3 ms·

While I agree completely with having an experienced designer spec hardware costs before the campaign, having run a kickstarter campaign for a small Bluetooth de

While I agree completely with having an experienced designer spec hardware costs before the campaign, having run a kickstarter campaign for a small Bluetooth device, where the campaign failed miserably, I can offer an interesting but sad observation. Everybody, but everybody, but everybody… And I do mean everybody, gave me the same advice over and over and over again. And that was to focus on the marketing and the message and to pick a price point before actually designing. Even putative campaign backers told me that they thought being able to match up specifications and deliver on promises was just a technical issue. The reality is, as many others have noted, producing inexpensive commodity consumer-electronics is a far more expensive proposition than many realize, especially when you're talking about low manufacturing volumes or surprisingly tricky and expensive operations such as FCC regulatory compliance. In the end I think this sort of thing is quite common… People hear and believe what they wanted to be true. And then they complain when companies cannot deliver on promises they made out of naïveté. Sorry for potato quality grammar and occasional incorrect homonym… And being forced to use text to speech!

ChuckMcM 12y ago
You make a good point, and the reality is that both have to be true. A good marketer will figure out at what price a device could reasonably be sold for. A good hardware designer could then tell you how close you could come to that, given the desired price point. A number of time folks have come to me with a "really great idea" but when priced out it didn't make sense. People just wouldn't buy it for what it would cost to make it, and "betting on the come" which is to say making 100,000 to get the price down on the hope that after you've sold 10,000 or so you will be able to sell another 90,000 is a risky business indeed. This though is part of the "nice thing" about Kickstarter. Which is you can set your price point for the quantity of units you offer on KS. Don't get funded? Couldn't make the market work, get more than you needed? Might be able to add a couple of points of margin to the bottom line. Danger of course is that if you get waaay more than you needed you'll be tempted to build V2 right away. Don't do that. Ship V1, pocket the margin and use that to design V2 with less risk after you've shipped V1. I'm curious about your bluetooth experience. I've got a project I'm advising and any previous experience is always good experience in my book. The BT vendor (TI in this case) has offered to help with EMC certification although the project already has a bid for that from a local lab at a cost of about $50,000.
- an_d_rew 12y ago
  Heh. Take a look at the EEVblog form for 'dogy crowdfunded projects' for some fun reading (http://www.eevblog.com/forum/crowd-funded-projects/ http://www.eevblog.com/forum/crowd-funded-projects/). The experienced people there can often rate a project's "bull$hit factor" just by costing out the BOM and assembly and comparing it to the offer price. Regarding Bluetooth (LE), I've got a history with ARM stuff so stuck with Nordic's offerings. Cypress & CSI have nice stuff, but they really lock you into their environments, IMHO. I've got a nice full-stack gcc-4.9 based dev/debug flow for the full Nordic stack, and that makes my development way easier (and much, much cheaper since I don't need to hassle with IAR or Keil). Mind you, I'm also comfortable with writing newlib CRT routines from scratch, so what's "comfortable and cost-effective" for me may not match your dev team's experience. Anyway, back to BLE and KS. I found that the vast majority of 'small BLE device' Kickstarters seemed to massively underestimate the cost and complexity of both EMC design/certification, and Bluetooth SIG compliance/certification. That stuff is expensive! As you may or may not know, if you design your own circuitry, you need to qualify your entire radio PHY and do a hardware re-qual. That's on top of FCC/IC/CE/TELEC certifications. It ain't cheap and it ain't a pretty process. I ended up going with an Insight SiP ISP130301 Bluetooth LE Module because they had awesome antenna characteristics, were mechanically robust, they included the optional timing crystal that I wanted, and were relatively inexpensive (~8 USD in Q1000). Using one of those, you do NOT have to do B-SIG qualifications (as long as you use the qualified vendor BLE stack), and as long as you use an approved antenna (usually fixed and included on the module) you only have to do FCC (or equivalent) testing that any/all consumer electronics need to go through. For under Q10k, that's a no-brainer in price and ease-of-integration.

ChuckMcM 12y ago

You make a good point, and the reality is that both have to be true. A good marketer will figure out at what price a device could reasonably be sold for. A good hardware designer could then tell you how close you could come to that, given the desired price point. A number of time folks have come to me with a "really great idea" but when priced out it didn't make sense. People just wouldn't buy it for what it would cost to make it, and "betting on the come" which is to say making 100,000 to get the price down on the hope that after you've sold 10,000 or so you will be able to sell another 90,000 is a risky business indeed. This though is part of the "nice thing" about Kickstarter. Which is you can set your price point for the quantity of units you offer on KS. Don't get funded? Couldn't make the market work, get more than you needed? Might be able to add a couple of points of margin to the bottom line. Danger of course is that if you get waaay more than you needed you'll be tempted to build V2 right away. Don't do that. Ship V1, pocket the margin and use that to design V2 with less risk after you've shipped V1. I'm curious about your bluetooth experience. I've got a project I'm advising and any previous experience is always good experience in my book. The BT vendor (TI in this case) has offered to help with EMC certification although the project already has a bid for that from a local lab at a cost of about $50,000.

an_d_rew 12y ago
Heh. Take a look at the EEVblog form for 'dogy crowdfunded projects' for some fun reading (http://www.eevblog.com/forum/crowd-funded-projects/ http://www.eevblog.com/forum/crowd-funded-projects/). The experienced people there can often rate a project's "bull$hit factor" just by costing out the BOM and assembly and comparing it to the offer price. Regarding Bluetooth (LE), I've got a history with ARM stuff so stuck with Nordic's offerings. Cypress & CSI have nice stuff, but they really lock you into their environments, IMHO. I've got a nice full-stack gcc-4.9 based dev/debug flow for the full Nordic stack, and that makes my development way easier (and much, much cheaper since I don't need to hassle with IAR or Keil). Mind you, I'm also comfortable with writing newlib CRT routines from scratch, so what's "comfortable and cost-effective" for me may not match your dev team's experience. Anyway, back to BLE and KS. I found that the vast majority of 'small BLE device' Kickstarters seemed to massively underestimate the cost and complexity of both EMC design/certification, and Bluetooth SIG compliance/certification. That stuff is expensive! As you may or may not know, if you design your own circuitry, you need to qualify your entire radio PHY and do a hardware re-qual. That's on top of FCC/IC/CE/TELEC certifications. It ain't cheap and it ain't a pretty process. I ended up going with an Insight SiP ISP130301 Bluetooth LE Module because they had awesome antenna characteristics, were mechanically robust, they included the optional timing crystal that I wanted, and were relatively inexpensive (~8 USD in Q1000). Using one of those, you do NOT have to do B-SIG qualifications (as long as you use the qualified vendor BLE stack), and as long as you use an approved antenna (usually fixed and included on the module) you only have to do FCC (or equivalent) testing that any/all consumer electronics need to go through. For under Q10k, that's a no-brainer in price and ease-of-integration.