Part of what went terribly was our servers initial ability to cope with the load. We had some "shark processes" lurking just beneath the surface that did not make themselves apparent until we were TechCrunched. They were throttling our database and we got bit hard. CloudFlare played a big part in not making this a total disaster. We managed to get a lot of the issues straightened out by noon. Next time we'll load test.

Kickstarter is a fantastic platform for supporting the arts. Part of what makes it so great is that everyone's incentives are aligned. Warhol quips aside, Art isn't usually done for financial reasons (speaking very broadly here.)

Design on the other hand walks this weird line between wanting to create something well and wanting to fuel an entity that's capable of creating something *better*.

We had to scrape together the resources to create the first version of Lockitron. Once we had it out there though, it was the *product* that enabled us pool together what we needed to create this new iteration.

With Kickstarter's current approach the funds are released to the backer before the product ships. I don't think this kills project's quality standards, but it certainly adds some pressure to "hang onto" as much of those funds as possible, especially when you are running the project *like a business* - especially as a first time business. It also increases the stress level for the participants where they are paying for the idea but take the risk of receiving a poor likeness.

Our guess was that we could mostly resolve this conundrum for our particular situation by changing the funding model. We created our crowd-funding service from scratch that collects payment information from backers but *does not charge them* until all the development obstacles have been overcome and the unit is ready to ship. This drives us to make the best product possible (so our backers follow through on their reservation), it qualifies our backers (non-spam), and gives us a much better idea of how many units to make. The absolute minimum number of units we had to hit where we could make this work was 1,000 (a lot of electronic component price breaks only start at 1,000) so we aimed for that. We built it quickly and on short notice, but if this experiment is a success, we would love to open source it.

It's like the the prisoner's dilemma right? In one-off interactions most people would choose to defect. In recurring trials though, the better strategy is to collaborate. We believe the value is in creating an entity that creates great things. Our business simply can not operate without trust - we literally have no choice but to collaborate. Our only chance for success is to commit to the long term. This may sound blindly obvious to any business owner, but sometimes as a consumer it's not always apparent, especially when "random" strangers on the internet are asking you for money.

I would be willing to pay anything to have it overnighted as soon as the plastic cooled