The Blueprint talks to Cameron Robertson, Co-Founder of


a product of Apigy

Lockitron, by Apigy, allows for keyless entry — controlled by your phone. The product lets you lock, or unlock, your door from anywhere in the world. Last fall Lockitron collected nearly $2.3 million in pre-orders for the second generation of the device on a crowdfunding platform they built from scratch. Today the team is just weeks away from shipping to their backers on July 15th. We sat down with Apigy Co-Founder Cameron Robertson to hear their story.

Tell us a bit about yourself and how you came to start working on Lockitron.

My background was all over the place. I started hacking and taking things apart as a kid when I was pretty young. I think the first computer that my dad had that I played with was a Macintosh SE/30. I was a few years old when I started playing around with that and I just loved it.

From there I would take things apart and I would hack with them. I started doing web pages when I was in middle school. I was just really fascinated by the internet and what was happening there.

However, I left that aside for a few years when I left to college and studied history, economy and finance. Near the end of school, I realized that I really liked the idea of making things and creating things and software was a really fascinating way to do that. While I was in school I actually helped to start a web development club at the University of Colorado with a friend of mine, which was kind of weird because I’m not a CS major and nor was he. That’s actually how I met Paul, because he was one of the first members of the web development club. We hit it off immediately. We found that we had very similar interests.

We weren’t going to close ourselves off to specific types of stuff. When we started the company, we said, “Look, we want to do something which really makes a difference and has an impact”, so we weren’t particularly interested in Facebook, Twitter, social type stuff. We wanted to really make something. When we first started the company, we didn’t know that meant hardware. It just so happened that this was Paul’s background, he had been hacking hardware for a long time.

How did you come up with the idea for a smart phone controlled lock?

We officially started the company back in 2009 and got in YC with a different idea, which even to this day is really hard to explain. That alone doesn’t bode well for the idea and is one of the reasons we killed it. But the simple version of it was that we were making a pizza button for your phone that you could press and have a pizza.

We had been working on the idea all summer throughout YC. We had zero customers and zero progress, and we went to talk to Paul Graham [Partner, Y Combinator]. We said, “We don’t like this anymore, we want to do something else.” He’s like, “Great, sounds good. What do you want to do?” We’re like, “Oh, okay, cool.” What a relief that was. Let’s do this.

We took a whiteboard, wrote down dozens and dozens of ideas, crossed a bunch of them out, looked at three of them and then we settled in on a smoke detector which sends you text messages when it goes off. That was very appealing to Paul [Gerhardt] and I. It immediately was like, wow we get to make something physical and have the software and the physical thing interact with each other and you can trigger commands from the virtual world into the real world.

We built the smoke detector in a couple days. We put together the website, a working demo and we went up and demoed it in front of investors and, in general, the reception was pretty good. We said, “Let’s feel this out. We’re not going to take any investor money or even try, because we’re not in a great position to raise and the market’s horrible. Let’s feel stuff out in this space.”

One of the original ideas that cropped up really quickly was well, what if we do this with a door lock? What if we connect a door lock to the internet? That was immediately appealing from a cognitive standpoint, just, wow, that’s a lot of utility. Being able to let people in remotely is a massive bonus. Dealing with keys is a pain. We didn’t really put together the first prototype until early 2010. This is our Lockitron product. It was a full replacement door lock.

We were at the point of two of us bootstrapping. We had no money. We had to start selling something. We put together our first prototype.

Story has it you build the initial prototype for Lockitron in a weekend. How has Lockitron changed over time?

Paul put it together in a weekend and then I started writing a real web front-end for it to actually manage users. We very quickly went to speaking to our network of people and interacting with potential customers. We found there was interest for it.

We got an initial order for several dozen, from a coworking space in the city, and we had to manually adapt Lockitron to all of these other special office locks. We learned a tremendous amount about locks and hardware and what all of that means and what we want to do with the product, but overall the general feedback was people love the utility of it. It solved their problem and it did what they wanted it to.

On our side, the problem was that we were buying all the components retail, they were very expensive off the shelf. The price we had to charge to even make a little bit of money was very high. It was $295.00. People buying it for $295.00 really, really wanted it, which was great because we learned a lot from the early adopters.

We got a little bit of money from the first batch and we knew that if we could lower the price to $150.00 or $100.00 then we can really shift the dynamic here to a much broader audience.

When you’re setting a price for the first version, how much profit margin were you allowing yourself?

We made under $100.00 per unit out of the $295.00, so 30%, or somewhere around there. It was decent, but not great and it was certainly not enough if we wanted to try and resell the product. Over the course of a year and a half, I think we sold $100,000 worth and it kept us going. We knew we wanted to eventually build a second version, and had a lot of beta users but we still hadn’t launched publicly. We finally launched the first version publicly in 2011 and got a TechCrunch article. It was really cool. We got a huge amount of press and our launch video got 140,000 views on YouTube.

We used that momentum to raise a little bit of money for angel investors. The goal of that was to get to the next generation of product, to really refine it and take what people loved about it and make it much, much better and fix the challenges at the same time. Around spring of 2012, we were were running low on money and knew we need to get this out there quickly. We need to either raise money to make the hardware or use this new thing called Kickstarter, which is looking really interesting. We were looking at Kickstarter, thinking about it and then we saw stuff like Twine, which was at the end of 2011, there was Elevation Dock, there was a few of these products which were starting to break out and raising $500,000, $1 million, and we’re like, “Oh, that’s really cool. That’s probably a really good way to do this.” We were kind of on that path and then we saw Pebble and we were like, “Wow, this is amazing. You can really bring in a community very quickly to your product.”

From that point forward we were geared toward the entire of summer of getting our product ready for a Kickstarter: making it easy to install, bring the price down, bring the quantity up, take it from a multi-component product to a single component that you put on your door and you don’t have to worry about anything else.

“The challenge in America is that you can find a lot of different suppliers, but finding contract manufacturers who bring everything together is tough.”

Why did you end up not using Kickstarter?

We had our video lined up and we were putting together our Kickstarter page. We had that ready and we submitted a Kickstarter application, because we thought, “Okay, we can submit and then we can edit the page later on once we get accepted.”

We submit. A day after our submission, Kickstarter posts this blog post that Kickstarter’s not a store and how the rules have now changed: you can only sell one of something, you can’t show renders in your video or on your page, and some other minor things. We thought, “Okay, that’s fine”. We got different renders to show some different colors, but we’ve got our prototype, we can rip that out. That’s fine. The next day after that, we get a rejection notice and we were like “Okay”. We kind of anticipated that.

Through Eric at Pebble, I’m like, “Hey, maybe we could reach out to someone directly at Kickstarter and see what happens”. We do and the response we got back was, “Oh no, you weren’t rejected because of the new rules, you were rejected because you were a home improvement product. We don’t let those on Kickstarter.” We went through all of the stages of grief very, very quickly.

I was already heading to New York that weekend, and I basically said, “Look, I’m coming to New York can I meet with you? We think this is an unfair assessment.” We brought up, “you sold light bulbs, you sold wall outlets, wall sockets, very home improvement-y things and our product has been used in small businesses, vacation rentals, individuals, and yes, homes. We care about replacing keys, not just putting this thing on your door.” There was no response to that.

What other options were there for you besides Kickstarter?

We had a few options at that point. We thought, “Gosh, we can go to Indiegogo, we can try to change the product to make it more indie for Kickstarter” and so forth. At the end of it, we felt that we didn’t want to be beholden to a platform.

We were so frustrated by it, we just wanted to show exactly our brand and our narrative purely by ourselves. Earlier that year Dalton Caldwell had where he was looking for $500,000 and he raised $800,000. We thought well, Dalton Caldwell is pretty well known. He’s got a big blog but maybe if we ask for the $150,000 that we’re looking for we can hit that $150,000 just on our own website. We took the leap and in a week or so, we built our own crowdfunding website.

Our designer put together a beautiful layout and had everything ready to go. We started trying to talk to a few people from the press. We didn’t have a PR agent or anything, and no one was responding to our stories. We were emailing people and no one responded. They weren’t even completely cold emails. TechCrunch, who had launched us in our V1 the year before, said, “Yeah, sure. We’ll do the story” and Matt Burns wrote about us. That was our one piece of launch press. We then put it on our video, our Twitter feed and our Facebook, which didn’t have many followers at that point. Within hours, it just blew up completely.

We were asking for $150,000. Within a few hours we had $500,000 and the video went absolutely crazy. Some folks posted it to Reddit and then it went to 9Gag and it very quickly rocketed to well over 2.7 million views. It just went absolutely crazy in and of itself. Then the other press finally followed on. When the dust had settled, we closed nearly $2.3 million from our campaign. It was a whirlwind. It was immensely gratifying, but there was a huge number of challenges with Selfstarter.

We built it ourselves so we had to build all the tools on the fly. Unlike a Kickstarter project, we didn’t have a framework for making updates and that was really bad for us, because we only got out a few updates during the thing. Like, “Oh gosh, we have to email 10,000 people. How do we do that?”

Setting up all these services and getting all that stuff together was a challenge, but we did think “Hey, we’ve done this and we don’t really need the software anymore. Let’s open it up.” That’s why we opened sourced Selfstarter and put it out there. We still essentially used the core of Selfstarter in our code. We just didn’t need to hang onto it internally for ourselves. We thought this wasn’t really directly related to what we’re building. It’s pretty skeletal, but if people want to use it, they can use it.

When you build all this stuff on your own, you can usually get out a great working prototype pretty quickly, but when you have to scale it that’s when things get tough. In the long term, we still highly recommend that people try to go with the platform, if they don’t have the skillset to deal with building the tools themselves or if they don’t want to.

It sounds like it exploded and you guys didn’t necessarily expect it. Looking back, can you point to things that might have lead to the success?

There’s a few factors that stuck out to us pretty quickly. We knew the value of video from our first launch way back and so we thought, “Oh cool, a video with 100,000 views is big,” and it is. That’s a pretty big deal. We knew getting the video right would matter. All of the 45 days we put into making it ourselves was worth it, every single one. So getting the video right was absolutely critical. The cool thing is it doesn’t mean you need a production company, although things are moving to be more and more professional. It just means you have to sell that narrative, tell that narrative.

Additionally, there are a couple of other things we tried out on our own and one of them was every time someone backed Lockitron, rather than giving them a generic Twitter message to tweet out, we gave them a backer number. It gave them a sense of ownership, and that absolutely blew up on Twitter. In the first few hours, it felt like every five or eight orders, the person was tweeting their owner number. There was this massive amount of excitement on Twitter with the early backer numbers. It gave people the sense of ownership. It gave them the non-generic thing to Tweet out as opposed to a generic “I backed this thing on Kickstarter” type message. We had a really great social media storm going on.

The backer number thing is funny because we almost didn’t include it. We built it in, I think at 4:00 a.m. in the morning the day of our launch. I was thinking about it, I was like, “Oh, we talked about this, maybe I should put it in there? Okay, yeah. We’ll throw it in there.” It just so happened that it was really incredible. The video, the backer number and the inherent graspability of the product itself helps a lot. It translates really well into video. If you’re doing something purely in software, it can be a massive challenge to put that into video. Whereas when you’re selling something physical, you should just show it off as beautifully as possible. People love that. By making something physical, you’re already at a huge advantage.

“If you're doing something purely in software, it can be a massive challenge to put that into video. Whereas when you're selling something physical, you should just show it off as beautifully as possible. People love that. By making something physical, you're already at a huge advantage.”

How did you find the investment community’s reaction to Lockitron in general before the campaign and after?

It wasn’t as night and day as you think it was. We didn’t really speak to investors before the campaign. Investors were interested in it, but only this spring were they really starting to warm up to hardware.

It was still pretty icy cold at the end of last year despite us, Pebble, Ouya, SmartThings and Oculus. A few people put together big angel rounds, but VC money really didn’t start to come into it until this year. I think Romotive raised money last year. They’re one of the few big ones that did. In general with the VCs they say, “Oh great you’ve satisfy two of the three check boxes, but the other one’s empty” and that checkbox is “can you make it and ship it and produce it? Is your team qualified for that?”

I think that’s what you see now with Pebble and a few weeks ago with Ouya: when people deliver on the product and customers start receiving the boxes and stuff, the VC community is warming up to that. You’ve delivered. You’ve gone through making stuff. You’ve gone through the craft of having to deal with contract manufacturers and shipping and fulfillment and all that other stuff and so you can make something.

Yes, they definitely received it better, but I’m amazed at how skeptical a lot of people are about connected devices and ubiquitous computing and, will it ever happen?

What you’re doing has been possible for quite some time but only recently has the connected home really moved forward with entrants like yourself and Nest. Why do you think it took so long?

There are several factors. One of them was the recession. The recession impacted traditional Fortune 500 and hardware companies. By hardware, I don’t mean electronics computer hardware, I mean building hardware. The Ingersoll Rands of the world. These companies that make thermostats and door locks and so forth. Because the housing market collapsed, a lot of those folks cut off spending on the research and development towards these new types of things. They were sort of aware of it and they were doing a few things, but they were really far behind the curve and I think today they’re still trying to catch up with that.

The other aspect of it was at the same time this massive rise of the smartphone. We sort of started to get a feel for that in 2009. That’s when we did the smoke detector, but even then it was an internet detected smoke detector as opposed to a smartphone one.

The more and more that we saw people using Lockitron, we saw that it was actually a rare intersection of where it’s both a very heavy mobile device and a very valuable internet connecting device. It doesn’t work really well when you just do one or the other. The original problem was just internet connected. I think if you just do it mobile connected offline you lose half the utility.

You’re getting relatively close to shipping the batch. You’ve been traveling a lot to Asia. How did you pick the manufacturing and the process? How are you managing that relationship?

Manufacturing is a slog and there’s no way to get around that. We thought that we would be able to deal with it with U.S.-based suppliers so we looked very closely at manufacturing all of it here in the U.S., but there’s a couple snags with that. Those are, we have metal in our product, as opposed to just plastic. We have a lot of components that need to be assembled together. The challenge in America is that you can find a lot of different suppliers, but finding contract manufacturers who bring everything together is tough. On top of that, the labor costs of assembling things are pretty high. We were just looking at all the different numbers we were getting back state side and even in the quantities we were looking at, the prices were almost all of what we were charging for the device.

It was very, very expensive. We really wanted to go that route because dealing with China everybody says, “It’s hard,” and it is hard. It’s very much about relationships and managing those relationships and quality control and things like that. As much as I still would like to do it state side entirely, we are not savvy enough yet I think to find the right suppliers with just the right mix of what we need to produce our product.

“I would say that even though it's in our name, we're not a lock company. We're much more about knowing who you are relative to the device. We see it as social hardware. We think that's going to be one of the main ways people interact with their devices.”

How important has the API development and the partnerships created with API been to the product itself?

It’s very important to us. It’s going to be very, very important going forward. We’ve grown up in an area of Macbooks and iPhones. We know how locked down they are. We thought, “Gosh, this product in essence, you’re doing a binary action, it’s very simple, why don’t we just let people control that programmatically?”

Then, this history of us thinking about the world from an API perspective and way, way back when being able to have software interact with businesses. It was pretty natural that we were like, yes we need an API at some point to have software interact with Lockitron.

We actually built out the API last fall and we have the V1 out today with virtual devices that people can play with. Very quickly we saw, “Oh gosh, if we offer this, we can do a lot of other cool things with people like Pebble and this and that.”

Where do you see the company being five years from now? Where do you want to take it?

The big thing for us is that we’re very focused on this whole idea of access and knowing who the person is and letting them in. When we shifted to the new version of the product, bluetooth low energy was a massive part of it for us. That actually dates back to the end of 2011 when we were thinking about this new product. At the end of 2011, the iPhone 4S is announced. I was thinking about it and then we saw something in there and we’re like oh, they have this new bluetooth low energy technology in the iPhone 4S that’s really interesting. That would be amazing in Lockitron because we could do some really, really cool stuff with that. So I immediately went out and upgraded my phone to an iPhone 4S and we got the development kits in and we started hacking on it back at the end of 2011.

We found that you could do some really cool stuff with it like walk up to the door and the door knows you’re there and unlocks for you. That sort of kernel which we’ve been working on over the last year and we’ll be pushing out in the product now is a very important piece to us. You walk up to Lockitron, it knows who you are and unlocks for you or it doesn’t unlock for you. Lockitron is a very minor device in actuality. That whole identity piece of hardware, knowing who you are is very interesting to us.

I can’t speak specifically to what that looks like in terms of products. That’s essentially where we’re heading in the next five years. I would say that even though it’s in our name, we’re not a lock company. We’re much more about knowing who you are relative to the device. We see it as social hardware. We think that’s going to be one of the main ways people interact with their devices.

What’s your current state of mind?

Excitement and impatience.

What is your greatest fear?

Bad customer service.

When and where were you happiest?

Sometime around the end of college in Boulder, Colorado when I decided to build a company rather than look for a job in investment banking. It was a rare intersection of bliss, fear and confidence that I have felt everyday since we committed ourselves to building Apigy and ultimately Lockitron.

What is your favourite book?

Marshall Berman's All That Is Solid Melts Into Air.

What is your greatest extravagance?

Sleep. While I'm fine with six, I prefer eight.

Which words or phrases do you most overuse?

Holistic, notably when describing the hardware and software elements that comprise Lockitron.

Which talent would you most like to have?

To be a great storyteller.

What is your idea of misery?

Retirement. I believe that the concept of planned obsolescence for people is terrible for all but physically demanding jobs.

Which living person do you most admire?

Jeff Bezos, Elon Musk and Warren Buffett are all candidates.

What’s your favourite quote?

"Aim high, time flies."

Ariel from InteraXon asked, if you were talking to a young start-up founder and wanted to get into hardware, what would be the first steps they should take to get started?

Young start up founder who wanted to get into hardware? It’s really interesting how many of us have grown up with Arduino. I only recently, as in the last three years, have learned how to play with Arduino, but it’s been around for almost ten years now. It was a very small community eight, nine years ago. The Arduino community is really, really, powerful. You can easily bang out commands to actuate stuff hooked in through wires. I think that Arduino is an awesome starting point for someone who wants to do hardware.

I would also say try to be agnostic of technology: see something that you want to do in the physical world. See something you want to manipulate or learn from or gather data from. Is there a certain thing in your life that you want to automate or is there a certain thing you want to learn about when you’re not around? That’s a really interesting starting point, because if you start from that you can back into the technology. I think that anyone who’s willing to work at it can use any of the technologies available to put together a prototype which achieves that.

Raspberry Pi is now a cool option on top of Arduino, but really understanding what it is that you want to do physically in the world. I’m saying this very much so from a software background standpoint. I think that software’s a really, really big part of that. There’s certainly hardware applications where the software’s are very, very thin layers across that, but I’m of course biased. I think that the software should be a very fat layer and let you enable a lot of stuff.

Where software meets the real world, think about the cool things you can build there first and then worry about the technologies later.

What question would you pose to our next interviewee?

Why do you get up every morning to do what you do?

The Blueprint talks to