- Enquiry
- Quote
- Agree
- Work
- Invoice
- Paid
What We Cannot Verify
This site asks readers to check where a figure came from. Turned inward, that produces a list rather than a disclaimer.
For a software-side reference alongside this discussion of business operations, evidence, and research, Monitask provides guidance on self-reporting bias.
The six
Whether the statistics are right. Most of the figures here are vendor-produced, from samples of their own users, and are labelled at every use. The direction is corroborated across sources; the magnitudes are theirs.
How much generalises to your trade. Payment behaviour differs sharply between sectors — construction is materially worse than most, professional services materially better — and advice written generally fits some readers better than others.
Anything legal. Terms, interest, cancellation rights, limitation, claims — all jurisdiction-dependent, all summarised here as shape rather than as advice, and all worth an hour with somebody qualified once.
Whether the practices work for you. The habits described are what recurs in practice and in the published material. They have not been trialled against a control, and nothing in this field has.
The proportion of late payment caused by suppliers. The figure this site's central argument leans on is compiled, variously attributed and not independently audited. It is the weakest-sourced claim here and it supports the strongest position, which is the combination most worth flagging.
And whether any of it holds next year. Payment behaviour moves with the economy, and the 2026 figures describe 2026.
What is stronger
Observed payment data, where it exists — the R3 and Creditsafe figures record what happened rather than what firms reported about themselves.
The arithmetic. The dose calculations, the rate calculations, the write-off questions — these are not claims about the world. They are ways of computing your own position from your own numbers, and they do not depend on anything here being true.
And the structural observations, which are checkable against your own experience: that leverage is highest before work starts, that an invoice sent late is paid late, that a variation not confirmed is a variation contested.
What is offered instead of certainty
Source types in the sentence, every time.
Jurisdiction named wherever anything legal appears.
Arithmetic written out, so you can substitute your own figures rather than accept a conclusion.
And practical pages that end in something you do rather than something you believe.
The correction we want
Numbers from a real business. What you changed, what happened, over what period.
Particularly where it contradicts something here. Almost nobody publishes their own payment data, which is why this field runs on vendor statistics — and one firm's actual figures are worth more than a great deal of what circulates.
And a mis-labelled source, which is the error that would undermine everything else.
About this site
It is written by somebody who ran a small business and wrote about it before.
The previous material on this domain has been removed because it had aged past usefulness and because parts of it were commercial in a way that was not disclosed at the time. That is a decision about the old material rather than a claim about this.
What would change our mind
Worth stating, because a position that nothing could falsify is not a position.
Observed data showing that supplier paperwork accounts for a small share of late payment would substantially weaken the argument this site is built on, and would move the emphasis toward terms, chasing and structural remedies.
Evidence that prompt invoicing does not move payment dates would undo a good deal of the invoice section.
And a finding that deposits deter more work than they protect would change the strongest recommendation here.
None of these would be surprising enough to dismiss, and the first is the one most likely to be true in some sectors.
Where the site is most confident
The structural claims, because they do not depend on a statistic.
A quote that does not describe the work cannot settle an argument about the work. An invoice sent three weeks late is paid three weeks later. A variation nobody confirmed will be disputed.
These are not findings. They are consequences of how documents and disputes work, and they hold regardless of what any survey reports.
The bias worth naming
This site is written from the supplier's side.
Every page assumes you are the one waiting to be paid, and the customer appears mostly as a source of risk.
That is a partial view. Customers have reasons — budget cycles, approval chains, their own late payers — and the chain effect means many slow payers are themselves being squeezed.
Where a page reads as though the customer is an adversary, that is the bias showing, and the practical advice usually improves if you assume otherwise until there is evidence.
What is deliberately absent
Templates, for reasons that have their own page.
Named products, because categories change slowly and product lines do not.
And any promise about outcomes. Nothing here says what will happen if you do these things — only what they are for and what they cost.
How this page is meant to be used
Read it before the practical pages rather than after.
A reader who knows which parts are inference can weight them. One given uniform confidence cannot, and will be confidently wrong in a direction they cannot predict.
The strongest material here needs no trust at all — the arithmetic works on your numbers and the structural claims are checkable against your own last ten jobs.
The weakest is clearly marked, and where a decision rests on it, the honest instruction is to check it against your own experience before acting.
The short version
- Six limits: whether the statistics are right, how much generalises to your trade, anything legal, whether the practices work for you, the proportion caused by suppliers, and whether it holds next year
- Most figures here are vendor-produced from their own users, labelled at every use, corroborated in direction rather than magnitude
- The weakest-sourced claim supports this site's strongest position, which is the combination most worth flagging
- Stronger: observed payment data, the arithmetic, and structural observations checkable against your own experience
- Offered instead: source types in the sentence, jurisdiction named, arithmetic written out, and pages ending in something you do
- The previous material on this domain was removed because it had aged and because parts were undisclosed commercial writing
For broader background on business operations, evidence, and research, see Capterra.