The product development process

Product development is easy to picture as one long job: have the idea, hire someone, get the product. In practice it is a series of smaller jobs, and each one exists to answer a different question. Skipping a question does not make it go away. It comes back later, when it costs more to answer.

The six phases below are described with the most generic names available. Companies and industries use different labels for the same steps (Stage-Gate®, NPI, EVT/DVT/PVT, and many house-specific systems), but the order is nearly always the same, and the questions are the same.

The six phases

1. Project Initiation

   Question: What is being built, for whom, and how will anyone know it works? Result: a shared understanding and a first written set of requirements.

2. Feasibility

   Question: Can this work at all, physically and financially? Result: calculations, research, rough tests, and a list of the biggest risks.

3. Proof of Concept

   Question: Does the riskiest part work in real hardware? Result: a rough build that proves the core idea. Usually ugly, and that is fine.

4. Prototype

   Question: Does the whole product work together, and does it look and feel right? Result: one or a few units that work like the product and look like the product.

5. Preproduction

   Question: Can it be made repeatedly, at the target cost, and pass its tests? Result: a design tuned for manufacturing, a small pilot run, test results.

6. Production

   Question: Can it be made and shipped at volume, reliably? Result: finished goods and a supply chain.

What a phase gate is

A phase gate is the decision point at the end of a phase. The people involved look at what the phase found, update the requirements, and decide: go on, change direction, pause, or stop. Each gate has exit criteria, meaning the things that must be true before the next phase makes sense.

Gates protect the budget. At every gate the amount spent is known, something real exists (a report, a test result, a working model), and the next step is chosen with better information than before. Stopping at a gate is not a failure. Sometimes the most valuable thing a Feasibility phase can report is that the product cannot hit its price.

Not every firm covers every phase

Engineering firms differ in scope. Some take a product from idea to production. Others specialize in the early, uncertain phases and hand off to a contract manufacturer or a production-focused firm. Neither model is better; they are priced and staffed differently. A firm that does only what it is good at usually costs less per phase than one spread across all of them. The practical point: ask any firm which phases it covers and where it hands off, and plan for the handoff from the start.

Where people get stuck

Jumping to Prototype. The most common first request is "a prototype." Often the idea has not been through Feasibility or Proof of Concept. A polished prototype of an unproven idea is the most expensive way to find a basic problem.

Treating a prototype as a product. A prototype that works on a bench is not ready for a factory. Preproduction exists because making one of something and making ten thousand are different engineering problems.

Spending out of order. Money goes to branding, patents or packaging before anyone knows what the product costs to make. The costs page tells that story in numbers.

Never writing it down. Without written requirements, "done" is whatever each person imagines. A requirements document is the fix, and Project Initiation explains it.

Find the current phase

People tend to place an idea one or two phases ahead of where it is. That is natural; the person has lived with the idea and can see it finished. The table below is a reality check. Pick the earliest row that is fully true.

If this sounds right...	The idea is probably in...	A sensible next step
There is an idea and maybe some sketches. Nothing has been built or calculated.	Project Initiation	Write down what the product must do. The planning kit is built for this.
The product's job is clear, but not whether it can be done at the target price or size.	Feasibility	A short study of the biggest risks: physics, cost, regulations.
It can probably work, but the key feature has never worked in real hardware.	Proof of Concept	A rough build of only the risky part.
The core idea works in a rough build. A unit that works and looks like the product is needed.	Prototype	A defined prototype phase against written requirements.
A working prototype exists. It needs to be made in quantity at a target cost.	Preproduction	Design for manufacturing, supplier quotes, a pilot run, testing.
It is being made. Improvements, cost reduction or a new version are needed.	Production	A scoped change, which is often a good fit for a fixed price.

Landing in the first two rows is good news. Changes are cheap there. A requirement changed on paper costs an hour of conversation. The same change after tooling has been cut costs the tooling.