule says who prices that change, by when, or what happens if the client declines, so the team absorbs the work and the margin pays for it.
Your accounts answer a different question
Monthly accounts tell you what the business billed and what it spent. They cannot tell you which project earned and which one the others subsidised, because the ledger groups cost by type rather than by job. Both numbers can look healthy while a third of the work loses money.
None of those four is a finance failure. They are the same structural gap in four places: the number that decides a commercial outcome is produced after the decision, and nobody owns it in between. That is a Financial Visibility problem, and it is fixed the same way every visibility problem is fixed — by moving the number forward in time and giving it an owner.
What “made money” has to mean before you can track it
A project made money when its delivered margin beats the margin you quoted. That requires three defined numbers: the margin you priced, the margin you delivered, and the gap between them. A firm that tracks revenue and invoices can say what it billed. It cannot say what it earned.
| Number | What it is | Where it breaks |
|---|---|---|
| Quoted margin | Price minus the cost you planned to carry: hours at a loaded rate, subcontractors, materials, anything bought for this job. | Quotes are priced on a day rate or a feel for the job, with no planned cost written down, so there is nothing to compare against later. |
| Delivered margin | Price invoiced minus every cost the job carried, including rework and the hours nobody booked. | Your system attributes hours and costs to the month rather than the job, so nobody can calculate delivered margin at all. |
| The gap | Delivered margin minus quoted margin, per job and by job type. | Nobody reviews it, so you quote the same underpriced job the same way next quarter. |
| Margin at forecast | Where the job will land if it continues as it is going — the only one of the four you can still act on. | It does not exist in most firms, which is why the answer arrives after delivery. |
Write the definitions down once, in one sentence each, and apply them to every job. A firm that argues about whether a cost belongs to a project has not lost the number — it never agreed what the number was.
The weekly project margin view: five numbers, one owner
The working version of this is five numbers per live project, updated weekly by the person running it, reviewed in one meeting where decisions are made. It fits on a single page. It does not require new software: a spreadsheet maintained every week beats a project accounting system nobody updates.
Budget
Price agreed, and the cost you planned to carry to deliver it. Set at the point of sale, not reconstructed afterwards.
Cost to date
Hours booked at a loaded rate, plus committed external cost — committed on the day it is committed, not when the invoice arrives.
Forecast to complete
What the project lead believes it will take to finish. One honest estimate a week, in hours and pounds.
Unbilled change
Work agreed beyond the original scope that has not been priced or invoiced. A running figure, in pounds, visible to the person who can act on it.
Margin at forecast
Price plus priced change, minus cost to date and forecast to complete. The number that says whether to keep going as planned.
Three rules make it hold. Every live project has one named owner — the person who runs it, not the finance lead, who owns the accuracy of the inputs but never the outcome. The view is updated before the weekly leadership meeting, not inside it. And the meeting acts on exceptions: any project whose margin at forecast has moved by more than an agreed threshold gets a decision in the room that week, not a discussion.
That is three of the four pillars working together. Financial Visibility produces the number early enough to matter. A decision-first cadence forces a decision on it weekly. Outcome governance keeps one accountable owner on each project rather than a committee watching a dashboard.
The same gap, three sectors
The structure is identical across project businesses; the leak is not. In agencies it is scope that grows without price. In construction it is variations agreed on site. In professional practices it is time written off at the point of billing. Each needs the same five numbers and a different weekly question.
The retainer that grew
A retainer priced for four workstreams now delivers seven. Utilisation looks strong, everyone is busy, and the account still loses money. The weekly question is what was delivered beyond the scope this week, and who is pricing it. See fractional COO for agencies.
The contract with variations
Variations are agreed on site, recorded in a photograph or a text message, and submitted late or never. Programme slippage costs preliminaries nobody has counted. The weekly question is which variations are unpriced, and what the job now forecasts against the contract sum. See fractional COO for construction.
The matter that was written down
Time is recorded, then discounted at the point of billing to keep the client comfortable. The write-down is a pricing decision made by whoever raises the invoice. The weekly question is which matters are heading for a write-down, while the scope conversation can still happen. See fractional COO for professional services.
Who owns the number, and what to do when a project is already going wrong
The project lead owns margin at forecast. Finance owns whether the inputs are right. The founder owns the standard — what an acceptable gap is, and what happens when a job breaches it. A project that is already losing money needs a decision within the week, not a post-mortem after delivery.
When a live project is heading for a loss, the sequence that recovers most of it is the same in every sector.
Re-baseline it. Hours to date, committed cost, and a forecast to complete written by the person doing the work. Most founders discover the job went wrong earlier than they thought, which tells you where the estimate is wrong on every similar job.
Price every outstanding change. Total the unbilled work, put a number on it, and take it to the client as one conversation. Firms recover more of this than they expect, because the client rarely knows how much has been absorbed.
Decide the shape of the finish. Complete as planned, renegotiate scope, or stop. All three are legitimate. Drifting is not a decision, and it is the option most firms take by default.
Feed it back into the quote. The gap between quoted and delivered margin is pricing intelligence. Review it by job type each quarter, then change the rate, the scope wording or the qualifying question that let the job in.
Founders often expect the answer to be a new system. In practice the constraint is authority: the project lead can see the job sliding and cannot re-price it, extend the programme or refuse the change without the founder. Writing down those limits — what a lead can approve, up to what value, without asking — turns a weekly report into weekly decisions. That is the threshold-based authority map, and it is usually installed in the same four weeks.
What changes when the margin arrives on time
Two engagements, both project businesses, both where the founder was the only person who could see a job going wrong.
The margin was never the problem. The problem was that the only person who could see it was the founder, and he saw it in the accounts six weeks after the job was handed over.David Schofield
What it costs to install, and the first four weeks
Installing a weekly project margin view takes about four weeks in a firm that already prices its work: one week to define the numbers, one to get them produced from real jobs, and two to run the meeting until the decisions are being made without you. The cost sits inside existing engagements rather than beside them.
Most founders start with the 4-Week Operational Assessment, because project margin is rarely the only number arriving late. If the bank balance is the sharper pain, start with the gap between profit and cash instead. A project can earn its margin and still leave you short, and the two problems have different fixes.
Common questions
How is project profitability different from company profitability?
Company profitability tells you whether the business earned more than it spent in a period. Project profitability tells you which work earned it. A firm can hold a healthy overall margin while a third of its jobs lose money, subsidised by the rest. Only the second number tells you what to stop selling, what to re-price, and which clients to keep.
Do I need project accounting software, or is a spreadsheet enough?
A spreadsheet your project leads update every week beats project accounting software they update now and then. The weekly habit is what produces the number: five figures per live job, one owner, one review. Buy software once the manual version works and the volume of jobs makes it slow, not to create the habit in the first place.
Who should own the margin number: me, my finance person, or the project lead?
The project lead owns margin at forecast, because they are the only person who can change it. Finance owns whether the inputs are accurate. The founder owns the standard — the threshold that triggers a decision, and the authority the lead has to act without asking. Splitting it any other way produces a number that is reported and never acted on.
What do I do about a project that is already over budget?
Re-baseline it with an honest forecast to complete, price every outstanding change and put it to the client as one conversation, then decide whether to finish as planned, renegotiate the scope or stop. Record the gap between quoted and delivered margin, then price the next job of that type for what it costs to deliver.
How long before the numbers can be trusted?
Two to three weeks for the inputs to stabilise, once hours are booked weekly and costs are recorded when committed rather than when invoiced. Expect the first pass to be uncomfortable: the early forecasts are usually optimistic, and the correction itself is the finding.
Related reading: Financial Visibility: the 30–90 day forward view · Profitable on paper, broke in the bank · Decision-first cadences · Outcome governance · Fractional COO UK
Find out which projects earn
A 30-minute Operational Clarity Call is a structural read of your business, not a sales call. Bring one project you suspect lost money, and we will find the point where the number stopped being visible.
