A spreadsheet will do for longer than the people selling software would like, and the thing that decides whether you need more is not your size. It is how many people have to touch the same record on the same day.
Written by JD Davenport, who still sells for a living. Updated 4 November 2026.
Software stores job cost. It does not create it. If your costs arrive two weeks late, land in the wrong column, get coded to the wrong job or never get a job number at all, a system will hold all of that faithfully and present it on a dashboard.
This is the expensive misunderstanding. A business that cannot state a job's true cost buys software expecting the answer, and receives a better-looking version of the same gap. The discipline has to exist first. Software is a way of scaling a discipline you already have, and it is very poor at manufacturing one.
So the honest sequence is: get the costs landing correctly on paper or in a spreadsheet, prove you can close a job inside a period you care about, and then decide whether a system is worth paying for. You will also be a much better buyer at that point, because you will know what you are asking it to do.
Four things, and they are worth naming precisely rather than hand-waving at scale.
Two people editing the same row at the same time. Shared sheets handle this badly and the failure is silent. If your office has more than one person entering costs on the same day, this is the first real wall.
Permissions. A spreadsheet is all-or-nothing. If a crew leader needs to enter what happened on a job but must not see everyone's pay, a spreadsheet cannot do that and a system can.
An audit trail. Who changed that number, when, and what it was before. Version history is not the same thing, and the day a commission dispute turns on it you will want the real one.
Volume. Not the number of jobs, the number of cost lines. A few hundred rows a month is a spreadsheet. A few thousand is not, and the tell is that recalculation gets slow and people start keeping side copies.
Notice that none of those four is about how much money you make. A one-truck business with a bookkeeper and an owner both typing hits the first wall earlier than a larger business where one person owns the file.
It shows its working. Every calculation is visible and changeable, which matters enormously when somebody disputes a number, because you can walk them through it in the same room in two minutes. A system that produces a figure from rules nobody can see produces arguments nobody can settle.
It also costs nothing, expires on nobody's schedule, and cannot be taken away from you. That is worth more than it sounds the first time a renewal price changes.
Ask what breaks first if you carry on as you are. If the answer is "the file gets slow" or "two of us overwrite each other" or "I cannot let that person see this tab", you have hit a real wall and software is the answer to a real question.
If the answer is "we still would not know what a job cost", the wall is not the tool. Buying a system will move the same problem into a more expensive container, and you will spend the implementation quarter discovering that the data going in is the part that was broken.
The spreadsheet version, with the twelve tests, the register and the rate card, is The Job Cost Standard. It is a file you buy once, and it works in Google Sheets or Excel. More answers on the questions page.