A builder rings about one line on your quote. They think the plasterboard allowance is high, and they want to know how you arrived at it. The number might be right. That is not what is being tested. What is being tested is whether you can show, on the spot, how it was produced, what it was produced from, and whether anything about it moved since the last revision you sent.
Plenty of software can produce a number. Explaining one is a different job, and that gap is where credibility leaks away. A figure you cannot account for gets treated as a guess, and once one line reads as a guess, the whole document becomes negotiable.
An unexplained line is not a price. It is an opening offer.
What a defensible line actually carries
Defensible does not mean confident. It means the line carries its own history, so answering a challenge is reading a record rather than reconstructing a memory. That history is not complicated, but it has to be captured while the work happens, not assembled afterwards.
- The rule that produced the quantity, and the geometry it used
- The plan revision and the confirmed measurements that geometry came from
- The supplier SKU, pack size and rate behind the cost
- Any manual change to the line, with the person who made it and the reason
- The version the line sits in, and who approved that version
Provenance: which rule produced this number
Provenance is the difference between "the software worked it out" and "that is the confirmed wall area from revision 4, run through the sheet coverage rule with the waste factor set for this job, then rounded up to whole packs". The second answer ends the conversation. The first invites another one.
It matters most where a quantity looks wrong and is right: an offcut allowance on rooms full of short runs, a hardware count driven by the opening schedule rather than by wall length. A rule you can show is a rule the builder can argue with on the facts. If a quantity was generated rather than calculated, there is nothing to show, only a plausible-looking figure.
Versioning: what changed between revision 3 and revision 4
Every job of any size collects revisions. The plan updates, a client changes a finish, a supplier discontinues a product. The trouble starts when each revision is a new file rather than a new version of the same thing. Two people end up holding different documents with the same name, and nobody can say which lines actually moved.
A versioned bill of materials makes that answerable. You can put revision 3 next to revision 4 and see which lines changed, which are new, which came out, and what the movement does to the total. It is also what makes approval mean something. Approving revision 4 is a decision about a specific set of lines, not a general nod at a PDF.
Overrides: the line someone changed by hand
Overrides are not a failure of the system. They are the reason a person sits in the middle of it. Someone who knows the site will sometimes correct what a rule cannot see: restricted access, a client-supplied item, a substitution agreed on the phone.
The damage comes from the silent override. Weeks later nobody remembers the figure was changed by hand, so it gets defended as though the rule produced it, or the rule gets blamed for a number it never produced. An override that records the original value, the new one, who changed it and why keeps both facts intact: what the rule calculated, and what judgement was applied on top.
Back-costing: what the job actually cost
Once the job is built, the actuals exist. What was ordered, what was delivered, what came back, what had to be ordered twice. Comparing that against the quoted lines turns a finished job into something the next quote can use.
Variance is not automatically a mistake. A line that ran over might point to a waste factor set too low for that product, a pack size that forces round-ups, or a rule that assumes a cutting pattern nobody follows on site. Each of those is a fix you make once, in the rule, and every quote after it is better. Without back-costing there is nothing to fix, so the fallback is to quote a little higher and hope.
- Which lines moved against the quoted figure, and by how much
- Whether the movement traces to a rule, a rate or a decision made on site
- Whether the waste factor for that product reflects how it is actually cut
- Which supplier rates have drifted since the catalogue was last reviewed
Defensibility is a habit, not a feature
None of this asks you to quote differently. It asks that the record be kept as the quote is built, rather than assembled afterwards from memory and a folder of files, which is where the hours go and where the confidence drains out of the answer.
The builder who challenged the plasterboard line may still disagree after you show them the working. That is a better argument to be having. You are discussing a waste factor or a plan revision, both of which have an answer, instead of discussing whether your quote can be trusted at all.