Manual sheet planning vs an optimizer: a side-by-side test

Every shop has an opinion about whether cut list software is worth the five minutes it takes to type in a part list. So we ran the same job twice: once with graph paper, a scale rule, and a pencil, and once through a browser-based optimizer. Same 22 parts, same sheet stock, same person, same day. The differences were bigger than we expected, and not just in speed.

The job: a 22-part kitchen run

The test piece was a small run of two base cabinets and one wall cabinet in 18 mm melamine-faced chipboard, sheet size 2440 x 1220 mm (8 x 4 ft). The list had 22 parts ranging from 250 x 400 mm shelves to 720 x 560 mm sides, all with grain running the long dimension on the visible faces. Kerf was set at 3.2 mm for a table saw with a scoring blade. Nothing exotic, this is a completely typical small-shop job.

Method: same inputs, two processes

For the manual pass, one of us used 5 mm graph paper at a 1:10 scale, drew the 2440 x 1220 mm sheet outline, and packed the 22 rectangles by eye, redrawing twice when a part didn't fit. For the optimizer pass, the same part list was typed into the free cut list optimizer with sheet, thickness, kerf, and quantities, and the tool returned a guillotine-cut layout in under two seconds. We timed both from a blank sheet of paper / blank browser tab to a layout the operator was willing to cut from.

Time spent

The gap widens with part count. On a 22-part job the manual planner is still tracking everything in working memory; on a 60-part job, that same planner starts making the errors we describe below, because graph paper doesn't check your work for you.

Sheets used and waste

The manual layout used 3 full sheets at an estimated 71% yield, once we measured the drawn rectangles against the actual part areas. The optimizer's guillotine layout used 2 sheets at 84% yield for the same 22 parts, because it packed the shelves into offcut space next to the cabinet sides that the manual drawing had left empty. At roughly €48 per sheet of 18 mm melamine-faced board, that's one fewer sheet, about €48 saved on this job alone, before counting the labor difference.

Mistakes made

The manual pass had three errors that surfaced only when we double-checked against the original part list: one shelf was drawn 10 mm short because the scale rule was misread, one part had its grain arrow drawn parallel to the panel's long edge when the design called for it across, and the kerf gap between two adjacent parts was omitted in one spot, which would have produced an oversized first part and an undersized second one after the blade removed material. None of these are dumb mistakes, they're exactly the kind of arithmetic and attention slips that happen after 40 minutes of squinting at a scale drawing.

The optimizer made none of these errors, because it works from the numbers you typed, not from a hand-drawn approximation. Its only failure mode is a wrong input: if you type 620 instead of 720 for a part length, it will confidently produce a wrong-but-consistent layout. The tool doesn't know your design intent, it just doesn't introduce new errors while doing the arithmetic.

Where manual planning still holds up

• One-off jobs of 3-5 parts where a full sheet layout is unnecessary and a tape measure against the offcut pile is faster than opening a laptop.
• Highly irregular parts, curved edges, or angled cuts that guillotine-style optimizers don't model well.
• Repairs and site measurements where you're cutting one part to fit an as-built opening, not optimizing a batch.
• Teaching contexts, where working through the packing problem by hand builds intuition about grain, kerf, and waste that's worth having even after you switch to software.

Where the optimizer clearly wins

• Any job over roughly 10-12 parts, where working memory starts to fail silently.
• Jobs with mixed part sizes where offcut reuse matters, since software checks combinations a human won't bother trying.
• Repeat jobs, where you want a saved project and a consistent, printable cut sequence for whoever runs the saw.
• Jobs billed by material, where a 10-15 percentage point yield improvement is real money, not a rounding error.

What the optimizer doesn't replace

Software won't check whether your sheet stock actually arrived at 2440 x 1220 mm or a hair under, and it won't catch a design error like a shelf that's 10 mm too short for its cabinet, because that's not a cutting problem, it's a design problem. You still need to verify your part list against the actual cabinet drawings before you type it in, and you still need to check your kerf setting against the blade you're actually using, since a wrong kerf value produces confidently wrong output just as fast as a right one produces good output.

A hybrid workflow that works

1. Build the part list from your drawings, with dimensions, quantities, thickness, and grain direction for each part.
2. Spot-check 2-3 parts by hand against the drawing before typing anything in, catching design errors early.
3. Run the list through the optimizer for the layout, sheet count, and cut sequence.
4. Print the layout and read through the numbered sequence once before cutting, checking it against the part list, not against your memory.
5. Cut the first two or three parts, measure them, and confirm before running the rest of the sheet.

Bottom line

On this 22-part job, manual planning took roughly 6.5 times longer, used 50% more sheets, and produced three real errors that would have caused rework or scrap on the shop floor. Graph paper is fine for a handful of parts or a one-off repair, but past about 10 parts, the time and material savings from a proper optimizer are hard to argue with, and the error rate drops close to zero because the arithmetic is no longer done by a tired human at the end of a shift.

Next steps. Try your own part list in the cut list optimizer and compare the sheet count against a rough hand estimate, or read our guide to sheet yield and nesting for more on how packing order affects waste.