Losing sleep about manufacturing startup risk

Startups talk a lot about risk: risk of not getting funding, risk of not making a splash on social media, risk of not landing that big order. But I hear a lot less worrying about the risks of execution once funding is wrapped up, once orders start pouring in.  Some of the risks those of us who live and breathe manufacturing worry about include:Counting sheep

  • Having material on hand to build
    • Seems simple, but it’s a rare business model that works along the lines of “once we get orders we’ll buy inventory.” So when that first order, or first big order, finally arrives, it’s likely too late to ask “gee, what is my lead time?”
    • While custom parts are the more obvious risk item here, standard off-the-shelf parts can be problematic too
    • Just because you can buy 100 components from a distributor and have them overnighted does not mean you can get 10,000 with two-day delivery
  • All those pesky certifications
    • In most countries and for most products, there are regulatory requirements to be completed before one can legally sell product. While prototypes are often exempt (or flown under the radar), there are huge risks (even go-to-jail risks) of not getting the paperwork squared off.
    • And sometimes it’s not just the bureaucracy, rather some very tough tests that must be passed (baby products anyone? Been there, lost sleep over that, the CPSC specs are draconian, and every country is different)
  • Processes
    • Having techs hand build 10 prototypes is one thing; having a production line build 10,000 is another.
    • Six sigma is not terribly meaningful for low volume runs, but in high volume statistics rule. There is no such thing as defect-free manufacturing. Or defect-free procurement. Or… you get the idea.
  • Logistics
    • All those materials have to get to the factory
    • End product needs to get to customers
    • Shipping times tend to be inversely proportional to shipping costs
    • Shipping internationally has a whole ‘nother layer of fun
  • Quality and Test
    • The product needs to work going out the door
    • It needs to work once it gets in the hands of the customer
    • It needs to keep on working for its rated life
    • Having the customer be the one to find product defects is a really bad idea
    • And again those pesky statistics
    • Don’t forget cosmetics—consumers tend to want a product that looks good and is functional, and who can blame them. There is nothing more frustrating (ok, that’s a wee exaggeration) than building up 100s of widgets only to have to reject them because the texture is wrong or the label is crooked or….
  • Iterations
    • Unlike software, iterating hardware is tricky
    • Iterations are the most common villain for blowing up manufacturing launch schedules: new components may be needed, new certifications, new testing, ick
  • Sleep is optional


Comments are closed.