We manufactured dive computers in Canada - due to the niche nature of the product, there weren't any obvious gains from manufacturing abroad, and huge gains in speed of iteration and innovation by keeping production in North America.
That being said, this shaped our strategy - we were constrained to be high price/high innovation. Our first line of products had machined aluminium casings and retailed for over $1000. As time went on I drove the change to injection molded casings and a price from $450-$750.
Thanks, Maggie. This is another great real-world example, and a really important one.
Your experience highlights something that often gets missed: keeping production close doesn’t mean “super high prices forever”, as so many assume.
Moving from machined aluminum to injection-molded casings is a perfect illustration of how design, manufacturing, and economics evolve together when feedback loops are tight. You couldn’t have made that transition intelligently without deeply understanding the product, the users, and the production process.
I’d bet that even beyond that shift, you would have continued finding ways to drive costs down over time (without offshoring to China)—through design simplification, process improvements, supplier learning... And of course, as volumes grow, parts, materials, and assembly costs should fall naturally as well.
Phenomenal framing with the Neil Armstrong thought experiment. The distinction between cost optimization and category creation is something most reshoring discussions miss entirely. I worked at a hardware startup that tried to iterate from overseas and it was brutal, 3 week turnaroud times killed momentum compared to when we moved to a domestic CM. The tacit knowledge loss angle deserves way more attention.
Thanks! That’s exactly the kind of real-world experience that gets lost in abstract reshoring debates.
The three-week turnaround point is such a good example: at that point you’re no longer iterating; you’re just batching guesses. And as you said, once the tacit knowledge is gone, no amount of Slack messages or CAD files really replaces it.
I appreciate you sharing that — it reinforces the argument better than any statistic could.
Great insight, Guy.
We manufactured dive computers in Canada - due to the niche nature of the product, there weren't any obvious gains from manufacturing abroad, and huge gains in speed of iteration and innovation by keeping production in North America.
That being said, this shaped our strategy - we were constrained to be high price/high innovation. Our first line of products had machined aluminium casings and retailed for over $1000. As time went on I drove the change to injection molded casings and a price from $450-$750.
Thanks, Maggie. This is another great real-world example, and a really important one.
Your experience highlights something that often gets missed: keeping production close doesn’t mean “super high prices forever”, as so many assume.
Moving from machined aluminum to injection-molded casings is a perfect illustration of how design, manufacturing, and economics evolve together when feedback loops are tight. You couldn’t have made that transition intelligently without deeply understanding the product, the users, and the production process.
I’d bet that even beyond that shift, you would have continued finding ways to drive costs down over time (without offshoring to China)—through design simplification, process improvements, supplier learning... And of course, as volumes grow, parts, materials, and assembly costs should fall naturally as well.
Really appreciate you sharing this!
Phenomenal framing with the Neil Armstrong thought experiment. The distinction between cost optimization and category creation is something most reshoring discussions miss entirely. I worked at a hardware startup that tried to iterate from overseas and it was brutal, 3 week turnaroud times killed momentum compared to when we moved to a domestic CM. The tacit knowledge loss angle deserves way more attention.
Thanks! That’s exactly the kind of real-world experience that gets lost in abstract reshoring debates.
The three-week turnaround point is such a good example: at that point you’re no longer iterating; you’re just batching guesses. And as you said, once the tacit knowledge is gone, no amount of Slack messages or CAD files really replaces it.
I appreciate you sharing that — it reinforces the argument better than any statistic could.