Design for manufacture, before tooling.
Also called: DFM review · Manufacturability audit · Producibility review · Tooling-readiness check
Walking the design through manufacturing eyes before tooling, so the decisions that cost ten times more to reverse get caught now.
DFM is a structured review of the design through a maker’s eyes: draft angles, tolerances, assembly access, supplier capability, tooling cost. Run it before the mould is cut or the board is fabricated. A change on a drawing is cheap; the same change in hard steel is not.
What design for manufacture is
Design for manufacture is the review where you stop asking “does it work?” and start asking “can it actually be made, at this volume, by these suppliers, at this cost?”. It is a deliberate pass over the design before any tooling is committed, looking for the features that are fine on a screen and expensive in a factory. The discipline is to think like the person cutting the steel or fabricating the board, not the person who drew the part.
The reason it matters is timing. A draft angle changed on a CAD model costs minutes. The same change once the mould is cut means re-machining hardened tool steel, or scrapping the tool and starting again. That is the ten-times rule in practice: decisions get cheaper to reverse the earlier you make them, and brutally expensive the moment something physical is committed. DFM is the structured chance to make those decisions while they are still cheap. I treat it as the last gate before money turns into metal.
What it actually checks
A real DFM review is not a vibe; it is a checklist against the specific design. For a moulded ceramic body and a populated circuit board, it walks five things: draft angles, so parts release cleanly from the mould; tolerance stacks, so mating parts actually mate; assembly access, so a human or a fixture can reach every joint and fastener; supplier capability, so the wall thicknesses and finishes you have drawn are ones your chosen supplier can hold; and tooling cost set against the run size, so you do not pay for a tool the volume cannot justify. Miss one and you usually find out at the worst possible moment, on the factory floor.
The DFM review, worked
The clearest way to see what a review covers is to lay it out against a real product. Here is the proofing box’s, the £149 box we ran through to launch, so you can see the shape of a real review rather than a generic template.
Notice that none of these are about whether the box works. It already worked on the bench. Every row is about whether the working design survives contact with a factory, at the price and volume the business can actually carry.
- A working prototype sent to a toolmaker without a review.
- No draft on the ceramic, so parts drag and chip on release.
- Tolerances assumed, never stacked, so the lid does not seal.
- Changes discovered in hard steel, at toolmaker rates.
- Draft, tolerances and access checked against the drawings.
- Supplier asked what they can hold before the tool is cut.
- Tooling cost weighed against the real first-run volume.
- Every fix made on a drawing, where it costs minutes.
How it fits the bigger picture
Design for Manufacture sits inside Stage 07 Engineer, after the design is functionally proven and before the compliance and marking work that closes the stage. It depends on a design that already does its job, and it hands forward a design that can be made at volume. With the review done and the fixes folded back into the drawings, the product is ready for the conformity work and then for Stage 08 Develop, where tooling is ordered and pre-production builds begin.
What it can do
It catches the expensive mistakes while they are still cheap to fix, on a drawing rather than in a tool. It turns the chosen suppliers from a name on a quote into a real constraint the design respects, and it keeps the tooling spend proportionate to a first run of 500 to 1,000 units rather than a volume the BOM of £38 to £55 has not yet earned.
What it can’t do
It can’t rescue a design that does not work yet; that is the job of the engineering before it. And it can’t replace talking to the actual supplier. A review tells you where the risks are; only the Stoke ceramics maker and the Manchester board house can tell you what they will genuinely commit to. DFM is the conversation that asks the right questions, not a substitute for the answers.
See the full 10-stage process →
Try it yourself
Take your design and read it as the factory will. For every moulded part, ask where the draft is and how it releases. For every pair of mating parts, stack the tolerances at their worst case and check they still fit. For every joint, ask how a hand or a fixture reaches it. Then ask your shortlisted supplier, in writing, what wall thickness and finish they can actually hold, and price the tooling against the run you can honestly forecast. Anywhere you are guessing, you are not ready to tool.
Not sure your design is ready for a factory yet? Start the Free Sprint → and pressure-test the route to production before you spend on tooling.
Your DFM checklist
Project notes: what breaks at the mould
▸ From the notebook · optional reading
Running the proofing box through a DFM review with Dan and Anna Hartley in Stockport, and the ceramic wall that nearly cost them a tool.
3 min read · click to open
The proofing box worked beautifully on the bench. It held 26°C ±0.5°C, drew under 30W, and Dan was ready to send the model straight to a toolmaker. The Sourdough School crowd were waiting. I asked for one afternoon first: walk it through a factory’s eyes before anyone cut anything.
The wall that was too thin
The first thing the review flagged was the ceramic. Anna had drawn a clean, slim wall on the Stoke-on-Trent tub because it looked elegant. The question I kept asking was “what breaks at the mould?” When we put the drawing in front of the Stoke supplier, the answer was blunt: at that thickness, across a run of 500 to 1,000, a meaningful share would crack or warp in the kiln. The drawn wall and the makeable wall were not the same wall.
What we actually did
We spent the afternoon on five things. We added draft to the tub so it would release cleanly. We stacked the tolerances between the body and the lid seat and found the seal was marginal at worst case, so we opened the seat. We checked a human could actually seat the Manchester PCB and the heating element, and reach the fasteners, then moved one boss that blocked a screwdriver. We thickened the ceramic wall to what Stoke would commit to. And we priced the tooling against the real first run, not the Year-one figure of around 3,000, so the £149 box stayed inside a BOM of £38 to £55.
Every one of those was a change on a drawing that afternoon. Found in hard tool steel, the wall thickness alone would have meant a new mould. The review cost an afternoon. The fix it caught would have cost a tool. That is the whole argument for doing DFM before, not after.
— Engineer stage, project notes, 2026
— Next activity → Engineer · CE/UKCA mark documents