Counting materials and forgetting everything else
The single most common error. Materials are the visible cost, so they get counted; wages, overhead, transport and the exchange fee do not.
The fix: use a calculator that includes all of them, and be suspicious of any profit figure that seems too good.
Assuming self-production is cheaper
It often is not, because the building doing the producing could have been earning more doing something else.
The fix: always ask what that building would otherwise be making. If the answer is "something better", buying the input is probably right.
Pricing inputs at the cheapest listing
The cheapest listing is frequently a small lot. Buying at scale means walking up the order book and paying more.
The fix: check the depth. If you need 10,000 units and the cheapest offer covers 200, price your plan on what you would actually pay.
Ignoring administration overhead as you grow
Overhead multiplies your wage bill and rises with your company. A production line that cleared 15% margin at ten buildings can be loss-making at forty, with nothing about that line having changed.
The fix: re-check your best lines whenever you expand. The calculators here take your current overhead as an input for exactly this reason.
Building before calculating
A building is a large, slow commitment. Constructing one to find out whether it pays is an expensive way to run the numbers.
The fix: work out the payback period first. If it is longer than you are comfortable betting on prices staying stable, it is not the right build.
Chasing whatever spiked today
The most profitable-looking product is very often the one whose price just jumped, and prices that jump tend to come back down.
The fix: look at the price history and the volatility before committing. Our opportunity scanner flags recent spikes for this reason.
Letting buildings sit idle
An idle building still pays wages. It is not neutral, it is a steady loss.
The fix: if a line is not worth running, switch it to something that is, rather than leaving it stopped.