Wow, you need to get a different supplier, you're getting scammed if you're paying anywhere near that much at any sort of purchase quantity for manufacturing.
I think it's still possible to sell hardware for a higher price if you can provide better support, etc. as in the examples of Sparkfun/Adafruit but a lot of customers are going to just get the lowest price board and there's no way to compete on price with a lot of this stuff. I can buy a board on eBay for less than it would cost me to mail it to my neighbor. If you can come up with a new product then there is no competition at first but if it's successful enough the clones are soon to come so you have to be constantly innovating.
liudr:
The 328P is about $4 each. Then a nano clone costs only $3 -$3.50 so why designing anything?!
Futurlec (ships from HK) has been charging $2.20 ea for single DIP 328P's. Shipping is not free but for me so far, about $7. They sell the DIP 1284P for $7 and have a wide chip selection. I've spent about $40 a pop there.
I am using all these cheap modules in a thru-hole design. This is an open-source data logger so is the previous one. On the outside they look the same:
The software should carry the sales price. Don't expect to compete with China selling empty boards.
Futurlec has held that $2.20 ea ($200 for 100) for the DIP 328P-PU since at least 2012. I saw DIPMICRO beat that at the end of last year but didn't get to move on it till too late.
Where you shop and how many you buy is where you get your price. A factory that produces 100,000 to millions of boards is not paying small quantity prices for chips.
How often do we get farmers wanting to automate chicken coops? Enough to show there's likely many more who don't attempt (or even think of attempting) the project. With the right package, some will buy and pay to have it installed right. Just don't be wrong, it had better be tested and work from the start.
Finding the right market is also crucial. Like GoForSmoke said, provide the right package and carefree operation. That's worth value added, not just the hardware, which is cheap.
liudr:
The price must have just come down from about $4. Now it is $2
Maybe related to the MicroChip buyout. I'm glad I didn't notice the price go up, I would have been stressed out. I bought 25 with a digikey order before that happened and haven't needed any since.
liudr:
Anyway, most makers aren't going to compete with spar-dafruit.
I wouldn't recommend trying to compete, rather emulate what they're doing right (documentation, libraries, customer support, branding) but with products they don't sell.
liudr:
This is what I do now:
I do like the look of a board with no modules, though I don't know why I have that hang-up. However, I'm also a big fan of through hole because it's so much more accessible so the board with modules has a significant advantage. They can order the kit and it probably won't take much over an hour to assemble even for someone with only a little prior experience with a soldering iron! In that way the second seems even more "open" to me. Of course some will be happy to pay you extra for the assembled device too. If you can provide this at a lower cost by using the modules that's great. Even though your shipping cost will be higher than what a Chinese seller would pay, it's not so bad as trying to compete on a single module sale since it's divided among a handful of modules and the extra shipping cost is well worth it to a US buyer anxious to get their new toy soon.
Robin2:
Well they did not manage to lure the Arduino founders away from Atmel chips.
...R
Both AVR and PIC serve the education and hobby markets. The hobby market is dominated by AVR, while education splits them about evenly: microprocessor courses taught in computer science or business departments most often use PIC. STEM and physics programs tend toward AVR, while engineering can go either way. ... or neither.
Banzi and the other's choice of AVR is rooted in history and is unlikely to ever change to PIC, solely due to the product manufacturer's merger, or the machine architecture. All the growth seems to be 'vertically', away from 8 bit processors. Hobby and education markets will hold on to 8 bit processors for a long time. My money is on AVR to be the last chip standing, but who knows the future plans of Microchip for the two, "competing" product lines?
Fish are attracted to a more organic lure though. AVR - particularly Arduino - has that, right?