After every cut there are usable pieces left beside the bench. Some of them really are waste; others are the material for the next job. What separates the two is not only the size of the piece but whether it was recorded at all. Glass that is not written down with its dimensions does not exist in inventory even while it sits on the rack: nobody looks for it, and nobody places an order relying on it. Remnant management looks like a habit of tidiness, but it sits right in the middle of the decision to buy material.
An offcut rarely disappears in a single moment. The cut ends, the piece is set aside, the day ends. The next day another job arrives and the piece is no longer on anyone's mind. A week later it is hidden under newer sheets on the rack. In the end it is either thrown out to make space, or a new sheet is opened because a piece close to the required size could not be found. In both cases the workshop has paid twice for material it already owned and still owns.
Counting an offcut as stock starts with writing down its dimensions: width, height, glass type and where it is kept. Those four pieces of information make it searchable on the next order. The moment it becomes searchable the order of decisions changes; available remnants are tried before a new sheet is opened, and the sheet is only reached for if they do not suffice. This two pass approach solves the same order against two different pools of material and gives priority to what has already been paid for.
In the workshop this shows up as the morning's decisions being visible in the evening. Which sheet was opened for which order, where each remnant went, and whether it was used on the following job, all stay on record. Keeping records looks like extra work; compared with the time spent walking between racks looking for a piece, it becomes clear which one costs more. There is also this: where nothing is recorded, nobody can prove the waste they noticed with their own eyes.
It is worth saying that remnant use has limits. Not every piece fits every job; edge condition, scratches, trim allowance and glass type mismatch are real obstacles. So what the system does is not to force remnants into use but to remind you of the suitable ones. If none is suitable a new sheet is opened, and that is not treated as a failure. The aim is not to consume a remnant on every order but to avoid forgetting what is already there; material that is not forgotten finds its own place in the queue.
Where to start looks simple and it really is: record today's usable pieces with their dimensions. Looking at that list a week later shows how many were used, and that number tells you whether the habit is paying off. The role of software is to keep the list and put it in front of you at the moment of ordering; the decision still belongs to the workshop. But a decision made with material you know you have is a different decision altogether.