When a client says "let me look at it again", the proposal usually has a problem with information density, not with quality. The client is not a designer. He cannot read your intent out of a pile of surface curves. The most effective method our team has used is to cut the proposal into three pages: page one covers form and proportion only, page two covers surface detail only, page three covers CMF and context only. Each page answers exactly one question, and the client can say yes or no within five minutes.
The short version: one page, one question. Page one: three-view or isometric, annotated with length-to-height ratio, thickness and the position of the main feature lines. Page two: light studies plus the key surface numbers — curvature, fillet, texture depth. Page three: colour options, real material samples and the product in context. If the client hesitates, the page has not given him anything to judge against.
Page one: form and proportion, so the client accepts the silhouette first
This page shows the silhouette and the key proportions, nothing else. Use a three-view or an isometric view, and annotate the length-to-width ratio, the thickness and where the main feature lines sit. On a handheld medical device we designed, page one held three images only: front silhouette, side thickness, top view.
Critical dimensions go straight onto the drawing as arrows — width to height came out at 1:1.6, and thickness was held at 22 mm. The client is judging proportion, not the light falling across a surface.
A mistake we paid for: we used to put form and surface on the same page. The client's attention drifted to the highlights, and he ended up with no opinion about the silhouette at all. After we split them, one client looked for three seconds and said "this proportion is wrong, it's too fat." We adjusted the ratio on the spot and fixed it in five minutes. Had the surface detail been on that page, he would probably have noticed it the next day.
The test for this page is direct: can the client say "I accept this form" or "this proportion needs to change" within ten seconds? If he hesitates, the silhouette itself is the problem — do not rush into surfaces. In our experience, once the silhouette is confirmed, the probability of later changes to surface and CMF drops by more than half.
Page two: surface and detail, argued with light and numbers
Page two covers surface detail and the feel of the product. Use high-resolution renders or light studies to show how the surface runs, what the fillets are, and which way the texture goes. Then annotate the numbers — the main surface curvature radius at R120, the edge fillet at R3, the grip texture depth at 0.3 mm.
On a blood-pressure monitor we put four light studies on this page: front light, side light, a close-up of the grip area, and a diagram of a finger on the grip.
The recurring failure here is not annotating the parameters and leaving the client with nothing but a feeling. One client said "this surface doesn't look smooth enough." We re-rendered the light study three times before discovering that the camera angle was the problem — the surface hadn't changed at all. Once key parameters are printed on the page, the client has something to hold on to. After seeing R120 he says "can that radius be R100? It would sit better in the hand."
The test for this page: can the client read the surface from the light study and respond in parameters. If he says "this area feels a bit flat", you know the radius needs work. If he says "change R120 to R100", he has understood the parameter — the proposal has moved into detail adjustment.
Page three: CMF and context, so the client sees the end result
Page three shows colour, texture, gloss and context. Use real material samples or high-fidelity renders, and show the colour options, the direction of the texture and the difference in gloss. The context shots show the product where it will actually live: on a desk, in a hand, next to other products. On a home monitoring device we showed three colour options — white, grey, blue — each with a context render and the matching material sample photographed.
A mistake we paid for: samples that did not match the render. A client said "the grey in the render looks blue, but the sample is pure grey." The render environment was lit too cool. We now require the CMF page to use the same material sample as the render, shot under a standard D65 light source. When the sample and the render agree, trust improves noticeably — the colour difference can be held within ΔE 1.5.
The test for this page: can the client imagine the product in his own environment from the context shots. If he asks "will this colour be too loud in an office", you know saturation or gloss needs adjusting. If he says "the blue, with a matte texture", the CMF decision is made.
What each page should and should not do
The number of pages can flex with project complexity — a simple project may only need two, form plus CMF. A complex project can add a page between surface and CMF explaining how internal structure limits the surface. What must not flex is the rule: one page, one decision.
Ask a specific question at the end of each page. Not "what do you think", but "is this proportion acceptable" and, if the answer is "needs adjustment", "too wide or too thick". The client cannot answer an open question quickly; he can answer a closed one in five seconds. When a client says "let me look at it again", the most likely cause is that the page asked him to make three decisions at once without giving him a basis for any of them.
FAQ
Does the three-page structure work for every product? It is built for appearance decisions. If internal component layout constrains the exterior — a handheld device where the battery defines the thickness, for example — add a page after surface detail showing how the structure limits the form. Keep the appearance section at three pages.
What if the client asks for a surface change on page two? First ask whether it affects the proportions confirmed on page one. A local change such as a larger fillet radius does not, so annotate the new parameters on the render directly. If it changes the overall surface flow, go back to page one and reconfirm the silhouette before continuing — otherwise the pages stop agreeing with each other.
How do you keep CMF samples consistent with renders? Same material, same process, same colour formula, shot under D65 at 6500 K, then tune the render against the photograph. If the client specifies a colour, confirm the formula against a RAL or Pantone chip first, then make the sample and the render. Colour difference within ΔE 1.5 is achievable this way.
Written by the Hezi Industrial Design team in Dongguan. We do product appearance design, structural design and CMF as one service, with 320+ products in mass production.
Top comments (0)