The Product Development Process: Six Gates, With Entry and Exit Criteria for Each

The product development process runs through six gates: concept definition, feasibility, industrial design, engineering, validation, and production release. Each gate has an entry condition you must satisfy before work starts and an exit condition you must prove before money moves to the next phase. A moderate consumer product clears all six in nine to eighteen months and costs $30,000 to $150,000 depending on part count, materials, and whether it contains electronics. Skipping a gate does not save time. It moves the failure downstream, where it costs ten to forty times more to correct.

What follows is the gate structure as it works on real projects, with the specific documents that open and close each one. This is the framework a firm runs internally whether or not it shows the client the chart.

Why the Process Is Gated at All

A gate is a spending decision. Before each phase, someone has to say yes to the next tranche of budget, and that yes should be based on evidence produced by the phase that just ended, not on enthusiasm carried over from the beginning.

The reason this matters more for an independent inventor than for a corporation is that the inventor is both the project team and the executive who kills projects. Nobody else is going to tell you the cost model does not work. The gate structure is how you build that check into your own project, because the alternative is spending $80,000 discovering something a $2,000 phase would have told you in week three.

Roughly two thirds of the concepts that enter gate 1 on a well-run pipeline do not reach gate 5. That is the system working. A process where everything advances is not a process, it is a queue.

Gate 0 to Gate 1: Concept Definition

Entry criteria: a written description of the problem, the user, and the mechanism. Two paragraphs, not a patent claim. If you cannot describe what the product does in plain language without gesturing, you are not through the door yet.

Work in this phase: requirements capture. A product requirements document listing must-have functions, nice-to-have functions, target retail price, target user, size and weight constraints, regulatory category, and expected annual volume. Twenty to forty line items on a moderate product. Each requirement gets a verification method: measurement, inspection, test, or analysis.

Exit criteria: a signed requirements document with a target manufactured cost. That target is the single number that governs every decision downstream. Set it by working backward from retail: for a licensed consumer product moving through distribution, manufactured cost usually needs to land near 20% to 25% of shelf price.

Duration and cost: one to three weeks. $0 if you write it yourself, $1,500 to $4,000 if a firm facilitates it.

What kills a project here: the target cost and the required function are incompatible at the outset. Better to find out now.

Gate 1 to Gate 2: Feasibility and Freedom to Operate

Entry criteria: approved requirements document.

Work in this phase: two parallel tracks. Technical feasibility asks whether the mechanism works at all, which usually means a rough functional mock-up built from off-the-shelf parts, foam, and printed brackets. Commercial feasibility asks whether anyone already owns the idea, which means a prior art search against the USPTO patent records and a look at what is already on the shelf.

The searchable question is narrower than inventors expect. You are not asking whether anything similar exists. You are asking whether the specific combination of elements that makes your version work has already been claimed. A professional $399 patent search answers that in about two weeks with a written opinion and copies of the closest references.

Exit criteria: a working proof of principle, however ugly, plus a search report that either shows a clear path or identifies exactly what has to change. Many projects exit this gate with a modified concept rather than the original one.

Duration and cost: three to six weeks. $400 to $6,000.

What kills a project here: the mechanism does not work, or a live patent covers the core with claims you cannot design around. Both are cheap discoveries at this stage.

Gate 2 to Gate 3: Industrial Design

Entry criteria: proven mechanism, cleared or mapped prior art, cost target.

Work in this phase: form, interface, and use. Sketch exploration, three to five concept directions, then narrowing to one. Appearance models, either printed or machined, that you can hold and put in front of buyers. Color, material, and finish specification. Human-factors review: grip sizes, reach, force required, visibility of controls.

Industrial design is not decoration. It sets part count, parting-line strategy, and the general architecture of the housing, all of which drive the tooling budget. Understanding what separates industrial design from product engineering helps you read the deliverables you are about to receive.

Exit criteria: one selected direction with a signed-off appearance model, a CMF sheet, and a preliminary part-count estimate. A firm should be able to give you a tooling cost band at this point, not a number, but a band.

Duration and cost: four to eight weeks. $6,000 to $25,000.

What kills a project here: the form required for the function pushes part count and tool actions past the cost target. This is where a redesign is still cheap.

Gate 3 to Gate 4: Engineering and Design for Manufacturability

Entry criteria: approved industrial design, cost target, target annual volume.

Work in this phase: the heavy lift. Parametric CAD of every part. Material selection with data-sheet justification. Wall thickness, draft angles, rib ratios, boss geometry. Tolerance stack analysis on every critical interface. Fastener and snap-fit strategy. If there are electronics, schematic capture, PCB layout, firmware architecture, and thermal analysis run alongside the mechanical work.

This phase is where design for manufacturability either happens or gets skipped, and skipping it is the most expensive decision available in the entire process. A change in CAD costs an hour. The same change after steel is cut costs $3,000 to $40,000 plus four to eight weeks.

Exit criteria: a released CAD package, 2D drawings with GD&T on critical dimensions, a bill of materials with supplier part numbers, and a mold-flow analysis on the largest cosmetic parts. The deliverable list is specific and you should ask for it by name, which is one reason to understand everything a full product development engagement includes before you sign.

Duration and cost: eight to sixteen weeks. $15,000 to $70,000.

What kills a project here: unit cost lands 40% above target after real material and process selection. Rare, but it happens, and the honest move is to stop rather than to hope volume fixes it.

Gate 4 to Gate 5: Validation

Entry criteria: released CAD, funded prototype budget.

Work in this phase: build and break. Functional prototypes from the released CAD, made by CNC machining, urethane casting, or high-resolution printing. Then testing against the verification methods written into the requirements document at gate 0. Drop tests. Cycle tests on any moving feature. Environmental exposure if the product lives outdoors. Compliance pre-testing if the product needs a safety certification.

Expect two to four iterations. Each round is two to five weeks and $2,500 to $15,000. Products that pass validation on the first build usually had a light test plan rather than a good design.

Exit criteria: every requirement verified, a test report per requirement, and any failures either fixed and retested or accepted in writing with a rationale.

Duration and cost: eight to sixteen weeks. $8,000 to $45,000.

What kills a project here: a failure mode that has no fix inside the cost target. This is the last cheap exit.

Gate 5 to Gate 6: Tooling and Pilot Production

Entry criteria: validated design, a selected manufacturer, a purchase order, and a written agreement covering who owns the tooling.

Work in this phase: tool design review, steel cut, first shots, part inspection, tool adjustment, second shots, first article inspection, then a pilot run of 200 to 2,000 units. Assembly documentation gets written here. Packaging gets designed and drop-tested here. This is also where the bridge from a validated design to a running production line either holds or breaks, usually on documentation quality rather than on part geometry.

Tooling timelines are the longest single block in the project. A simple aluminum prototype tool runs about $3,000 and four weeks. Production steel runs $12,000 to $60,000 per tool and eight to sixteen weeks including sampling and adjustment.

Exit criteria: first article inspection passed, pilot units built on production tooling with production process, and a signed golden sample retained by both parties.

Duration and cost: twelve to twenty-four weeks. $15,000 to $120,000 depending on part count.

What kills a project here: almost nothing kills a project at gate 5. That is the problem. Once steel is cut, projects push forward on sunk cost. Everything meaningful had to be caught earlier.

The Whole Process on One Page

Add the low ends and you get about 36 weeks and $44,400. Add the high ends and you get about 93 weeks and $270,000. Most independent-inventor products land between those, and the honest planning number for a moderate single-material consumer product with no electronics is 12 to 15 months and $55,000 to $95,000 through pilot production. A fuller picture of where the money goes across the whole journey sits in the complete inventor cost breakdown.

Where the Patent Track Runs Alongside

The patent timeline is a separate track that touches the product track at two points. A provisional application filed at gate 1 gives you twelve months of priority while the design settles. The non-provisional gets filed at or before that twelve-month deadline, ideally after gate 3, because by then the design is stable enough that the claims describe what you will sell rather than what you sketched a year earlier.

Filing too early is the common error. A patent that describes the concept version, not the engineered version, protects a product nobody will make. The interaction between these two clocks is laid out in the month-by-month patent timeline, and the USPTO’s own process overview covers the examination side.

The Three Gates Inventors Skip Most Often

Gate 1, because prior art searching feels like looking for bad news. The result is a design that gets engineered around a claim nobody read until a manufacturer’s counsel found it.

Gate 3’s DFM pass, because engineering that produces pretty renders feels finished. The result is a mold quote that comes back 60% above the estimate, or worse, a mold that gets cut anyway and produces parts that will not assemble.

Gate 4’s test plan, because the prototype worked when the inventor held it. The result is a warranty rate discovered at unit 4,000. Deciding when to bring in a product design firm is largely a question of which of these three you are equipped to run yourself.

Running the Process With a Firm Instead of Alone

A firm compresses gates 2 through 5 because it has done the sequence before and owns the software seats and the vendor relationships. It does not compress gate 0 or gate 1, and those are the two you can run yourself for close to nothing.

The practical starting move for an independent inventor is to write the requirements document, set the cost target, then buy a prior art search before spending anything on design. That is a few hundred dollars against a decision that governs $50,000 or more of downstream spending. If the search comes back clean and the cost target is plausible, a staged product development engagement picks up at gate 2 with the expensive questions already answered.

FAQ

How long does the product development process take?

Nine to eighteen months from concept to pilot production for a moderate consumer product without electronics. Add three to six months for products with a PCB and firmware. Add six to twelve months for anything requiring a safety certification or medical clearance. Tooling alone is eight to sixteen weeks and cannot be compressed much regardless of budget.

What are the stages of the product development process?

Concept definition, feasibility and prior art, industrial design, engineering and design for manufacturability, validation testing, and tooling with pilot production. Six phases separated by five decision points where money either moves forward or stops.

Do I need to complete every stage?

You need to satisfy every exit criterion. You do not need to pay someone to do it. Requirements definition and initial market work are genuinely doable alone. Tolerance stack analysis, mold-flow simulation, and tooling review are not, because the failures they catch only appear after money is unrecoverable.

What does each stage cost?

Roughly: concept $0 to $4,000, feasibility $400 to $6,000, industrial design $6,000 to $25,000, engineering $15,000 to $70,000, validation $8,000 to $45,000, tooling and pilot $15,000 to $120,000. The band is wide because part count and material choices drive it more than anything a firm controls.

Can I stop partway through and still have something of value?

Yes, at two points. After gate 3 you own a released CAD package and drawings, which is a licensable asset and the thing a manufacturer needs to quote. After gate 4 you own tested prototypes plus that CAD, which is what a licensee wants to see. Stopping mid-gate leaves you with files nobody else can pick up cleanly, so plan your exits at the boundaries.