An engineering process, applied to Amazon.
We design for margin, manufacturability, and FBA fit, in that order, with gates.
How does a product move from idea to FBA-ready?
Through ten gated stages, run in sequence. Each stage has a defined output, and the work does not advance until that output holds up.
This is the single place the methodology is walked end to end. Every stage produces something you can read, check, and carry forward, so you always know what you are paying for and what you are getting.
- 01 Discover Understand the seller, the listing, the margin structure, and the white-label product being escaped. Establish what success has to look like commercially. Output: opportunity brief
- 02 Define Turn the brief into a requirements set: target unit cost, key differentiators, category constraints, and the FBA parameters the product must fit inside. Output: requirements specification
- 03 Concept Generate and screen concept directions against the requirements. Resolve the product into a small number of buildable routes rather than a mood board. Output: screened concept set
- 04 Feasibility & FTO first-pass Pressure-test the leading concept for buildability and run an in-house freedom-to-operate read. Where a formal opinion is warranted, it goes to a named partner. Output: feasibility note + FTO first-pass
- 05 Design engineering (CAD) Take the chosen concept into parametric CAD: geometry, tolerances, materials, and the engineering decisions that determine cost and performance. Output: CAD model + drawings
- 06 DFM review Review the design for manufacture against the intended process. Strip cost and risk out of the part before any tooling money is committed. Output: DFM report + revisions
- 07 Prototyping Build physical prototypes to validate fit, function, and feel. Close the gap between a model that looks right and a part that works. Output: validated prototype
- 08 Tooling strategy Decide the tooling route that meets the unit-cost target at the realistic volume, and set out who owns the tooling positions at the end. Output: tooling plan + cost bands
- 09 Manufacturing & QC Move into production with a vetted partner under defined quality control. Hold the part to the spec, not to whatever the factory finds convenient. Output: production parts + QC records
- 10 FBA-ready handoff Land the product into Amazon’s world: packaging that meets FBA rules, FNSKU labelling, and the documentation needed for clean inbound. Output: FBA-ready stock + handoff pack
Stage names follow Cast Iron CAD’s framework.
What separates an engineering process from an agency pitch?
Gates. A process you can fail at a known point, on evidence, before it costs you money.
A defined check the work has to pass before the next stage starts. DFM reviews, FMEAs, and FEA where appropriate. If the part will not hit the unit-cost target, that surfaces at the gate, not at the quote from the factory.
A forecast with no checkpoint behind it. The work moves forward on optimism, and the first hard test of whether it was buildable arrives when the tooling invoice does. By then the decision has already been made for you.
We use DFM reviews, FMEAs, and FEA where appropriate, not because it’s fashionable, but because it’s the only way to hit unit-cost targets with factories you don’t own.
How do you handle IP and compliance?
We coordinate IP and compliance via partners — we do not provide legal advice.
A first-pass freedom-to-operate read happens in-house, early, so an obvious collision is caught before design effort goes into it. That read is engineering judgement, not a legal opinion, and we are clear about the difference.
When a formal opinion is needed, it comes from named partners: patent attorneys for IP, regulatory consultants for category compliance. Dave is not qualified to give legal, patent, or compliance advice, and offering it would be the wrong service to sell. Coordinating the right specialists at the right point is the correct way to handle it, and it is the way we work.
What does designing for FBA actually constrain?
More than most product design accounts for, because the marketplace is part of the specification, not an afterthought.
MOQ implications shape the tooling decision from the start. The volume a factory needs to run at to hit your unit cost has to square with the stock you can realistically sell and fund, so the tooling route and the MOQ are designed together, not negotiated apart.
FBA packaging rules, FNSKU requirements, and category compliance are all designed in, not bolted on. Pack format and dimensional weight feed back into the product geometry, FNSKU labelling is planned into the packaging artwork, and category-specific compliance, electronics, cosmetics, and food-adjacent goods each carry their own, is mapped before commitment rather than discovered at the border.
Lead-time realities sit across all of it. Tooling lead time, production lead time, and inbound shipping have to line up with Amazon inventory planning, because a product that arrives late or out of stock loses rank regardless of how well it is engineered.
Where do you have it made?
UK-first to prove the concept, Asia-for-scale when the unit economics demand it. The choice is made on numbers, not by default.
For proof-of-concept and early validation, UK manufacture keeps the loop tight: shorter lead times, easier quality oversight, and the ability to iterate without a six-week round trip. It is rarely the cheapest unit cost, and we say so.
When the volume and the target unit cost can only be met offshore, we move production to Asia through vetted partners, with the trade-offs stated plainly: lower unit cost against longer lead times, larger MOQs, and quality control that has to be designed in rather than assumed. The point of running the engineering and DFM work first is that the product file is settled before it goes to a factory you do not own, wherever that factory is.
Cast Iron CAD prototyping bench · Plus X Innovation, Brighton
What does this look like on a real product?
“The discipline that gets a part to a unit-cost target is the same discipline, whether or not it ends up on Amazon.”
The clearest read on the process is a project that ran through it. The engineering patterns are the same ones an Amazon seller hits moving off white-label: DFM, IP, unit economics, supplier-managed tooling.
This is Cast Iron CAD’s product development process, primed for Amazon. Forty-plus projects’ worth of engineering discipline, applied to a marketplace that now rewards differentiation.
This is the process. The Edge Sprint is where it starts.
A 1–2 week, fixed-price read on whether your product is worth engineering. £1,250 + VAT. Delivered by Dave.