Kickstarter is coming

The Makerfaire is over and now the time has come to get the Gerbil Kickstarter up and running by end of June.

Two components of the Kickstarter that we need to solve are the intro video and the second is to re-evaluate the production costs to ensure we can deliver this for the price on Kickstarter.

The video has been a very difficult item to do since I have very little media experience. My initial idea was to get some freelancers to shoot an animated movie of 30 seconds but no offers came forward and it seems way too hard.

So back to the drawing board and shoot a movie myself. This requires a script that I have been working on with help from Dan. To make an effective script you need to know what main message is that you need to tell to your audience. Facebook was very helpful for me to discover this. One response on my post to FB was from a gentleman who sees my product as a simple Arduino with a shield that you could make yourself for a few Dollars. Obviously he has not understood the value proposition about the engraving complexity that requires 16 bits PWM and a different AVR chip with more capabilities than the UNO. Great feedback and teaches me to explain what I bring to the potential backers and what the actual cost structure is.

It’s not easy but let’s see what the cost elements are:

  • BOM parts costs: revisited the various suppliers and compared cost between them in Octopart see http://Www.octopart.com (online tool for sourcing electronic parts). I have seen huge differences in prices like 50%. Next is the available stock. Most suppliers don’t have stock beyond 1000 components. I have optimised my BOM and ordered the parts to assemble ten boards so I know that the BOM is correct and the boards are working well. What I learned from testing my BOM was that:

Stock is available: so you can order your amount, low on stock: now you have to manage multiple suppliers to make up the numbers which can be time and resource intensive or no stock: you either have to find a replacement part or order the parts with potential lengthy delays.

Testing the BOM part list. It is extra upfront cost that I bear now but testing the BOM and see whether my parts are correct and do work is essential for a successful Kickstarter. For example a crystal for the new AVR chip needs to be of high quality because this chip does not have the builtin full swing oscillator circuit (the makers from Atmel had to make room on the chip die for more functions so something had to go). So ordering the wrong crystal would be disastrous and require extensive SMD rework.

Next I found when testing the BOM was that three parts that did not work well: the SMD USB connector was not entirely flat but had protrusion frame parts so had to replace that with a different part, the SMD capacitors were not matching the footprint so had to find smaller ones and the cheap female stack headers were too low so back to the expensive ones.

New boards next to the original China manufactured board marked QC
Surface mount device assembly of the controllers

Also saw difference in price, the USB uart chip from ftdi was expensive, 100% more than another supplier based on just one letter on the chip “R= Tape and Reel”, “U=Tube”. The tube delivery was cheaper which I can’t explain since that are smaller quantities (50 pieces per tube). I could opt for a different USB chip like the Chinese ch340 but with the current windows driver issues for not digitally signed drivers would make USB driver installation very difficult for non-tech savvy users.

  • Assembly costs: used my current PCBs supplier to see the quantity brackets and alternative suppliers.
  • Insurance: nothing worse than losing a batch of 500 or more controllers but it is possible. The cost of insurance is about $1.50 per $100 value.
  • Import duties: there are import duties to be paid for parts going from the US to China (assembler of the pcbs).
  • Shipping agent costs for the assembled controllers from China to the world (big batches) or to Australia to the rest of the world (small batches)
  • End stop flag and L Shape bracket manufacturing. These parts are very basic and can be made locally here but the batch numbers determines the supplier for these. Alternatively we can ask the customer to make these themselves when prices are too high,
  • Firmware flashing: All controllers need to receive their firmware image.
  • Testing: All controllers need to be tested before shipping them out. That requires a test harness and machine.
  • Packaging: Controller and Parts need to packed into a suitable carton and protective foam.
  • Shipping agent: Depending on the number of backers, we could use a shipping agent to move these controllers in one go to a specific country to make it more cost effective for the backers.
  • 5% Fail out; to provide great customer service we need to be ready to immediately replace any defective products and keep plenty of spares for long term requests for spare parts.  5% is a worst-case estimate of what might be needed.

So that brings me to the order strategy: Ordering everything in one batch would be a big risk so, it is smart to use a few batches but not too many since shipping and batch size assembly cost will go up. All comes down to the number of backers, too few and the cost are high. High demand would drop the part, assembly and shipping cost but testing and packing will require outsourcing (hire labour) which is an extra cost.

Okay that’s for today, next is shooting the video.

4 thoughts on “Kickstarter is coming”

  1. Your controller board sounds great! I was thinking about designing something similar myself, but there is always this time issue…

    Will you publish the software and the design files as well or are you focusing on selling ready made boards? So far I could not find any download links here on your blog.

Leave a Reply to Paul DeGroot Cancel reply