A client last month wanted CMF settled after the mould drawings went out. Fine, I said, on one condition: nobody tells the tool shop yet.
10 years of form work here, 320+ products through to launch. When a shape comes back for a second pass, the cause is rarely that it looked bad. It is that the three elements were signed off in the wrong order.
The three are volume and proportion, human scale, and CMF — colour, material, finish. Not a menu you pick from. Volume sets the skeleton, human scale sets the working face, CMF sets how the thing finally looks. Skip ahead and you do not save a step. You buy it back later, at a price set by whoever owns the steel.
The Three Elements, and What Each One Owns
Volume and proportion covers the length-width-height relationship, main wall against rib thickness, draft angle, the fillet family. It is the element that shapes everything downstream, and the one that cannot be recovered once a tool is cut.
The test we use is plain and costs nothing: fill the silhouette solid black and show it to someone who already owns one. Can they name the category? If they hesitate, the proportion is wrong, and no amount of surface work will rescue it.
A case from our files. A power bank shell on a main wall of 2.0 mm. In the first pass the main wall and the ribs came out close in thickness — nearly the same number, which looks safe on a drawing and is not. At T0 the outer face showed sink marks. The fix was not in the tool: we pressed the rib thickness down to 0.5–0.6 times the main wall and the part ran clean.
Plastic part tolerances get looked up, not felt. We read them off GB/T 14486 and pick the grade by function. A hand-feel number on a mould drawing becomes an argument at T1.
Human scale covers grip diameter, key travel, finger recess depth, screen tilt — the geometry the body meets. It has its own deadline, below.
CMF covers colour, material and finish. It decides how expensive the product reads on a shelf, and it is the element the budget reaches for first.
Human Scale Has to Be Locked Before Structure Starts
Human scale is where the product stops being a shape. Grip diameter, key travel, how deep the finger recess sits, the angle the screen leans at. None of it shows in a render. All of it shows in a hand.
A blood pressure monitor from a few years back. The battery compartment cover went through three revisions. The client kept sending it back as a looks problem. It was not. People could not get a fingernail under the edge — the finger recess on the cover was too shallow, and on an older hand the nail slid straight off. We cut the recess deeper and the complaint stopped.
That one carries a rule. Human scale must be settled before structural design begins. Structure allocates space — ribs, bosses, battery format — and every allocation is a hard boundary once it exists. Hand it a grip diameter afterwards and it can only shave wall to find the millimetres, which is how you get a part that fails a drop test in a way nobody predicted.
CMF Runs Last, and That Is Exactly the Risk
In the same batch, ΔE ≤1.5 is the line where the eye stops telling two parts apart. Above it, a customer holding the unit sees two colours on one product with no instrument involved.
Texture changes the tool, not just the look. A grained face needs more draft than a polished one — commonly 1.5° or more — or the side wall drags on ejection and the grain shows a smear exactly where the thumb travels. Ask for texture after the draft is set and structure either moves the parting line or accepts the drag marks.
So CMF has one hard deadline: no later than the release of the mould drawings. Not because colour needs the time. Because the draft and the parting line do.
What Gets Cut First (Always CMF)
When the budget tightens, the cut lands on CMF every time. The substitutions are standard:
- Double-shot moulding becomes paint plus pad print.
- Metal-look finishes become a no-paint moulded material.
- Electroplated parts become brushed film.
Each line saves real money on its own. Stacked together they cost the product a full tier of perceived quality. That is not a design opinion; it is what the sample shows when you set it next to the version you signed off.
The problem is not the substitution. It is that nobody goes back to check whether the volume and proportion still hold. A parting line that double-shot moulding can absorb may not be containable under paint. Weld lines that vanish on a light part are obvious on a dark one, and no-paint material makes them worse, not better.
Cut the CMF budget and you pay it back in volume: tuck the parting line into a fillet, take a dark colour down to a light one, simplify the split.
One test settles it. Any CMF substitution gets rechecked on the sample at three places: the parting line, the weld line, the colour difference. All three pass, then talk about volume. That is worth more than the number on the quote sheet.
When Each One Must Be Signed Off
The whole article in one table.
| Item | Volume and proportion | Human scale | CMF |
|---|---|---|---|
| Sign-off point | sketch and first 3D | before structural design starts | before mould drawings are released |
| Cost of change | proposal redrawn from scratch | affected structure parts revised | mould surface treatment or a material change |
| Who decides | product owner | structure and design together | brand side and supply chain together |
| Can it be added later | no | small tweaks yes, big changes no | yes before the tool, sharply more after |
Where small and mid-size companies go wrong: the concept goes to the factory first, to ask whether it can be moulded, and only then back to a designer. Structure-first feasibility locks the volume and proportion before design has had a say, and what is left for form is surface decoration on a proportion nobody chose.
The fix is a scheduling one, and it costs nothing. Put the volume proposal and the structural feasibility review in the same meeting. One room, one afternoon, both answers on the table before either gets frozen. That single change removes most of the rework above.
Budget and Price Bands
In RMB, the bands here are stable: appearance only 5,000–12,000; appearance plus structure 15,000–40,000; full project 30,000–80,000.
The three elements mostly live in the first band. Volume and proportion, human scale, CMF, done properly, is appearance work. But the later they get settled, the more likely the job slides into the second band — and the second band is what pays for structure chasing decisions that should already have been made.
Where to spend inside the budget: leave enough review rounds for the proportion and the human scale stages, and leave physical CMF samples in the line. Money saved on review rounds comes back later as tool modification. It always does, and it comes back larger.
Two numbers from our side: 200+ enterprise clients, a 65% repeat rate. The ones who come back are mostly the ones who put the review points in the right order. Not the ones with the biggest budgets.
Sources: Compiled from Hezi Industrial Design's own project experience; no external citations.
Article by Ou Ruolan / Hezi Industrial Design — Dongguan, China. Originally published at hezidesign.com.
Top comments (0)