Not as much as you fear, and not nothing. Here is the straight answer — the inputs you actually need, the money and time buckets, the skills you can hand off, and a readiness checklist to run before you spend on tooling.
The short answer: making a physical product takes four things — a clear definition of what it is and who it is for, a manufacturable design, money for tooling and a first run, and someone to manage the making (which can be a partner, not you). You do not need engineering skills, CAD, or a patent to begin. On the patent question: this is general information, not legal advice — talk to a patent attorney before you publicly disclose or sell, since that can affect your rights. The one hard requirement is a settled design, because tooling assumes the design is final. Everything else you can buy as a service or sequence cheaply up front.
Here is the whole path from idea to a packed product: what you bring, what a manufacturing partner handles, and the failure that bites at each stage if it's rushed. You don't need all of it on day one — most people start at stage 1 with a sketch. (DFM thinking doesn't wait for stage 4: a good partner folds make-it-cheaper feedback in as early as the prototype, so the design and the DFM review shape each other.)
| Stage | What you provide | What the partner does | What breaks if rushed |
|---|---|---|---|
| 1. Define | A one-page description: what it does, who it's for, size, rough material, use | Turns it into a plan; flags the open questions first | A vague spec → quotes you can't compare and redesigns later |
| 2. Design / CAD | A sketch, a sample, or CAD if you have it (not required) | Production-ready 3D CAD + 2D drawings with tolerances; suggests the material | No tolerances or wrong resin → parts that don't fit or don't last |
| 3. Prototype | Feedback on look, fit, function | A 3D-printed or machined prototype to prove shape; iterate cheaply | Approving a print that can't be molded → tooling rework later |
| 4. DFM review | Your must-haves and constraints | Reviews wall thickness, draft, undercuts, material, part count — cost-cutting changes before steel is cut | Skipping it → sink, warp, short shots, or a tool that must be recut |
| 5. Tooling | Sign-off on the locked design; tool ownership in writing | Cuts the mold (aluminum or steel to match volume); runs first (T1) samples | Locking a design that's still moving → paying to modify or recut steel |
| 6. First article + pilot | Approve the first-article parts | Measures parts to the drawing; a small pilot run to shake out real-speed issues | Approving without measuring → a full run out of spec |
| 7. Production + QC | Agree the inspection standard (an AQL plan) | Full runs inspected to that standard; reorders from the proven tool | No agreed standard → defects reach your customers |
| 8. Packaging + launch | Branding, retail needs, any market compliance | Retail-ready packaging and labeling; freight to you | Ignoring compliance/packaging → held shipments and returns |
The decisions are yours — you define at 1, approve the prototype, tool sign-off and first article at 3, 5 and 6, agree the inspection standard at 7, and set requirements at 4 and 8 — while the production craft (CAD, DFM, cutting the tool, inspecting) is the partner's. The physical gate artifacts worth insisting on: a frozen STEP file + drawing revision before tooling, a signed first-article report before full production, and a written tool-ownership letter so the mold is yours. A single accountable partner running the whole table keeps you from stitching separate vendors together and owning every seam between them.
This is where most would-be inventors talk themselves out of it. You can hand off almost all of the technical work:
| Keep (only you can) | Hand off (buy as a service) |
|---|---|
| Define the product & the buyer | CAD modeling & drawings |
| Validate that people want it | Design-for-manufacturing review |
| Fund it & set the budget | Tooling & mold engineering |
| Run the selling & the business | Production, assembly, inspection, logistics |
Translation: you do not need to be an engineer. You need to know your customer, make decisions, and fund the work. The engineering is a service — our DFM review and tooling answers show what that service covers.
Any single dollar figure is an estimate, not a quote (it depends on the part), but it helps to budget by bucket, cheapest first:
A useful way to frame it: everything above the tooling line is cheap insurance that de-risks the one big, mostly non-refundable commitment — the tool. Spend in that order (cheap steps first, tool last); tooling before you've validated is the costly way round. See what it costs to manufacture for how the drivers move these numbers, and get a real figure by having your actual part reviewed rather than guessing.
Driven mostly by how settled your design is. A locked, simple design moves through the stages — design, tooling, first-article inspection, pilot, production — far faster than a complex design that keeps changing. The single biggest schedule-killer is making design changes after the tool is cut, which is why locking the design first saves both time and money (see prototype to production). A manufacturer gives you a real timeline once they have seen the part, rather than a generic promise. It also helps to think in phases rather than one lump of time: design and DFM, then tooling, then first-article approval, then a pilot, then production — each phase has a clear exit gate — design frozen, first article approved in writing, pilot passed — so knowing which gate you are at tells you what is next and what is holding you up. Your own review turnarounds (signing off the design and the first samples) are part of the schedule too, so being responsive at those points keeps the whole thing moving.
If most boxes are unchecked, you are not behind — you just know your next moves. Start with the cheap ones (definition, validation) before the expensive one (tooling).
Say your product is a simple one-piece molded phone stand. Walk it through the four requirements: (1) Definition — "a folding desk stand for phones and small tablets, for remote workers, hand-sized, polypropylene (the right resin for a living hinge)." (2) Design — hand it to a partner; it is a single part with a living hinge, so DFM gets specific: the hinge is thinned to a consistent web so it can flex without cracking, the gate is placed so the plastic flows across the hinge (not along it) to align the molecules that let a living hinge survive repeated folding, and the fold faces get enough draft to release — three changes that quietly decide whether the stand lasts. (3) Money — validation and a few printed prototypes are modest; the injection mold is the big one-time number (a simple single-cavity tool sits far below a complex multi-cavity one — treat any figure as an estimate until quoted); then the first run is a low per-part cost. (4) Manage — one partner runs design → tool → first articles → production. Contrast a "smart" stand with electronics and an app: same four requirements, but #2 and #3 balloon. Same test, very different scale — which is exactly why you get a real cost before committing.
Two quick definitions: tooling is the precision metal mold that shapes every part (the big one-time cost); a first-article inspection is a documented check that the first real parts match your drawing before full production.
Send Jack a sketch, a sample, or a description — he reviews it himself and tells you exactly what it would take to make, and what to do first. NDA first.
Book a 30-min callThe full pathFour things: a clear definition of the product, a design that can be manufactured, money for tooling and a first production run, and someone to manage the making — which can be a partner, not you. You do not need engineering skills, CAD, or a patent to start; you need a clear idea of what the product does and who it is for, plus a willingness to validate demand before you spend on tooling. The biggest single requirement is a settled design, because tooling assumes the design is final.
Fewer than people assume. You need to understand your customer and your product; the technical skills — CAD, design-for-manufacturing, tooling, quality — can be handed to a manufacturing partner. What you cannot outsource is deciding what the product is, validating that people want it, funding it, and running the business of selling it. Treat engineering as a service you buy, not a barrier to entry.
It varies with the part, so any single figure is an estimate, not a quote, but plan for several buckets: validation (cheap), prototyping, the one-time tooling, the first production run, plus packaging, freight, and the cost of selling. For plastic products the tooling is usually the biggest up-front number and falls to a low per-part cost at volume. Get a real manufacturing cost by having your actual part reviewed before you commit.
It depends mostly on how settled the design is and the product's complexity. A simple, locked design moves through design, tooling, first articles, and a first run far faster than a complex one that is still changing. Rather than a generic estimate, a manufacturer can give you a real timeline once they have seen the part. Changes after tooling are the main thing that stretches a schedule.
None are required to start making a product. A patent is a separate legal decision — this is general information, not legal advice, and you should talk to a patent attorney before you publicly disclose or sell; a prototype helps but you can validate demand before building one; and you can begin exploring manufacturing as an individual. What you should have before tooling is a settled design, a validated demand signal, and a real cost — in that order.
No. The engineering is a service you buy from a manufacturing partner. You bring the idea, the customer knowledge, and the funding.
Many do, by handing the technical work to a partner and sequencing the cheap validation steps first. The constraint is usually funding and time for selling, not engineering.
A settled design plus a validated demand signal. Tooling assumes the design is final and that someone wants the product.