Why your quoting spreadsheet eventually lies to you

Nobody sets out to run a fabrication business on a spreadsheet. It happens because the first one works brilliantly. Somebody sits down, writes the arithmetic for a work table, and for about a year it is the best thing in the shop.

The failure is not that the spreadsheet is bad. It is that a spreadsheet is a document, and the thing you need is a system, and the difference does not show up until you have several hundred documents.

The copy is the problem

Here is how every estimating spreadsheet dies, in the same four steps.

Someone copies it for a new job. Of course they do — that is what it is for, and starting from a similar job is faster and safer than starting from nothing.

They tweak it for that job. The customer wants a deeper table, so a cell changes. The frame is 202 to hit a price, so a rate is overtyped. Perfectly sensible edits.

That copy becomes the base for the next one. Because it is the most recent, and it was for a similar customer.

Nobody knows which file is canonical. Six months in there are forty files, several of them partially edited descendants of each other, and the one your estimator opened this morning has a hardcoded 202 rate in a cell that is supposed to be 304.

The spreadsheet is not wrong. It is doing exactly what it was told. It is just that what it was told was decided in a hurry, on a different job, by someone who is not in the room.

The three specific ways it gives a wrong answer

Stale rates, confidently applied. A quotation built on last quarter’s sheet price is out by whatever steel did since. Material is around nine tenths of a fabricated table’s cost, so a 10 per cent move in nickel is a 9 per cent move in your cost and — if you did not notice — a 9 per cent hole in your margin. There is no warning. The file calculates cleanly and returns a number that looks exactly like a right number.

Assumptions that live in a person, not the file. Ask why the top is cut four inches over on each dimension and someone will tell you it is the bend allowance. Ask why a tray-type top carries an extra 25 per cent of labour and someone will tell you it is the pressing. Now ask the estimator who joined in March. The formula is in the cell; the reason is in a head, and heads take annual leave and take other jobs.

Margin taken on the wrong base. Twenty per cent margin on selling and twenty per cent markup on cost are different numbers — ₹23,119 against ₹22,194 on an ₹18,495 table. Multiplying is easier than dividing, so spreadsheets tend to do the markup and call it margin. It is small per line, systematic across every line, and invisible for years.

The test that settles it

Not an opinion, a measurement. Take one job you quoted eighteen months ago and requote it today, from the current file, with current rates, and compare it to what you actually sent.

Two possible outcomes.

The numbers match, allowing for genuine rate movement — your process is sound and this article is not about you.

Or they do not, and the gap is not explained by rate movement. That gap is what your quoting is costing you, and it has been doing so on every job in between.

Most shops that run this test find something. It is usually one of the five below.

Where the gaps are, in order

1. Quoting the finished size, not the blank. A 72 by 30 top cut at 76 by 34 is 20 per cent more sheet. Easy to fix once, easy to lose on a copied file where someone “cleaned up” the allowance.
2. A rate that moved and a cell that did not.
3. Margin on the wrong base.
4. The small parts. Undershelves, splashbacks, extra tiers — quoted as small because they are small, when an undershelf is nearly two thirds of a top.
5. Gauge assumed rather than specified. 18 SWG where 16 was quoted, or the reverse.

What actually fixes it

Not discipline, and not a better spreadsheet. Both of those have been tried in every shop that has this problem.

What fixes it is separating the three things a spreadsheet mashes together:

The rates. One list, one place, one edit when steel moves. Not a number typed into forty files.

The logic. The arithmetic, written down once, with its reasoning attached — this allowance is for the bend, this multiplier is for the pressing — so it survives the person who wrote it.

The job. The dimensions, the grade, the options, the quantity. This is the only part that genuinely changes per quotation, and it is the only part anybody should be typing.

That separation is the whole design of Sazinga Quote, and it is what a metal fabrication shop is actually buying when it moves off the estimating folder. Do it and copying a file stops being a thing that can happen, because there is nothing to copy — there is one set of rates, one set of formulas, and a job that references both.

Is a spreadsheet ever still the right answer?

Yes, and it is worth saying plainly. If you send four or five quotations a month, if one person does all of them, and if what you make is genuinely varied enough that no two jobs share arithmetic, a spreadsheet is fine and a system is overhead you do not need.

The point at which it stops being fine is fairly sharp, and it is not really about volume. It is when more than one person quotes, or when the same product goes out at different sizes over and over. The first means the assumptions have to leave somebody’s head. The second means you are retyping arithmetic that has not changed, several times a week, and every retype is a chance to be wrong.

Both of those are structural. Neither is fixed by being more careful.