How to choose a prototype company
A buyer checklist: the questions to ask, the red flags, what quotes should contain, and realistic price and timeline bands.
August 29, 20262 min read
Written by Konstantin Dolgan, Ph.D., NPDP
Founder & CEO, Product Development Engineer
Published August 29, 2026
Choose a prototype company by matching its accountability to your stage: parts vendors are accountable for geometry, development firms are accountable for whether the product works. Ask for products they took to production, confirm you own the CAD and the IP in writing, and start with a small fixed-scope engagement before committing to a full program.
LA NPDT works this way by default, starting with a small, fixed-scope engineering assessment rather than a program commitment.
Match the company to your stage
Where you are | What you need | What to avoid |
|---|---|---|
An idea or a sketch | Concept design and feasibility | Buying parts before the concept is settled |
A rough mockup that sort of works | Engineering and a real prototype | Jumping to tooling quotes |
A prototype that works but cannot be made | Design for manufacturing | Another prototype of the same design |
A released design | A contract manufacturer | Paying a development firm for production |
Ten questions that separate real firms from the rest
- Which of these products reached production, and who manufactured them?
- Who owns the CAD, the source files, and the IP when we are done? Show me the clause.
- What is the first deliverable, what does it cost, and what does it not include?
- What happens when the first prototype fails a test?
- Who on your team does the mechanical work, and who does the electronics?
- How do you handle design for manufacturing, and at what point in the project?
- Can I take the files to another shop for quotes?
- What have you built in my product category?
- What is your change-order process when requirements shift?
- What is the realistic total to reach a sellable product, including the parts you are not quoting today?
Red flags
- A guaranteed outcome, a licensing promise, or talk of "getting your product in front of major companies."
- A large upfront fee with no defined deliverable.
- Refusal to release CAD or to let you get outside quotes.
- Renderings presented as prototypes.
- No named engineer on the project.
- A quote with no assumptions, no exclusions, and no revision allowance.
- Pressure to file a patent through their partner before any engineering happens.
What a real quote contains
- A scope statement in plain language, with explicit exclusions.
- Named deliverables: CAD, drawings, a physical unit, a test report.
- A revision allowance and the rate for work beyond it.
- A schedule with decision points, not just an end date.
- An assumption list, so you can see what the price depends on.
Price and timeline bands by product type
Product type | Prototype cost | To a testable unit |
|---|---|---|
Simple molded plastic product | $8,000 - $20,000 | 4 - 8 weeks |
Mechanism or moving assembly | $15,000 - $40,000 | 6 - 12 weeks |
Connected device with custom board | $35,000 - $150,000+ | 3 - 9 months |
Medical or regulated device | $60,000 - $300,000+ | 6 - 18 months |
Estimate your own with the prototype cost calculator, or compare engagement types on our solutions.
Ready to develop your product?
Talk to an engineerFrequently asked questions
Should I choose a local prototype company?
Proximity matters less than it used to for CAD work and more than you think for hands-on testing. Video calls handle design reviews well; physical iteration does not travel. What actually predicts success is response time, a named engineer, and whether the firm has produced products in your category. We work with clients nationwide from Ruston, Louisiana, and ship physical units for testing as part of every prototype engagement.
How much should the first engagement cost?
Enough to buy real work, small enough that a bad fit is survivable. In practice that is a few hundred to a few thousand dollars for a review or a feasibility study. Any firm that requires a five-figure commitment before you have seen a deliverable is asking you to take all the risk.
What if I already have a prototype that failed?
Bring it. A failed prototype is a data set: it tells an engineer which assumptions were wrong, which is faster than starting from a description. Most of the redesign work we do begins with a unit somebody else built.
Related articles
All articles
EU Cyber Resilience Act 2026: What Connected Hardware Founders Should Build Into the Prototype Now
For founders and product teams shipping connected hardware into the EU: what the CRA already requires now (reporting) and what must be designed into the next prototype before full application in 2027.
CPSC Lithium Battery Rules Are Moving in 2026: What Hardware Founders Should Change Before the Next Prototype
For founders and product teams building battery-powered industrial, pet, outdoor, or consumer hardware: what CPSC proposed on June 24, 2026 for lithium-ion micromobility batteries, and how to keep enclosure, BMS, and charger choices from locking you into a redesign after tooling.
Pfas Rules Are Moving in 2026: What Hardware Founders Should Change in Material Specs Before Prototype
For founders and product teams building industrial, medical, pet, outdoor, or consumer hardware: what changed in US TSCA PFAS reporting timing and the EU REACH restriction process in 2026, and how to keep early material choices from locking you into a redesign after tooling.
Services related to this guide
- Best product development companiesHow the leading US firms compare, and which one fits your stage and budget.
- Best product design companiesDesign studios compared by scope, industry and what they hand off.
- Product engineering companiesEngineering firms for physical products compared by discipline, rigor and budget.
- Consultant vs firm vs freelancerWhat each option costs, what it delivers and where to find them.