Last month I walked into a factory in Bao’an. The boss was proud. Real proud.
He showed me this “golden sample” of a smart speaker. Perfect finish. Clean seams. The weight felt right. He said his team “developed” it in three weeks.
I turned it over.
Saw the competitor’s logo scratched off the back.
That’s your product development partner right there. A guy with a Dremel tool and zero shame.
The Truth About “Co-Development”
Here’s what factories mean when they say “let’s develop together”:
They want you to pay for their learning curve.
You’re not partnering. You’re funding their mistakes. Every failed prototype? Your wallet. Every tooling revision? Your wallet. Every “oops we ordered the wrong chip” moment? Also your wallet.
I’ve seen this play out maybe 200 times.
The factory acts excited. They send engineers to meetings. Everyone nods a lot. Then reality hits and suddenly they can’t figure out why the Bluetooth keeps dropping or why the plastic warps in shipping.
But hey, they learned something new.
On your dime.
When It Actually Works
Sometimes it works.
Not often. But sometimes.
I worked with a lighting factory in Zhongshan two years back. Client wanted custom RGB strips with a specific diffusion pattern. Factory had the injection molding expertise but zero experience with the LED programming side.
Here’s what made it work:
They admitted what they didn’t know. They brought in a sub-contractor for the PCB work. They ate half the prototyping costs themselves because they wanted the capability for future clients.
That’s rare.
Most factories will just say “yes” to everything and figure it out later. Or not figure it out. Usually not.
The Liar’s Dictionary
You need to speak factory. Here’s the translation guide:
|
What They Say |
What It Means |
|---|---|
|
“We have experience with similar products” |
We made something vaguely related five years ago |
|
“Our R&D team is very strong” |
We have two guys who can use SolidWorks |
|
“We can start tooling right away” |
We haven’t actually reviewed your design yet |
|
“The timeline is no problem” |
We have no idea how long this will take |
|
“We’ll handle the certifications” |
We’ll Google how to fake a certificate |
|
“Trust us, we’re partners” |
We’re about to steal your design |
Learn this table. Tattoo it on your arm if you have to.
The Conversation You Need to Have
I sat through a product development kickoff last week. The factory kept saying “no problem, no problem” to every requirement.
So I stopped the meeting.
Me: “Show me a product you developed from scratch in the last year.”
Boss: “We make many products—”
Me: “From scratch. New mold. New everything.”
Long pause.
Boss: “Most of our products are based on existing platforms.”
There it is.
That’s not bad by itself. But now we know what we’re working with. They’re not an innovation lab. They’re a modification shop.
Factory: “We can make anything you design.”Me: “Who owns the tooling?”Factory: “Well, you pay for it, so…”Me: “So I own it?”Factory: “It stays here for production efficiency.”Me: “That’s not ownership.”Factory: “But you’re our partner!”Me: “Then put it in writing that I can move the mold anytime.”Factory: sudden meeting about another topic
If they won’t commit to mold ownership in writing, you’re renting, not owning.
What You Actually Need
Stop looking for a “partner.”
Start looking for a vendor with specific capabilities that match your design.
Here’s the checklist:
-
Do they already make something 70% similar? If yes, you’re modifying. That’s good. Modifications work. “New” doesn’t.
-
Can you see their in-house tooling workshop? If they outsource tooling, your costs will balloon and timelines are fiction.
-
Do they have a real QC lab or just a testing corner? Product development means constant testing. A “testing corner” with one caliper won’t cut it.
-
Will they sign an NDA that actually means something? If it’s a Chinese-only NDA with no enforcement clause, it’s toilet paper.
-
Have they worked with your target market’s regulations? EU? USA? Each has different hoops. If they’ve never done it, budget triple the time and money.
Most buyers skip this. They fall in love with the factory tour and the free lunch. Then they’re shocked when the project dies six months in.
The IP Protection Box
Your design will leak.
Accept it.
But you can slow it down. Sometimes that’s enough to get your product to market first.
What actually works:
Split the manufacturing. If your product has three main components, have three factories make them. Do final assembly yourself or through a fourth party. Yeah, it’s more logistics work. But now no single factory has your complete design.
We do this for clients through our sourcing network. Factory A makes the housing. Factory B does the electronics. Factory C handles packaging. Nobody sees the full picture.
Watermark your drawings. Not the obvious kind. The sneaky kind. Change a dimension by 0.2mm in each version you send to each