Small businesses run projects constantly and almost never call them projects. The new website. The move to different premises. The software change. The recruitment drive. Each one has a start, an end and a cost, and each one is run out of somebody's head between other jobs. Then eleven months later the website is still not live, and nobody can say exactly why, because there was never a written statement of what it was supposed to be.
The model — often called the project triangle — says three constraints govern any piece of work, and you may fix two. The third has to flex. Fix all three and something still gives, and the thing that gives is quality, whether or not anybody agreed to that.
The three corners, and what sits in the middle
Scope — what gets delivered
The single most under-written of the three, and the one that moves without anybody deciding. Scope creeps by one small favour at a time, each too trivial to raise.
Time — when it lands
Sometimes genuinely immovable, because a lease ends or a season starts. More often it is a hope with a date attached, which is a different thing and should be said out loud.
Cost — what it consumes
Money, and your people's hours. In a small business, cost is usually the corner nobody will move, because there is no budget line for it. Internal hours are the invisible half: nobody costed the website, so nobody noticed it absorbed sixty hours of the operations manager's year, which is a fortnight she did not spend on the thing you promoted her to do.
Quality — the thing that pays for the rest
Nobody ever agrees to lower quality. It is simply what happens when the other three are held rigid. Usually it is the testing, the training or the handover that gets sacrificed — exactly the parts you pay for again later.
Why does this matter to an owner?
Two reasons, and they cost different money. Internally, undefined projects consume your best people in dribs and drabs, and because the project has no end date it never formally finishes, so it stays on the list. Owners then look at nine live initiatives, feel behind on all of them, and start none properly. Six half-run projects produce less than two finished ones and cost considerably more in attention.
Externally, scope is where margin dies. In client work the profit leaves the job quietly, and because nobody logged it, the next quote is built on the same wrong assumptions.
Decide which corner flexes, before you need to
The useful discipline is choosing the flexible corner at the start, in daylight, rather than in week nine when everyone is tired and blaming each other.
If the date is genuinely immovable, then scope is your flexible corner and you should be planning what launches later, not pretending everything lands at once. If the specification is genuinely fixed, because it is a regulatory requirement or a customer's non-negotiable, then the date or the budget moves and somebody senior needs to say so in week one. And if cost is the corner that cannot move, say that plainly: it means either the date or the scope must give, and the honest conversation is which of the two you would rather explain.
A worked example
Illustrative arithmetic, recognisable to anyone who sells time. An agency quotes a website build at £12,000: 100 hours at £120, with a delivered cost of £60 an hour. Planned profit £6,000, a 50 per cent margin.
Then the project runs. The client asks for a booking form, described in the email as a small addition, which takes 14 hours. Two extra rounds of revisions beyond the two quoted take 9 hours. Content arrives three weeks late, and the restart costs 6 hours of re-briefing because the developer had moved on to something else.
That is 129 hours against 100. Revenue is unchanged at £12,000, so the effective rate is £93 an hour rather than £120. Delivered cost is £7,740, and profit falls from £6,000 to £4,260 — nearly 30 per cent of the profit gone, with the margin down from 50 per cent to 35. Not one of those three changes was unreasonable. Every one was a legitimate variation that should have taken sixty seconds to raise. Nobody raised any of them, because each felt too small to make a fuss about, and because raising the first one is the awkward part.
How to apply it this week
- Write one page for your biggest live project. What it delivers, by when, at what cost, and who owns it. If nobody can write that page today, that is the finding.
- Write the exclusions list. Five lines of what is not included. Every line is a conversation you will not have to have from a weak position later.
- Name one owner, not a committee. One person accountable for delivery, everybody else contributing. Shared ownership is how projects die slowly.
- Put a number on internal hours. Estimate them before you start and log them as they go. Internal projects feel free, and they are the most expensive thing in the building.
- Agree the variation rule up front, and use it the first time. Anything outside scope gets a note, a price and a yes or no before the work starts. The first small variation sets the precedent for the whole job.
- Review the three constraints weekly, for fifteen minutes. Which corner moved this week, and what did we consciously let move to absorb it? That question takes a minute and prevents most overruns.
The mistake most owners make
They confuse scope with a to-do list, and they write only what is included. The half of the document that saves you is the exclusions: what is not in this price, how many rounds of revision, what we assume you will provide and by when, and what happens to the timeline if that is late.
The second mistake is treating variations as a relationship risk. Owners stay quiet because they do not want to seem money-grabbing. What actually damages the relationship is a project running late and over budget while the client is told nothing, followed by a difficult conversation about an invoice they were not expecting. Raising a variation at the time is a sign of control. Raising it at the end is a sign that you lost control weeks ago.
The third is fixing all three constraints and hoping. Everyone nods, quality absorbs the difference, and you pay for it twice.
Questions to ask yourself
- Out of 10, how clearly could you state what your biggest current project delivers and what it will have cost when it lands?
- Which project has been almost finished for the longest, and what would happen if you formally killed it this week?
- On your last three client jobs, how many hours went in beyond what was quoted, and who absorbed them?
- When you last fixed scope, time and cost all at once, what actually gave way instead?
The takeaway: a business that cannot land a project on a date is not short of talent. It is short of a written agreement — internally as much as with customers — about which of the three corners is allowed to move.
Where this fits
Systems, delivery and margin sit inside Business Coaching. The internal-projects half of this is really a time problem, so read Productivity alongside it. And because overruns show up in the bank before they show up in the accounts, Profit vs Cash is the natural companion.
Get the next one in your inbox
One practical, plain-English guide for business owners each week. No spam, unsubscribe any time.
Common questions
What is the difference between scope creep and good service?
Good service is doing the agreed work well and occasionally choosing to absorb something small as a deliberate decision. Scope creep is absorbing it by default, without a note, a price or a conversation, until the job's margin has gone and nobody can say where. The distinguishing feature is whether it was decided or allowed. Log every variation even when you choose not to charge for it, because the log is what makes the next quote accurate. A firm that never records the extras keeps quoting from assumptions its own delivery history has already disproved, and repeats the same underpriced job indefinitely.
How do you raise a variation without damaging the client relationship?
Raise it the first time it happens, in a sentence, as a normal part of running the job: this is outside what we quoted, it will take about this long, it costs this, shall we go ahead? Raising it early reads as control. What actually damages a relationship is silence followed by a late, over-budget project and an invoice the client was not expecting. The awkward one is always the first, because it sets the precedent for the rest of the job. Agreeing the variation rule at the start, in writing, means neither side has to invent the process while under pressure.
Why do internal projects never seem to finish?
Because they have no client, no deadline anyone is accountable for, and no cost that appears anywhere. The hours come out of your best people in small pieces, and since nothing is logged, nobody notices how much has gone in. Without a written end point the project never formally finishes, so it stays on the list, and a list of nine unfinished initiatives makes an owner feel behind on all of them and start none properly. The fix is unglamorous: one page defining what it delivers and by when, one named owner rather than a committee, and an estimate of internal hours tracked as they are spent.