Today we cover Part 2 of Hardware Investing with Avidan Ross of Root Ventures. In this segment we address:
- Examples of Trojan Horse hardware businesses that Avidan has invested in
- Many view hardware as very capital intensive, more difficult to commercialize and thus, too risky in an already very risky asset class. What’s your response to this and why do you think that now a good time to be investing in hardware?
- There are clearly some advantages that software companies have over hardware companies. Can you talk about the distinct advantages that hardware-centric startups have over that of software?
- Are there certain hardware business models that you target and do you think it’s possible we will see more SaaS-like, subscription-based models with hardware products in the future?
- How is your approach to dealflow and connecting with founders or potential founds different, considering the areas you’re investing in?
- Any other thoughts on hardware or core science that you’d like to touch on for the listeners before we wrap things up?
1- The Extremes of Consumer Experience
2- Platform In or Platform On
Recently, we’ve spent a lot of time discussing platforms that companies have built which allow others to build upon them. Examples including the iPhone, Slack and Tesla have been cited as brilliant examples of the platform approach, allowing a form of crowdsourced innovation. But Avidan mentioned another example today that we haven’t spoken much about. Not platforms that others are building on but rather platforms that can be built in to others products or services. Here we discussed Uber and how their platform that efficiently connects providers and consumers, is one that can be built in to a number of other businesses. It would not be very surprising to see many companies ‘powered by Uber’ in the future.
3- Hardware, Why Now?
10 years ago it was not possible to innovate in hardware the way it is today. There were a number of factors Avidan mentioned including: Wifi was not ubiquotous, Bluetooth was not ubiquotous, cell providers were charging an arm and a leg to get access, amongst many others.
Now hardware startups, courtesy of innovation like Arduino, can do rapid hardware prototyping w/ cloud connectivity, for thousands of dollars, vs millions. Here Avidan brought up the example of Nest, who did not raise a $750k angel round; they raised $20-30M out of the gate.
Today, hardware innovation has incredible potential and economic viability due to:
* ubiquioutous connectivity
* unlimited processing
* connectivity to other data streams and
* the ability to prototype very cheaply
And, as mentioned in my blog post, Hardware as an Anchor, it can function as the beginning of a long-term, healthy, annuity if the value is there. Which relates to this week’s Tip, which is called:
Tip of the Week: Delightful Trojan Horses
Avidan: Just as we were talking about earlier, does the value of that asset appreciate or, more importantly, are you able to monetize on that consumer day in and day out, month after month, week after week, year after year? And so that exists for consumers, it exists for enterprises, it exists for everyone in between. And for us that’s what we would consider a trojan horse.
Avidan: It looks like you’re receiving a certain product, but in reality there’s something deeper inside. There’s something else that’s going on. So I think a couple really interesting examples of that, you know, there is, there is a company we invested in called #Prynt. And they built a consumer product, they had a fantastic Kickstarter campaign, did $1.6M in their first 30 days. They shipped almost on time, but you know, in Kickstarter world it was very very on time. They shipped in time for Christmas this past year. And what they do is they provided a, the created a polaroid printer for your iPhone. Now that must, it sounds very basic, right? Your phone docks in this thing and you take photos with your phone and it prints out polaroids. But it’s a Trojan Horse. There’s something significantly deeper there. What you’re doing is you’re actually printing out a video. Because just like with your iPhone 6s you take live photos, so where it’s recording a couple seconds before and after you take the photo, the Prynt Case does the exact same thing and it prints, records 7 seconds after the video has been taken. That video is stored in the cloud. And now that Polaroid photo that you’ve taken, when you hand it to anyone, it’s an AR marker, that if they point their phone, they point the app at that photo, the video becomes alive.
Avidan: Now it’s, since this is all audio I can’t really show the demo. I could show you the demo and you’ll go wow!
Nick: It sounds awesome
Avidan: But really what you’re doing at that point is you’re getting people into an ecosystem where they are sharing their videos through tangible products. So like we were talking about earlier swapping business cards with people and why does it still exist. Now you could hand someone a photo that means something to you, but what you’re really doing is handing them access to digital content. Now the trojan horses, you’re also, you’re also then able to share with things other than video. You want to share a spotify playlist, you want to share a digital gift card, you want to share whatever you want to share that’s digital, now that physical tangible asset you’re handing to someone is where your real value is for the customer. On top of it there’s some pretty healthy margins on the paper and some healthy margins on the printer. But that’s a great example of a trojan horse. And I think others just, you know, they look at the, the physical product as a way to hold data. So I think a great example of that is something like Maschine. So Maschine is a computer vision based self checkout system.
Avidan: So, you go to a grocery store, you see that self checkout line staring at you because there’s four people in your line, and you’re like I can do it, I know I can! I can do it!
Nick: And I always screw it up somehow
Avidan: Everybody screws it, you always , and the worst part is like on your fourth or fifth product you’re standing away and you go back, you look at that line that you were in, and the person that was right in front of you just finished, you’re like yeah I could have been someone else could have been scanning this stuff for me. And I think that the problem is like bar codes are a legacy, you know, it’s 50-60 year old technology. And this company has provided a product for retailers and commissaries where you lay down a tray of up to 20 items at a time and it uses an array of high resolution cameras, identifies every product on the plate using computer vision, and it’s an under 5 second transaction time for 20 items.
Nick: Awesome. Is this only at the checkout counter or can you do it in real time as you fill in your cart?
Avidan: Aha. I like your style, man. That is, that’s the beauty of where the future can go, right. Initially you don’t want to change consumer behavior that significantly,
Avidan: So initially you want them to still fill the basket, have the same shopping experience they’ve ever had. But now, when they go to the self checkout line, they don’t actually have to search through a bar code, they just put the items down under the hood, right. So there’s basically a kiosk, and each time you put products down it identifies everything on a computer screen and tells you, ‘great, we know that, we know that, we know that,’ and eventually you’ll get to a point when those same cameras can be miniaturized and put on your shopping cart. And as you put products into your shopping cart, they’re identified and they’re transacted.
Nick: It’s about time
Nick: I think grocery shopping experiences have been crying for disruption for a long time
Avidan: Yeah, I agree, I agree. So it’s that also I think the camera, that’s one of those pieces of hardware that’s improved dramatically over the years. So high resolution cameras. And we can thank the folks at Apple and Google and even Blackberry. Well I don’t know if I would thank Blackberry, their cameras have always kind of sucked. But, but even Nokia, right, like companies that were building smartphones and pushing better, cheaper, smaller, lower cost camera sensors
Avidan: now allows us to have a bank of cameras for 50 bucks or 60 bucks you can have multiple camera sensors. And then again, going back to that process in the MacCloud, you have unlimited compute cycles in the cloud, so you push all those images up to the cloud. And then, you know, you roll up a huge instance of Amazon or TensorFlow or any of the sort of computer vision or all networks that are being put out there to identify products and objects. And you can do object recognition nearly instantly with an extremely low cost physical piece of hardware. That to me is, is I, I love, on a computer vision front I think that’s opening up so many, so many interesting use cases.
Nick: That’s awesome. Yeah, I mean it’s making me think, I used to work for a robotics company. We did some wafer handling
Avidan: Which one?
Nick: It’s called # Motion Engineering Incorporated. It was a division of a conglomerate called # Danaher
Nick: But they were based in Santa Barbara. So that was sort of my tour of duty there. But we were on this vertical called ASRS, Automated Storage & Retrieval System. So imagine high speed cantilever systems at an Amazon warehouse, running down an aisle, going out and picking, you know, a box, or picking a product and then running back down the aisle to deliver it. So these are really, you know, heavy loads out the ends of this cantilevered arm. And we were able to sprinkle accelerometers around the end of these arms to do much better motion control, to avoid swaying and vibration on these arms. And the only reason we were allowed to design these really effective systems was because of Nintendo and them allowing accelerometers to become really cheap. Because they sold all these wee hand remotes which brought down the price of accelerometers and allowed us to design a system that previously we couldn’t because we could control the motion from the motors and the drives at the source.
Avidan:Absolutely. Those sensors that they, the ubiquity of these ultra low cost sensors makes for a very interesting, very very interesting use cases. We, we have another company called # Shaper. And this one is for the, are you, have you ever got into any woodworking or building like
Nick: Yes. Yeah I have built a few things, a fence in my back yard, I built a shuttle board table in the basement
Avidan: Ah, nice
Avidan: So there’s a, you know, a world of people who want to build more. It’s like, so I’m a big believer in this democratizing of access, right. The idea that if I want to build a crib for my kid, I should be able to build a crib for my kid. I don’t mean assembling something for my kid. But I should be able to go buy lumbar at Home Depot, bring it home and easily follow instructions online that allow me to build something fantastic.
Avidan: And the hard part is, is that experience is actually really essential in the woodworking space because you need to be able to build fixtures and jigs and measure three times before you cut. And what these guys have done is they’ve used a lot of technology that has become available or cheap in the last 10 years to create auto-correct for woodworking. Or auto-correct for cutting or any form of power tool. So basically using a combination of accelerometers and computer vision, as you move the tool around the surface, so let’s say I was using a router and I wanted to cut a perfect circle, as I move that router around that circle, if I’m too far to the left the router bit inside compensates for my error and moves to the right. So as I move the tool around, I need to provide just general 1 inch margin of error and the tool provides two thousand. So I can basically just buy a slab of wood and say I want to build a desk or a chair, whatever I want to build, and the tool downloads the instructions off the internet and as I move the tool around it creates a perfect cut.
Avidan: That was mind blowing. I met these guys while they were at MIT working on their PhD and I was just floored. And like that to me is that’s, that’s what gets me excited everyday about doing what I do is that people now, you know, computer scientists for years have been able to like build a startup. They were like yeah I have this great idea for an apple, let me show you it. But now mechanical engineers, electrical engineers, industrial engineers, all this amazing engineering talent is now able to create product. And these guys were able to build a prototype of this tool with very very little funding. I mean, no funding it was just like with the stuff that was hanging around in their lab at school they were able to build this. And that’s, that’s a pretty significant change, that’s a paradigm shift.
Nick: So I wanted to ask you about this. So why has that changed? You know, fundamentally people have always looked at hardware and core science startups as being extremely capital intensive, sometimes too risky for the venture capital world. What has changed so much that allows these hardware startups to achieve, you know, product design, maybe even finishe concepts on limited capital?
Avidan: So I think it’s two fold. I think I’ll get elephant out of the room pretty quickly, which is software became much more expensive, right. People building an enterprise SaaS company used to be a cheap proposition but nowadays it’s not so capital efficient. You still need to build up a pretty big sales team that runs around with a deep thirst for whisky and stake and a American Express Card and they just run around and try to get you business. So, so the sort of idea that software is cheap, yes it’s true it’s cheaper than hardware. But hardware you can get to revenue much quicker and on a relative basis at probably the same cost. I think what makes hardware interesting from an investor perspective is the fact that the opportunity can be massive, the fact that the opportunity is not just about the margin business, right. People, I’m not interested in somebody who’s building a product that is going to be sold and has no connectivity. That is just a margin business that someone’s going to try and sell every December, you know, and hopefully get in the black every December and make good money. But that’s a great business for the people who are building it. However I don’t believe that it’s a venture bankable business. I don’t believe that it can reach a venture style returns. So what we are focused on are these companies that have the opportunity to be explosive in their growth, and that growth comes from network effects, that growth comes from the data and analytics that come with the product, and the trojan horses of the hardware space. I think the other piece of why it became, you know, interesting for investors to get involved is a couple key successes, right. You have # Nest, you have # Fitbit, you have # GoPro. Those have great successes and the ability to point to something and say it’s not just about that initial margin sale. There is an ongoing people buy into the # Fitbit ecosystem and they’ll be renewing their products, and they’ll just buy new products as they come out. And they compete with their friends and family on # Fitbit on the platform. So I think that it still has a long way to go. I think the vast majority of funding is still heading towards software where you, you know, you can get, if you hit it right in software, it’s not super expensive to scale up the business, right? Despite the enterprise SaaS stuff being expensive, a lot of the consumer software stuff,you can raise an A round, a B round and then be done. Whereas in hardware to scale up and manufacture a ton of hardware, you are going to raise twenty/thirty/fifty/a hundred million dollars
Avidan: But a lot of that stops being venture capital money. Eventually you get to a point where you are going to far more traditional investors who have done asset backlending before. Where you say I have a fact or I have a purchase order from Amazon. Or I have a purchase order from Target, Best Buy or Walmart. And they will provide you far less the loot of capital. So investors look at it and say this is a very interesting opportunity where you can get a lot of stuff done with 2 to 5 million dollars. You can prove out what size of market you have. And then the most exciting part about it is people pay for hardware. Consumers do not expect to get a physical product for free. Which means you can actually test out a consumer’s willingness to spend on your product from Day 1, right?
Avidan: It’s like the, it’s the exact opposite. Like everyone was like oh Twitter, Foursquare, Facebook, amazing products, people I’m sure people will pay for it. I’m sure they’ll pay. You tube, I’m sure they’ll pay for it. Look at how much they love it. And but that’s an experiment, right. Whereas if you wanted to find out if people would pay for #Nest, all you had to is check the receipt because they already did pay for #Nest, right? Now the question is are they willing to pay for #Nest on an ongoing basis? And if you look at attachment rates for a thing like Dropcam, they’re mind blowing, right?
Avidan: Like they’re mind blowing how much people are willing to pay a monthly service fee for a product they’ve already paid for. But because people physically can feel their products, they believe that they have inherent value. There’s an inherent value in holding a product. And I think with software, there is, there is a belief that you have to prove to me the value of it first before I’m willing to pay for it. Whereas in hardware it’s actually the reverse. You get the opportunity to prove after some things paid. Which makes it also a dicey proposition. You better make sure your product is pretty solid to begin with because, you know, you don’t have you don’t really get second chances although every once in a while someone gets a second chance.
Nick: Yeah, I mean, one of the misconceptions I’ve dealt with is people think that the end of the value exchange happens at sale of the hardware product, which couldn’t be further from the truth. I mean, in my experience I’ve sold, I’ve developed and sold razor razorblade products, so you’ve got a consumable strain. And I’ve also had a SaaS follow on product. So you sell the hardware, you’re embedded and then you have some SaaS information sharing product or what not that they’re paying on an ongoing basis month to month. And so it’s almost like hardware plause. And there’s a stickiness component, too, right? Somebody buys the hardware, they can’t just switch it out to the degree that they would with software. They understand the user experience, they can embed it with a product, they enjoy it and the solution costs are just high
Avidan: And they might even modify their infrastructure to support that piece of hardware, right. We have a company called #Skycatch that built drones for construction sites. And those drones people pay for drone but in reality what they’re paying for on an ongoing basis is the data that the drone is extracting from the work site, right? So you could fly the drone all you want, but what you really are after, you’re not after four propellors in the air. What you’re after is the information that the drone is extracting from the ground. And that is a monthly service fee. That monthly service fee for processing all the pictures, the 3D reconstruction, the pixel maps, everything that’s coming off the ground, that’s where the value comes from. And, you know, there’s a, there’s a very important world of building your business above and beyond. And I think that’s why we’re seeing a lot of entrepreneurs who are in the hardware space who are actually ex-software and SaaS people. Whereas you could have assumed hey maybe the guys from the thermostat world are going to be the ones to build Nest whereas it wasn’t that at all. Because the people in the thermostat world, they think of things in bomb margin get that product out the door and if you could outsource the support of this product and the warranty replacements of the product, just outsource it all, right?
Avidan: The relationship ends actually before the consumer has it. The relationship ends when you ship it to Home Depot. You’re trying to cut yourself out of the equation.
Avidan: Whereas the new paradigm with hardware, entrepreneurs and startups are trying to insert themselves as deep into the relationship and as late into the relationships they can possibly go. They want to own that relationship with the consumer no matter where the consumer bought their product. Whether they bought it at Amazon or Target, they want to say interact with us, log into our site, get credentials through our site, use our app, be a part of our ecosystem. And I think that’s a big paradigm shift. Specially, I mean, when we talk about this razor razorblade model it’s like Target is worried about it because where do you buy the razor blades? Are you buying the razor blades by going back to Target? Are you buying the razorblades through the app that automatically detects that you’ve run out of paper and you need to get a paper replenishment. Or you’re on a subscription service, or you basically say I will pay a certain amount per month, just make sure I never run out of paper, or I never run out of whatever it might be that’s my consumable. That’s what they do on Amazon now right?
Avidan: Auto order set-up. I got a dash button but for some reason something keeps on pressing that button when I know it’s not me. I believe the button feel somewhere by the washing machine and like gets hit by a broom handle and like we have more paper towels that I could ever use, so I filled the garage with paper towels.
Nick: Sweet, I’ll take some off your hands before the end of this
How about deal flow ? Is deal flow and the approach to deal flow different from maybe your approach and the hardware approach than a lot of standard approaches to connecting with founders?
Avidan: Deal flow I think comes from the fact that I have a fantastic group of entrepreneurs that we’ve invested in, and they are very well connected with friends who they know and they’re working on something interesting and new. We look to take meetings long before people are raising money. Because the way that we look at it is anything we can do to provide advice or assistance or just anything we can do to help entrepreneurs trying to build hardware, we’ll do. We’re not going to get into every great deal, you know. We can only do 6 to 8 deals a year. So the way we look at it is we don’t need to be in every great deal. We just need to make sure that every deal we’re in is great. And so we are helping entrepreneurs left and right despite the fact that we’re not investing in all of them. And it’s a lot to do with, you know, the idea of a rising tide raises all ships. The more people that can build great hardware, the better off we all are. And I think the other piece of generating deal flow is the fact that we hang out with engineers, you know. Instead of having an office, we decided to build a makers space, hackers space with digital fabrication tools and a coffee shop. And people would just come and hang out and, you know, geek out on the laser cutters and the 3D printers and cmc mills are build whatever it is they want to build and drink coffee or drink beer, whatever they want to do. And we talk about new products that people want to make. And, you know, we are happy to talk to people at their earliest, earliest, earliest stages of ideation. So that oftentimes means going to meet-ups and going to universities and checking out labs and just talking to people while they’re still working at Apple and still working at Google and still working at Facebook, and they’re thinking about making the move to hardware. Because the one thing that a lot of really smart engineers need is somebody to coach them through what are the pitfalls and what are the, you know, the road bumps and sort of the path to building a hardware company. And anything we can help with, we try.
Nick:Awesome. Any other things related to hardware or hardware based startups that you’d like to share with the listeners?
Avidan: I would say do as much of it yourself as you possibly can feel. It’s a painful process but get in it. And really, you know, focus on something you’re deeply passionate about. Hardware is basically the most open ended term I can possibly think of. Because you’re basically in the same product and then you say what’s product? Physical objects. You’re in the physical product business. So find a product that resonates with you. Find something that you love and you’re deeply passionate about. So that when you’re slugging away at 2 in the morning on a Friday night, you care about it, right? Like don’t just be opportunistic and find some like bluetooth accelerometer and come up with some creative business model to track like people’s arm movements to tell how well they’re brushing their teeth. If you’re not interested in something, you’re not going to stay with it when it’s hard.
Avidan: And one thing is absolutely for sure, building hardware companies is extremely difficult. It is not easy stuff. It’s not for the quick buck. So that’s why I would say focus on something you’re deeply passionate about. And then that ends up resonating with investors because we can tell when you’re passionate about something. And then what we want to do is be as passionate or try to be as passionate as we possibly can about what you’re working on. And really at the end of the day, you want investors who are as excited as you are about what you’re building so that they roll up their sleeves in disproportionate amount to help you out.
Nick: Yeah. There’s a lot more to hardware than I think a lot of people realize. The last product I built, I think the requirement stack was about 6000 line items. And, you know, we had to design packaging, we had to design a case for the device. In addition to all the device stuff, too. So it’s got to be a passion, an obsessive sort of interest.
Avidan: Yup. And then you have to build the entire software stack and then you have to handle supply chain logistics manufacturing, distribution, you know. If it’s a multi parts of IC you have to deal with firmware distribution and it gets crazy. But it’s a full company, that’s the crazy part is with an entire, when you’re building a single product, you’re building a almost a multi-part company that’s extremely multi-disciplinary. So I would say don’t, don’t start on something, and find something you really love because it’s going to take a lot of work.
Nick: # Avidan, if we could address any topic related to startups or venture, what do you think should be addressed and who would you like to hear speak about it?
Avidan: Oh wow. I mean, there’s so many, so many people and so many areas of what I’d love to hear about. I’d love to hear from an entrepreneur who has gone through a high and a low, and hear from their perspective, I think there’s been a lot of conversation recently about CEOs and entrepreneurs making decisions around their late stage fundraising rounds and whether or not they take liquidation preferences or, you know, sort of more interesting terms on their, on their raise. And I think that it’s easy to be, you know, a Monday morning quarterback, right. You’re like after the fact everyone’s sort of passes judgement on these guys and is hypercritical saying oh they, you know, they sold out their employees or they just went after the big money. But I think that in reality it’s a very hard part about raising money as eventually, you know, your job isn’t to raise the money, it’s to actually turn it into value. And people don’t just connect those two things. I think that would be very interesting piece to talk about. Or recruiting. Oh man, recruiting. I would love to talk to , I would love to hear somebody share their experiences just being on a recruiting tear and knowing when do you recruit on your own, when do you bring in externals, how do you go about getting recruiting done for your first ten, twenty, hundred employees. Because really at the end of the day these companies are built by amazing people. And getting those people on board is easier said than done. And it’s really easy for an investor to say oh you just need, you should scale the T from 10 to 30. Like, like all of a sudden there is some like website you click on Uber for employees and they just all show up and they’re all amazing. So I think that recruiting is one of those pieces that is very rarely, because to a certain degree it’s a sales job, right. The idea of selling people on joining your company. And for technical founders everyone’s looking for that sort of automated shortcut to get it done. And recruiting is not an automated process. Recruiting is still dealing with humans. And it ends up showing itself later on. If you’re not very good at recruiting, you’re probably not going to want to bring on a salesperson until too late because sales people are recruiting customers. And so that idea of recruiting human capital or human talent is underrated and really needs more focus I think in the startups space.
Nick: Bet you got some scars from #CIM
Avidan: Yeah. Definitely, definitely a good time recruiting in sales. As an engineer, not my favorite part of the job, but definitely one of the most important.
Nick: So I thought you’re going to recommend that I talk to somebody about coffee. I know about this book that you’re writing about coffee, is this, is this real?
Avidan: Yeah, yeah I mean, I, it’s going to be published next spring. And it basically is it’s a book about the best coffee around the world. Because I take a ton of meetings at coffee shops. And I basically got to a point where I was starting to compile the best independent coffee shops in America and around the world. And that included a road trip I did with my wife where we ended up going to coffee shops across the US. And, you know, I look at it, I don’t, I don’t really like to to take meetings at Starbucks because I want to share that sentiment with an entrepreneur that, you know, a small independent shop that cares a lot about their product, that cares a lot about their physical, the physical good sold. And the experience is, that’s what we’re after, right? We’re after that great experience. So, you know, coffee I think is, there is a lot of hardware in a coffee.
Nick: Well, if you pre-sell it, shoot me a link
Avidan: Yeah I will
Nick: Because I’ll be the first buyer.
Avidan: I’ll kickstart it for you
Nick: Alright. And then, finally to wrap up, what’s the best way for listeners to connect with you?
Avidan: So root.vc r-o-o-t, like super user root, root beer hardware, and I guess roots of a tree. And if you’re Australian I apologize for the vulgar name. But you can also reach us, you can reach me on Twitter at @AvidanRoss or @rootvc
Nick: Awesome. Well, thank you so much for carving out the time. This was a lot of fun and hopefully we can do it again.