Turnover Is an Opinion. Cash Is a Fact. One Contract, Followed From Quote to Bank

You already know the saying. Turnover is vanity, profit is sanity, cash is reality. Everyone nods. Then everyone goes back to talking about the order book, because the order book is the thing you can say out loud at a networking event without feeling sick.

Here's the sharper version, I think. Your turnover is an opinion. It's a number you formed on a particular day, about what a job ought to be worth, based on what you knew at the time. Cash is a fact. It either arrived or it didn't.

The gap between the two is where most mid-sized UK manufacturers quietly lose their year.

So rather than talk about it in the abstract, let's follow one order. Just one. A real sized job of the kind you probably quoted last month: £186,000 for a run of fabricated stainless enclosures, quoted at a 28% gross margin, which means you expected to make around £52,000 on it. Good order. The sort you'd mention to your other half over dinner.

We'll walk it from the quote going out to the money clearing. And at every stage, we'll stop and ask the only question that matters: how much of that £52,000 is still there?

One thing before we start. Nobody in this story is incompetent. There's no villain, no disaster, no machine fire. That's deliberate. The leaks that matter aren't dramatic, they're ordinary, and they happen on jobs that everyone agrees went pretty well. That's exactly why they survive year after year. A crisis gets a meeting. A slow leak gets a shrug.

And the backdrop isn't especially kind. Make UK's latest survey has price pressure easing slightly but still elevated, with the UK prices balance down from +36% in Q2 to +30% in Q3 and expected to climb again in Q4. Growth is forecast at 0.9% for 2026, after a 0.2% contraction in 2025. So this is a year where volume won't quietly cover a sloppy job. The margin has to come out of the job itself.

Right. The quote's gone out.

1. The quote: your margin is decided before anyone touches a machine

Here's the uncomfortable bit. By the time that quote leaves your building, the margin on the job is basically fixed. Everything that happens afterwards can only make it worse. The shop floor can protect it, but it can't create it.

So what happened when yours went out?

The material was priced off a supplier quote from January. You bought in March. Steel moved about 4% in between, which on a £48,000 material line is roughly £1,900 straight out of your profit before a single sheet was cut. Nobody did anything wrong. The estimate just aged, quietly, the way estimates do.

Then there's setup. Your estimator allowed two hours because two hours is what it says on the sheet, and the sheet was written when a different person was running that machine. The job takes closer to three and a half. There's a well-documented version of this in job costing reviews: one manufacturer found a project quoted at 720 hours had actually taken 845, which was over £10,000 of margin on a single job. That business was targeting 32% gross margin and achieving 29.7%. Not a catastrophe. Just a permanent, invisible haircut.

Then the engineering. Twenty-two hours of drawing review, DFM suggestions and a sample before the order landed. Free, obviously, because that's how you win work. At a loaded £65 an hour that's £1,430 you gave away to get the chance to give away more later.

And then the last five minutes of the phone call. They asked for a bit off. You said 2.5%, because 2.5% felt small and the order was worth having. On £186,000 that's £4,650. Except it doesn't come off the price in any meaningful sense, it comes off the profit, and £4,650 is nearly nine percent of your expected margin gone in one sentence.

Running total: about £8,000 of your £52,000 has left the building, and you haven't started the job.

To be honest, this is the section most people skim, because it feels like finance rather than manufacturing. It isn't. It's the highest leverage hour in your whole business. An hour spent fixing your estimating basis is worth more than a week of chasing efficiencies on the floor, because it applies to every job you quote from now on.

What actually helps:

• Quote from your own actuals, not from memory or a rate card nobody has touched since 2023. If you close jobs without ever comparing estimated hours to real hours, you're guessing with confidence.
• Put a date and an expiry on every material price. Thirty days is normal. Beyond that, requote or carry a price variation clause.
• Decide your floor price before the call, not during it. If you're going to discount, get something back: a deposit, shorter terms, a longer run, tooling paid up front.
• Price pre order engineering into the job or set a limit on how much of it you'll do free. Both are fine. Doing it unconsciously isn't.

2. The terms: you just became their bank, and nobody told you

The purchase order lands. Everyone's pleased. Somebody puts it on the board.

And almost nobody reads the back of it, which is where the money is. Their standard terms say 60 days from the end of the month in which the invoice is received. Read that slowly. Invoice on the 2nd of the month and you're waiting nearly 90 days. Their terms attached to their PO, sent after your quote, so their paperwork is the one that counts. That's not sharp practice, it's just how it works when nobody pushes back.

This is the point where your business stops being a manufacturer and starts being a lender. You're funding their production.

And manufacturing carries this worse than anyone. Analysis of Department for Business and Trade data found that manufacturers wait around 47 days on average to be paid, the longest of any UK industry, which is put down to long, multi-tiered supply chains. Across the wider economy, the average UK invoice runs about 21 days beyond agreed terms and 52% of SME invoices are paid late. So the 90 days you agreed to is really a 110 day expectation with a straight face on.

Now count the money going the other way. Steel is ordered in week two and paid on 30 day terms. Wages go out weekly regardless. Subcontract finishing invoices you before the customer sees a panel. And the VAT on your £186,000 sale, £37,200 of it, falls due on the return covering the invoice date, whether you've been paid or not. Cash accounting would fix that, but the scheme caps out at £1.35m of VAT taxable turnover, so if you're a mid-sized manufacturer you're funding it.

Carry roughly £134,000 of job cost for about 90 days and, at overdraft or invoice finance pricing of somewhere near 9%, that's about £3,000. Gone. On one order. Nobody codes it to the job, so nobody sees it.

Here's the part I find genuinely odd, and I've raised it with a few managing directors who were surprised it existed. You can look this up before you quote. Large UK companies have to publish their payment performance: 2,933 of them are in the regime, and 35% admitted paying more than half their invoices outside agreed terms in their most recent filing. It's a free, searchable government database. Ten minutes with it tells you whether your lovely new customer pays in 34 days or 82. Almost nobody looks.

The law is moving, slowly. The Small Business Protections Bill, introduced in May 2026, would cap payment terms at 60 days, make interest at 8% above base mandatory, and give the Small Business Commissioner power to fine persistent late payers. Useful. But it's not expected to bite until 2027, so it doesn't help this order.

What actually helps:

• Check the payment record before you price, not after you're owed. Price the risk in if it's bad, or ask for staged payments.
• Stage the money against milestones: deposit on order, materials on delivery, balance on despatch. It's normal in manufacturing and far easier to ask for at the quote stage than halfway through.
• Get your terms accepted in writing, and keep the statutory interest clause in there. Right now that's 11.75%, base plus eight points, with £40 to £100 of fixed recovery costs per invoice. You may never charge it. Having it in writing changes conversations anyway.
• If you bid for public work, watch your own record too. Bidders for government contracts over £5m have to show an average payment time of 45 days or better.

3. The floor: where small favours quietly become free work

The job runs. It runs well, actually. Your team are good, which is rather the problem, because good teams absorb trouble instead of reporting it.

Four things happen.

A drawing revision comes through in week three. Six panels are already cut to the old version. That's about £2,200 in material and machine time in the skip, and because it went in the skip rather than to a customer, it never generates an invoice, an argument or a conversation. Somebody mutters. Everyone moves on.

The revision costs you two days, so the back end of the job goes to overtime to hold the despatch date. Forty hours at premium, call it £1,100.

A fixing kit arrives late from a supplier, so you pay £640 for next day freight to avoid missing the slot.

And then the customer's engineer phones your production manager directly, the way he always does, and asks whether you could just move the bracket while you're in there. Eighteen hours of work. Nobody raises a variation because it's only a bracket and he's a decent bloke and you want the repeat order. That's £1,170.

None of that is visible as one number anywhere in your accounts. Scrap sits in material variance. Overtime sits in wages. Freight sits in distribution. The bracket sits nowhere at all, because it was never priced, so it doesn't exist except as four thousand pounds of capacity that didn't earn anything.

That scattering is exactly why this stuff survives. The American Society for Quality's long standing rule of thumb puts quality related costs at 15% to 20% of sales revenue for many manufacturers, and scrap plus rework alone runs from about 0.6% of revenue at top performers to 2.2% at weaker ones, with the real bill typically three to five times the visible scrap cost. Those are estimates rather than gospel, and I'd treat any precise sounding benchmark with suspicion. But the shape is right, and the shape is the point: what you can see in the scrap bin is a fraction of what it cost you.

Running total on our job: roughly £13,000 of the original £52,000 is now gone. A quarter of your profit, on a job that everybody agrees went fine.

And here's the line I'd put on the wall if I were you. To earn that £13,000 back at the same margin, you'd have to win another £46,000 of work. Same heat, same risk, same late payment, a whole extra order just to stand still. Suddenly an afternoon spent on change control looks like a reasonable use of time.

What actually helps:

• Compare estimated against actual while the job is still open, not at year end. A weekly look at hours booked versus hours quoted catches drift while you can still price it.
• Route every change through one person with authority to say what it costs. Not a form. A two-minute rule that a revision triggers a price and date check before the work starts.
• Book expedited freight, overtime and rework to the job, not to a general overhead. If you can't see it against the order, you'll never believe it's real.
• Tell your team that reporting a problem is not the same as causing one. The instinct to absorb it quietly is the expensive instinct.
  

4. The invoice: the most underrated product you make

The enclosures go out on the 12th. Everyone's relieved. The job is done.

Except it isn't, because you haven't been paid, and the thing standing between you and being paid is a single sheet of A4 that nobody in the building thinks of as their job.

The invoice goes out on the 22nd. Ten days of drift, for entirely reasonable reasons. The proof of delivery hadn't come back. The person who raises invoices was off on the Friday. It didn't feel urgent because the money wasn't due for months anyway.

Then it bounces. The part number changed when you moved that bracket, so the line description doesn't match their purchase order, and their accounts payable system rejects anything that doesn't match. It sits in a queue for a fortnight before anyone tells you. You resubmit on the 9th of April.

And there's the killer. Their terms run 60 days from the end of the month in which the invoice is received. Your March invoice is now an April invoice. Ten days of internal drift and one wrong part number have just cost you a full calendar month of cash. Same job, same quality, same customer. One field on a document.

Meanwhile the VAT on the original sale is already sitting in your return, doing you no favours at all.

I'd genuinely argue the invoice is a product. It has a customer, it has a specification, and if it's out of spec it gets rejected exactly like a panel with the wrong finish would. The difference is that you'd never dream of shipping a panel without checking the spec, and yet most manufacturers I've come across have never asked their customer what a compliant invoice looks like. It takes one phone call. What's your PO reference format, where does it go, do you need a goods received number, which portal, who approves it, and how long does approval take?

This matters more than it used to, because the law is starting to notice. The new Bill includes a statutory time limit for raising disputes, and a right to a fixed sum where a purchaser raises a dispute late or without sufficient information. In other words, the quiet bounce is on borrowed time. Your paperwork being right is what lets you use that.

What actually helps:

• Invoice on the day of despatch. Not weekly, not when someone gets round to it. Every day of drift is a day of borrowed money, and near a month end it can be thirty.
• Agree the invoice specification at order stage, in writing, and keep it with the job file.
• Confirm receipt and acceptance within about five working days. Most terms run from receipt, so if you can't prove when they received it, you can't prove anything.
• Give one named person the job of getting invoices out clean, and measure them on days from despatch to invoice. It's the cheapest performance metric you'll ever run.

5. The last mile: getting paid is a process, not a personality

So the corrected invoice goes in on the 9th of April. Sixty days from the end of April makes it due at the end of June. It actually lands on the 17th of July, roughly eighteen days late, which is about average and therefore nobody escalates it.

Add it up. You quoted in early January. The money cleared in mid-July. You funded materials, wages and subcontract from late February. That's five months of your own cash inside someone else's business, on a job that was finished in March.

And you're not unusual. Sage's analysis of around 150,000 businesses found 49% of SME invoices overdue, with firms waiting an average of 27 days past terms. The same research found SMEs are now taking 37.1 days to pay their own suppliers, up from 31.9 days a year earlier, which tells you everything about how this travels down a supply chain. Someone stretched you, so you stretched the next one along.

The thing most manufacturers get wrong here, in my experience, is treating credit control as a personality trait. The sales director doesn't want to chase because he's protecting the relationship. The office manager chases when she's had a bad week. It's emotional, so it's inconsistent, so it doesn't work.

It's a cadence, and it's fairly dull. Confirm the invoice arrived. A week before the due date, check it's approved and in the payment run, because that's the moment when a problem is still fixable. The day after it's due, call, don't email. Then a firmer call a week later, then escalation, and escalation means a named person with a date, not a strongly worded message into a shared inbox.

Chase approval, not payment. By the time an invoice is 65 days old, the conversation is about someone else's cash flow. At day 7, it's just an admin question, and admin questions get answered.

So what was that order actually worth?

Where the money went Cost

Discount agreed in the last five minutes of the call              £4,650

Material price drift between quoting and buying                   £1,900

Pre order engineering never charged for                                £1,430

Panels scrapped after a drawing revision                               £2,200

Overtime to hold the despatch date                                        £1,100

Expedited freight on a late supplier part                                 £640

Bracket change nobody priced                                                 £1,170

Funding the job for five months                                               £3,000

Total £16,090

You expected £52,000. You made about £35,900. You quoted 28% and delivered a bit over 19%.

Nothing went wrong. Nobody was careless. The customer is happy and will probably reorder, which is lovely, because next time you'll do all of this again at exactly the same price.

That's the part worth sitting with. The turnover line in your accounts says £186,000 and it's telling the truth, technically. It just isn't telling you anything useful.

Bringing it together: one job, five stages, about thirty percent of your profit

Let's be clear about what happened, because it's easy to lose it in the detail.

The quote decided the margin before anyone switched a machine on. The terms turned you into an unsecured lender for five months. The floor absorbed problems quietly and turned them into free work. The invoice sat in a drawer for ten days and then bounced on a part number. And the chase happened on feeling rather than on a schedule. Five ordinary stages, none of them a scandal, and roughly a third of the profit gone between them.

That's the whole argument. Not that your business is badly run, but that the leaks live in the joins between departments, where nobody owns them and no single report shows them.

It's worth remembering what this eventually does to firms that never look. There were 1,886 manufacturing insolvencies in the 12 months to February 2026, about 8% of all UK cases. Most of those businesses had an order book. Quite a few of them were profitable on paper right up until the week they couldn't make payroll. Profit is an accounting position. Payroll is a Thursday.

So here's what I'd actually do, and it's smaller than you're expecting.

Pick one job. Just one, ideally a closed one from the last six months that everybody agreed went fine. Get the quote, the works order, the timesheets, the material invoices, the despatch note, the sales invoice and the bank statement on one table or one screen. Then walk it, stage by stage, and write down two dates and one number at each step: when did cash leave, when did cash arrive, and what did that step cost that the quote didn't allow for.

It takes about two hours. I've never seen anyone do it and find nothing. Usually there's one leak that's far bigger than the others, and usually it's the one nobody would have guessed at the start of the session.

Then fix that one. Not all of them. Just the biggest, properly, before moving on.

And this is where it loops back to growth, because most growth plans start from the wrong end. The instinct is to win more orders. But if a job leaks thirty percent of its profit and pays in five months, then winning more of them doesn't fix anything, it just scales the leak and puts more strain on the same working capital. Marketing that brings you more of your slowest paying customers isn't growth. It's a faster treadmill.

The better question is which work is worth chasing. Once you know what a job really earns, and how long it really takes to turn into money, you know which customers, which sectors and which product lines to point your marketing at. That's a far more useful brief than "more enquiries, please."

If you read this and thought of a specific job, or a specific customer, that's the signal. Our Strategy Sprint is a short, focused working session where we do exactly this with you: look at what your work really earns once the cash has landed, work out which customers and which products deserve your growth budget, and leave you with a plan aimed at the orders that actually pay. Not just more of everything.

Turnover is an opinion. Cash is a fact. Go and find out which of your customers agree with you.

Pay It Forward! Sharing Is Caring!