A client forwarded me two quotes for the same small kitchen appliance last month. Same product, same three views, same glossy PDF full of renders that would survive a pitch deck. One studio wanted RMB 5,000. Another wanted RMB 50,000. He asked me which one was robbing him.
Neither. I have quoted both numbers myself, out of the same office in Dongguan, and the gap is not drawing speed. It is not who renders better either. It is the delivery boundary: how many lines of deliverables the quote puts in writing, and who is still answering the phone when the tool shop calls at eleven at night.
For RMB 5,000 you buy a few renderings good enough to report upward, plus a rough shape. For RMB 50,000 you buy a line from brand tone deduction through Class-A surfaces, a CMF specification book, and handboard-to-tooling follow-up, with someone accountable when it goes wrong.
Three Price Bands, and What Each One Sells
In RMB, the market here settles into three bands:
- Appearance design only: RMB 5,000–12,000. Shape renderings plus preliminary 3D. Revision rounds: 1–2.
- Appearance plus structure: RMB 15,000–40,000. Appearance work, structural coordination, mould-ready surface data, CMF recommendations.
- Full project: RMB 30,000–80,000. Adds brand tone deduction, multiple handboard validation rounds, a CMF specification book, and tooling follow-up.
The extra money buys three things: deduction rounds, surface quality grade, and scope of responsibility. Nothing else. A studio charging RMB 50,000 that skips the deduction rounds is not expensive, it is badly scoped. A studio charging RMB 5,000 that promises a mould-ready model is not cheap, it is lying about scope.
Curvature: You Paid for G2 and Got G0
The cheapest place to hide a shortcut is a fillet, because renders forgive it.
The low quote fillets both sides of the parting line at R3 and calls it done. That is positional continuity only — G0, two surfaces meeting at a line. Under soft studio light the seam disappears and the director approves the image. Then the mould shop polishes the joint, the highlight band breaks, and the part looks assembled from two pieces of different tooling.
The high quote works a variable-radius transition with curvature continuity — G2. Run a zebra stripe test and the reflected stripes cross the boundary without kinking. That is the physical difference, measurable on a screen, not a matter of taste.
A case from our files: a handheld product, first version with plain side chamfers. The client loved the renders. The handboard came back with the side highlight band broken and the unit reading as two parts glued together. We rebuilt the side as a variable-radius surface. Modelling cost two extra days; at the tool shop it removed one polishing step. The client paid twice — once for each version.
Every quote sits in one of three grades of surface delivery:
- Viewable — good enough to render and present.
- Usable — structure can attach to it and survive.
- Mould-ready — the tool shop can cut steel from it without rebuilding anything.
The RMB 5,000 band almost always stops at the first grade. Sometimes that is all the client needs, and it is legitimate. It becomes a problem when he believes he bought the third.
CMF: One Colour Chart vs a Stack of Chips
For RMB 5,000 you get a colour picture — a rendering with a material description under it. For RMB 50,000 you get a stack of physical chips you can hold against a sample in the light of the actual shop.
Texture has a public standard that most buyers never hear about. VDI 3400 numbers texture by depth, and the same number on different steel, etched by a different shop, does not feel the same in the hand. Both VDI 24 and VDI 27 count as matte. One feels like fine sandpaper, the other like coarse felt — on a surface your thumb crosses daily, that is what the customer comments on.
Colour is worse. The cream white you pick from a Pantone book comes out of the mould slightly yellow or slightly grey, because pigment, melt temperature and cooling rate all push it. The industry holds ΔE at 1–2 on a matched part. Above 3, the eye reads two components on one unit as two different colours with no instrument.
Getting there is a process: shoot a chip, cure it, check it under daylight and under shop lighting, adjust the masterbatch, shoot again. One to two weeks per round. A low quote has no line item for this, so it does not happen. The sample arrives, the client says "why doesn't it match the picture", and the rework costs more than the design fee he saved. He remembers the cheaper quote fondly right up until the T1 sample.
Hand Feel Does Not Render
A screen model cannot tell you whether the product slips in a wet hand. Neither can a cheap handboard.
The SLA resin handboard everyone uses for form checks is the wrong tool for feel. Its density, friction and thermal mass have nothing in common with a real moulded part. It tells you the shape, nothing about grip.
To judge grip you need a CNC handboard with paint or bead blast, so the surface friction is at least in the right family — and you test it on hands, not a desk.
Adult male hand width is mostly 78–88 mm. On a cylindrical handle, two locations decide whether it feels right: where the palm makes contact, and where the web between thumb and index takes the load. Those set the diameter, and the diameter that suits one hand is uncomfortable for another. On the handboard, have different hand sizes each hold it for five minutes, mark the pressure points with a pencil, then change the geometry. One handboard plus one revision round is about two weeks. That line item exists in a full-project quote, not in a concept quote, and it decides whether the product feels expensive.
The Comparison Table
| Dimension | Appearance RMB 5,000–12,000 | Appearance + structure RMB 15,000–40,000 | Full project RMB 30,000–80,000 |
|---|---|---|---|
| Form deduction rounds | 1–2 direction proposals | about 3, including form convergence | multiple rounds, includes brand tone work |
| Surface quality delivered | render-ready | mould-ready data | mould-ready data + highlight band check |
| CMF delivered | colour suggestions | material and process suggestions | CMF spec book + sample follow-up |
| Handboard validation | not included, quoted separately | can be added on request | appearance handboard + grip test included |
| Tooling and follow-up | not included | structure interface, follow-up assessed separately | follow-up and first-article approval included |
| Best fit | concept validation, internal sign-off | first mass production of an own product | retail channel, brand consistency required |
Cutting the Price Without Cutting the Deliverable
Some clients should not buy the full project, and saying so saves a year of resentment. If you only do small-batch custom runs, if the appearance follows an existing public shell, if you cannot get two products to market in a year — the RMB 5,000 band is enough. Put the rest into packaging and product-page photography, where it moves units.
A quote is not right or wrong by itself. The mistake is paying concept-validation money and expecting mass-production delivery, or paying full-project fees for a public shell.
When you negotiate, do not cut the total. Cut the scope. Push form deduction down to 2 rounds. Take the handboard in-house. Hand tooling follow-up to the tool shop. The price drops and every boundary stays written down. Halving the number outright does not halve the work — it makes the other side quietly remove the surface quality and the sampling steps, and you find out at T1.
Before you talk price at all, write the deliverables out line by line and ask each supplier to quote line by line. With those lists side by side, the gap between RMB 5,000 and RMB 50,000 stops being a mystery and becomes a shopping decision.
Sources: Compiled from Hezi Industrial Design's own project files, quote sheets and tooling records in Dongguan; no external citations.
Article by Liu Ye / Hezi Industrial Design — Dongguan, China. Originally published at hezidesign.com.
Top comments (0)