E35: Peter Semmelhack, Founder, President & CEO of Bug Labs – Interview

June 28, 2016


This wonderful interview is with Peter Semmelhack, founder and CEO of BugLabs. They provide tools for IoT including messaging and visualization tools. It’s fascinating. He’s also the author “Social Machines: How to Develop Connected Products That Change Customers’ Lives“, which was released in 2013.

We talk about the challenges of hardware-centric startups and what it takes to scale. And also how BugLabs and their new tools dweet.io and freeboard help IoT developers develop and iterate faster.

Peter has a deep background in hardware and it is a treat to learn from his experiences.


David Kruse: Hey everyone. Welcome to another episode of Flyover Labs, and this is Dave Kruse from Madison, Wisconsin. And today we have Peter Semmelhack with us, and he is the Founder and CEO of Bug Labs, which is a pretty sweet name, and they provide tools for the internet of things, including messaging and visualization tools and hopefully I got that right, but Peter will definitely correct me.

So it’s some fascinating stuff they’ve been doing and definitely making the lives of some of the developers easier. And Peter is also the author of Social Machines, which was published in 2013 and we can talk a little bit about that as well. So Peter, definitely thank you for coming on the show today.
Peter Semmelhack: Well, thanks for having me.
David Kruse: And so Bug Labs is based on New York City, which I’m a little jealous of. Where is your office in New York?
Peter Semmelhack: Yes, so we are Manhattan based. So New York City is a bunch of burrows, but we are in Manhattan and we are located, our head quarters located in the upper west side.
David Kruse: Wow! Okay, nice. That sounds like a nice way of lifestyle.
Peter Semmelhack: Around Central Park so.
David Kruse: Oh! Yeah, really, okay. All right. Do you live in Manhattan too or outside?
Peter Semmelhack: Yes, I live not that far from the office, so added benefit.
David Kruse: That is nice. All right, well before we get into Bug Labs do you mind maybe telling us a little bit about your background and then how Bug Labs came to be?
Peter Semmelhack: Sure. So I am a graduate at Brown University in 1987. Went to work for Oracle Incorporation right out of school, which was interesting, because Oracle at the times was probably a $90 million company, so it was a good time to join, yeah, I was a tiny little thing. So it was fun. But I started my first company in the mid-90s. At the time it was when the web was just sort of invented, taking off and so we created a company where we’d do websites and cooperate developments and things for big companies in New York and that worked out pretty well. My second company was a software company.
David Kruse: Who were some of your clients?
Peter Semmelhack: We’ve had big companies like Disney and Time Warner and so it was looking – back then the web was brand new and no one had any idea what HTML was. So if you could go in and show him what you could do, there was lots of business to be had. So there were lots of small companies at the time down in SoHo, which is where we were located. Some of the big companies that came out of there were companies like Razorfish, and Organic and others, things that are – companies that are still around today. Meanwhile, but then I got…
David Kruse: I’m sorry. I was just going to ask, how much did it cost to develop one of those websites? Was it a lot more expensive?
Peter Semmelhack: Oh! Gees. Well, everything was just – there were very few tools like there are now. So everything took you know months, so you know hundreds of thousands of dollars you know, big projects. And so we started noticing that one of the things that was becoming more popular was trying to access information with mobile devices. It sounds sort of funny now, but back then you just had sort of normal cell phones and BlackBerry was starting to become popular and BlackBerry required a special software for companies to do things out of an E-mail and so I started my second company called Antenna Software and we focused specifically on providing applications for mobile devices. And basically enterprise stuff, not consumer friendly things and that company did pretty well. We grew that company to up 500 people and solid to Pegasystems up in Boston, so that was a success.
David Kruse: How long did that take by the time you…
Peter Semmelhack: It took a long time actually, almost eight, nine years.
David Kruse: That’s typical.
Peter Semmelhack: Actually seven years. And what was interesting about that learning experience for me was Antenna was a – when we got started with Antenna, we raised lots of money, we raised like $17 million and you had to do that back them because back in those days you had to spend a lot of money of software licenses for Microsoft, Oracle and others so – and what started to happen after that was, so our software took off and the company was into what I had done, sort of the back of the napkin kind of calculation on that, and I had said, you know I probably could have started this whole company for $10,000.
David Kruse: Really.
Peter Semmelhack: All the software is available for free. And so it was very interesting to watch our open source software just completely transformed the software business, whereas before you just have to raise a lot of venture capital to start a software company. You could not do it for virtually nothing and as a result of that you had an enormous amount of innovation happening in software that we are still benefiting from.

So in 2006 when I started Bug Labs, and 2006, its 10 years this month, the idea was could we bring open source IP, the idea of open source to the hardware world and at the time it was kind of just an idea, radical notion. And so we got started. A couple of other companies that worked sort of the same space at the same time was what’s called Dovknow, which was fairly popular, open source hardware platform today, and a couple of other ones. And so I raised some money round this idea that we’re going to build a modular open source hardware kit. It was like Lego. You could build devices by snapping things together like Lego. And if you go and Google Bug Labs today and you look up the images you will see all of the hardware because that was predominantly what we were know for actually.
David Kruse: I remember that way back then.
Peter Semmelhack: Those were fund days. But it’s hard to remember though, back then raising money as a startup – as a hardware start up in Manhattan made no sense to anybody, and so – but we were able to raise some money because the idea was so crazy and VC’s are good at funding crazy things. And so we got to a market, we launched CES in 2008. We actually won Best of Show, which was kind of neat for a small company.
David Kruse: So what did they do back then, when you first launched, what did the product do exactly?
Peter Semmelhack: Well, the product, it was a base. So we think about roughly the size of an iPhone and it had modules slots on it, so two on the top and two at the bottom. So you could plug different types of functions. So you had modules that were camera modules, GPS module and motion sensors, air qualities, all kinds of sensors we just plug into the slots and by plugging them in you were basically building a little device. The device had WiFi in it, so you could have a wireless connection.

So you could literally build like a telematic solution by putting these things together and that was the idea, and it’s still a good idea. But as it turns out, it’s very complicated products. Because you have lots of new increases, lots of different things that can delay and become obstacles in the production and supply chain. And as a software guy, I had really no appreciation for all the things that could go wrong.

And to make matters worse, we launched the company at probably the worst time in history to launch a company, which was 2008, which was not a good year for anything really. But we survived. I mean we raised a bunch of money, so we had sort of survived through that. But it became clear that in order to really become a successful hardware company we’ve had to raise a lot more money, just because hardware is expensive and to get rights is really expenses and there was very little appetite after that.
David Kruse: Is that because you need, you need the inventory and like you guys create all the moulds for your hardware parts or why is it so much more expenses than let’s say your software.
Peter Semmelhack: Well, this is actually a much longer story that we probably have time for now. Something we call, something I call a two beer stories. But it really has to do with the fact that in hardware, unlike software there is a lot of friction in the system and the friction is everything from procuring the parts, procuring the right parts, getting them on a timely fashion. Building things in a way that is easily manufactured and finding manufactures who will give you the proper terms and they are reliable and so on. There is just a lot of thing that can get in the way of delivering a high quality products. And in hardware as you’d imaging, the bigger your spend, the better your suppliers.

So if go to a supplier and say, I’m going to order 10,000 of something, then that immediately gives me up a selection of suppliers that are size X, and they are usually not very large companies. And so those are not the suppliers that supply Apple and Samsung and those guys. So if you want to get to that level of quality, you got to be ordering hundreds of thousands of devices, right. So if you just start doing the math then you’ll recognize, well, if I’m going to be talking about those kinds of volumes, that’s a lot of money. And if you are going to be making that many things, then you better be able to be support it with marketing and sales people and just cascades.

Unlike the Facebook idea where you can never just put up a website and it will go viral with very little friction. And so from our standpoint we just really did not have a good appreciation of what it was going to take. And so we had to – we had to reorganize the company really to get around our software. So we got out of the hardware business a number of years ago, five years ago.
David Kruse: 2011 or so. So you tried it for what, for well three years, you had the hardware in the market.
Peter Semmelhack: Yes, that’s basically what happened. And so we ended up extracting all the software and all the knowledge. One of our big advantages today I think in the market is that we’ve got all this experience and we know what works and doesn’t work. We have all these scars on our back to sort of help our customers not make the same mistakes and so on. And I think our customers rely on that. And for us it’s been, it’s been a big learning curve frankly and I think the market has also evolved. So back then as I mentioned trying to start a hardware company in Manhattan was almost comically difficult. Today, now there is probably a 100 hardware startups in Manhattan, all being funded by well known VCs.

The other thing that happened in the world of financing was the idea of Crowd funding, which made going after small markets, very idiosyncratic or IOT markets made that possible, because VC’s, most VC’s won’t go after small markets, trying to return on capitals alone. But – so a lot of things sort of transpired to make it a lot easier for you guys. So starting Bug Labs now would be a lot easier than 10 years ago, but all that experience has improved our software, it has improved the services we provide and its improved how we help our customers go to market with products.
David Kruse: How would you start it now if you – with the same idea you had in 2006?
Peter Semmelhack: I’m sorry, what would I do? What’s the question?
David Kruse: How would you start it now if you came off with the idea for Bug Labs with the hardware components, how would you go about the financing?
Peter Semmelhack: I would raise a lot, lot more money. So for example, there is a company Samsara on the West Coast, but its backed by Andreessen Horowitz and their Series A was $25 million. That’s the right way to do it.
David Kruse: Okay, all right.
Peter Semmelhack: Because my feeling about software is you can bootstrap software start-ups, because open store software you pay for people that maybe not going to pay as much because you can get equity and so there is all the costs related to the people. In hardware, you’ve got lots and lots of upfront costs you have to absorb. And because of the uncertainty around hardware development you just, you need people, you know deep pockets, you need to be able to make, as I mentioned you need to be able to make orders and bets and having more money is the way to do it.
David Kruse: And that’s probably why some of these crowd funding campaigns go, you raise a ton of money, but it’s not. Still you never raise $10 million, the most successful ones, but then they need another $30 to actually pull it off, their finding. They should call you off.
Peter Semmelhack: That’s exactly right. While that’s a pretty good – the problem with the crown funding approach is what you just said, which is that it’s easy, how do you do – it’s easier to get that first trench of capital to get started. It’s very hard – unless your enormously successful on that first round, it’s hard to get the follow-on capital, because you continuously need capital. Because you don’t think about hardware, especially in the consumer space, there is an expected refresh rate of like twice year, now. You can’t make a product and then update it three years later, that’s not going to happen in the consumer space. So that just requires a lot of just investment and forward booking, investors who are willing to be there for you and it’s hard to pull it all together.
David Kruse: And do you see – I won’t get into kind of what you are doing at Bug Labs now. Anyway this is potently segway, but do you see kind of like what open source software did for the software world. Do you seem more companies coming online that are trying to make it easier for other companies to come online, a lot less expensive. Of course you have 3D printing which could really help down the road. I mean right now it’s still, it’s not where it needs to be, yes.
Peter Semmelhack: Yes for a trend standpoint I think there are number of things that are conspiring to make it easier. One, it is Open Source IP, so for example as you know the entire board design and everything about it is Open Source, they could use that. There are other open source IP based hardware kits out there, lots of those. And so on the one end you have this idea of open source and maybe just moment, I’m trying to lower the initial costs of hardware design, so that’s one.

Two, there is a recognition by most big players in the market that IOT as a concept is not one market. It’s probably 1,000 small markets and if they are going to play, and make anything out of that, they have to play horizontally. And so big companies like Airelectronics as an example, which historically was really one of the biggest players who would sell chips to other big players has now put in place all these programs to go after smaller IOT makers if you will, in hopes of aggregating smaller number of markets, but then in also hopes of finding sort of the next Steve Jobs, you know. So as they are making their new cool thing, they are in the mix and can stay in the mix of this scale. So that’s starting to happen.

Crowdfunding, I think we mentioned that, and also VC’s, I think we’re starting to recognize that these forces are coming together in such a way that they can in certain ways and in certain verticals envision situations where they do get good return on capital, there have been some interesting exits over the past couple of years that helped people think that way, and last week the I think the whole notion of cloud computing is starting to help because you are offloading what used to be native functions on the device itself are now being handled in the cloud. So the devices themselves could be simpler, they could be cheaper and they can be just more flexible when it comes to offering value to whoever their audience is. So I do think that’s a lot easier today than it was 10 years ago for those reasons.
David Kruse: Okay, got you.
Peter Semmelhack: Lastly, I think there is other thing which we are not seeing yet, but I wouldn’t be surprised if that’s happening and that’s something what we are calling, we are saying that in some ways and some markets there may be an opportunity where the prototype is the product. What I mean by that is that the processes and the products that you were using now to create prototypes and 3D printing is a good example.

In some cases its good enough to actually be the product itself and maybe it’s not a product to turnaround and sell, but it’s a product that you would use yourself internally. So let’s say in the past if you needed a 100 laws, XYZ’s whatever they might be and you would pay ABC Corporation to give those to you. And you pay some premium, some markup because you just have to be like a party.

Today maybe what you can do is actually build those things yourself and do a 3D printing enclosure and so on and get it to your own needs for a fraction of the cost or faster or maybe more customer lag and as I mentioned, I don’t think we are there yet, but it’s not inconceivable that in the next 10 years there is an opportunity for people either themselves, either individuals or enterprise to make their own products if they need to internally for their own use or even from the other side maybe actually communities making products for retailers that they can sell.

And some things have to happen otherwise your cost takes them down and expectations actually get adjusted a little bit, but there are I think plenty of examples of that sort of idea working and I for one think that’s a great approach for IOT because as I mentioned IOT is not a market it’s a lots and lots of little markets and people who are in those markets are in the position to innovate in those markets.
David Kruse: That makes sense. And so let’s go back to Bug Labs in 2011. You guys have been doing hardware for three years and now software. So you’ve got lot of a knowledge in the IP, let’s call it and so how do you kind of restructure. Did you have to go find new clients, could you use the same – go to the same clients or how did that all transfer?
Peter Semmelhack: Well, as it turns out, all of our customers really were – First of all, right around 2010 we stretched from focusing on consumer, that’s where we are know. We are just really focusing on the early adopter products in the market. We switched to enterprise, so we went to really talking to product developers themselves, not armatures. And so as we found when they were starting to use the kit they were always building connected devices. They weren’t building devices that would sit in isolation. They were always building devices that would connect to the internet to do things. So in many ways those were internet of things, types of engagements. If you didn’t know what’s the problem, it could take your time.

And so it because clear is that the device itself was really just a collection of IO. It was a bunch of sensors connected to a board and the board was connected to the internet and most of the value, business value had to do with the application and the software that was running off the device itself. So it actually wasn’t as hard as it might seem to say, alright well look. We are going to get off the hardware business and it wasn’t so hard at that point to actually swap in other hardware that’s providing the same functions and keep the valuable components ahead to do our software still within the organization.

So we have built a product called Swarm that was really a cloud based system that would allow you to publish and subscribe to information from devices that were in that particular Swarm. It would do commissions management, account management and so on. Device management, and so we had taken everything we had learnt as a whole systems provider and just swapped out the hardware, because it seemed to be the least valuable component anyway and just keep going from there.

So from there we started selling Swarm to big companies like Comcast and Ford was a big customer then. So and since that point, that’s what we were really focused on, we totally rewrote our platform in 2014 and now we have two products, one is called Dweet and Freeboard which have been, sort have been huge successes for us with respect to usage, because lots of customers, tens of thousands of users.

So for us that has been just this gradual leaning process as to how do you continually focus on making the products simpler to use, easier to use, more affordable and allow those who are in these markets, as I mentioned fragmented small markets, allowing them to explore new ideas quickly and easily using whatever hardware they want, to see if it’s for them and the next step is to provide easy pathways to go from super proof of concept to prototype two and potentially tens of thousands of devices connected, and then for us right now, that’s really what our core focus is.
David Kruse: Interesting and can you give an example or can you tell our audience what is, Dweet and Freeboard and Dweet is a sweet name I’ll give you that, haven’t really thought of that. That’s good. But yes, can you give an example of how developers are using both those platforms.
Peter Semmelhack: Sure. Well Dweet, I mean obviously the name spells like Tweet and so far we haven’t had any phone calls from Twitter guys, but the way – the reason we came up with that name is that we consider the twitter for devices right. So with twitter here you go to platform you sign up, you give yourself a unique ID, so then you can tell the world what’s on your mind and I can subscribe to what you have to say and you can subscribe me and we follow each other in that aspect.

Dweet is the same concept, but it’s for devices. So now a device can go to the Dweet platform and say, okay look, I am a device and I have temperature, I have acceleration, and I have humidity. I want to publish that information on this platform. So Dweet takes that information and instead of the device having to sign up and they wait for a name, the Dweet platform actually gives the auto sign to the same. And then from that point forwarded, any buddy or anything can follow that device on the internet.

So if your take your phone as an example of you and your Smartphone you go to Dweet.io and go to into that website and you see there is button on there, perking interest saying ‘try it now.’ If you click that try it now button on your phone, what will happen is you will see a white screen come up and on that white screen you will see some numbers and the numbers that you see initially are called tiltX, tiltY, tiltZ and that’s an accelerometer on your phone and the accelerometer in a telephone basically keep track of its position in 3D space. So if you move your phone around and you check, you check those numbers those serial numbers change.

Now what’s happening with Dweet thought is that above that box you’ll see there is a hyphenated name and its usually some funny combination like my phone is wasteful rout. Those are the things called mute apparatus, it’s just some funny names. But that’s a name that’s auto signed by Dweet to the device. So now what’s happening is that your phone is Dweeting this information under that name on the Dweet platform.

So now you or someday can really watch the data if you want. And then what happens is another device will come in and say I’m going to follow that device, because you do the information, do you something for an application. So the metaphor around Dweet is this idea of publishing some scribe, but it’s built around an analog that people understand which is ID following. And so that’s been a successful direction for us in that and so it’s fun to get started but it’s also a very high performance in sales of the hundreds of millions of transactions. So it’s very easy for an end, but a very powerful platform.
David Kruse: What type of – yes, that sounds quite slick. And what type of protocol do the machines talk to each other in?
Peter Semmelhack: There is two right now, ones HTTP and the other is MQTT.
David Kruse: Got you and so definitely familiar with HTTP and what was the other one?
Peter Semmelhack: MQTT.
David Kruse: Okay. Is that more of a hardware protocol or is that the internet too?
Peter Semmelhack: Yes, no its definitely – it was designed by IBM. It’s sort of low latency, low – it’s very concise and so on.
David Kruse: So any machine that sends out a messaging that those protocols, you can take in and then push out to whoever is asking for it or following that machine.
Peter Semmelhack: That’s right, yes.
David Kruse: Wow, that’s slick; interesting. And what about the Freeboard. What’s the example of…
Peter Semmelhack: Freeboard a consumer of the data. That’s a visualization tool. So you can use it build dashboards, applications and other things. So you would have the information coming from Dweet and then Freeboard would be the way that you could build beautiful visualizations of the data. So you could put up graphs and gauges and charts and so on that’s sort of showing in real time the data that Dweet is presenting. And so they basically work hand in glove to build ILTS and so if you think about it, the websites make it easy to use, but our other line of business that we have companies take these products and wrap their own brand around them and use them for your products, for example, Verizon last December launched a portal called ThingSpace and ThingSpace is a developer’s tool to let them – let developers use the Verizon network services more easily.

And they took our platforms to Freeboard and they basically white labeled them under the Verizon brand or the ThingSpace brand and I know promoting and marketing that portal with our technology built in. And we have a number of companies who are doing that now. Ford is trying to do it, Renaissance is starting to do it and more will follow, but the idea here is that every company who makes a thing or has some service that isn’t interesting or usefully for IOT, now need to provide easy ways for people to use them and so that’s another area that we’ve specialize in.
David Kruse: So how is their architecture setup. I mean you much have. I mean you are going to have a huge number of transactions at some point. Is that all going through your service and platforms?
Peter Semmelhack: Yes, yes, it is, but for every customer we set up separate sort of extenuations of it. So I mean we do ask for data privacy and data security between the customers. So it’s not as if it’s like one giant servicer. We have lots of ways of sustaining it.
David Kruse: No, sure indeed. You’re not sharing data between all the companies. Okay, all right. And what other kinds of products or features do you think the world IOT needs that you want to build.
Peter Semmelhack: Well, you know honestly the biggest – there’s lots of challenges in IOT today. The biggest one is still cost and it’s starting to happen in the world on radio is just good. So IOT, Internet of Things, the real challenges in the word of. We understand the internet, we understand things, but try and put those together, that’s where the complexity lies. And sometimes I tell people that we are in the business of, of – we are trying to make that piece easy.

And so communications is a challenge, WiFi its easy once its setup and it’s a pain to setup. It’s not always reliable. It’s not always super secure. Cellular on the other hand is easy to setup, it’s very secure, but it’s also expensive. But with this new models coming out called CAT 1 modems which will be cheaper, the air time will be cheaper, the data plans will be cheaper, and so that trend will continue. We’ll get cheaper and cheaper modems, cheaper and cheaper data plans that carriers are going to be playing, I think an important role in that transformation going forward.

I think open source has to continue to mature, so that people can really count on it to build much of the prototypes than actually products in the same way about Linux and MySQL and all the other open sources foundations that everyone uses today. The same things needs to start happening in the world of hardware. I think it is starting, but it’s probably another 10 years before that’s business as usual.

Hardware takes a much longer time than software to become something. Well because again, we are thinking about, as I said early about friction. Its just – it would just take longer. Everything takes longer in the physical world than it does in the digital world. But all these things will happen. I think it’s just a matter of time before the business models catch up and make sense of all the different opportunities, but we are obviously very bullish on it.
David Kruse: Got you. And we are kind of running out of time here. But I’m curious about your book, what – it was published in 2013, is that right?
Peter Semmelhack: That’s right.
David Kruse: Okay and what prompted you to write the book and do you think you’ll write another one?
Peter Semmelhack: Well, what prompted me to write the book was just two things. We were struggling really as a company to try and figure out what does IOT mean. What does ITO – how is ITO different than what it currently existed, which at the time was and still is just called M2M, which is machine to machine, which is a space that’s been around for 20 years, not that exciting, it’s sort of boring. Its serves a function, which is a very important function, but it’s really – if you are going to go invest in M2M software, the ROI typically is around saving money. Again, it’s you go and invest in money because you are going to invest $100,000 because you are going to save $200,000, you know that kind of thing, which is fine.

What we started to see is that internet of things included M2M, but they also had other interesting opportunities and if you think about what I mentioned about earlier about Dweet, were we are using this notion of a follow. That represents a little bit of a social networking type of mentality and we did a number of interesting projects with vending machines, where vending machines would send samples of whatever they are holding, like a beverage or samples.

If you did things like pressed like on a Facebook page and you think about the ROI around that kind of limitation, the RIO is not about saving money at all, it’s about making money, it’s about market awareness and positioning and so on. And so when we came up with the idea of social machines, it was really based on this idea of what happens when machines start participating in social networks.

What happens when machines start following you on Facebook or you start following the machines on Facebook. You could imagine why you might want to follow your teenage son’s car. I wouldn’t even have to tell you the reasons why you want to follow the car right, there is reasons and so when machines start becoming smart enough and responsive enough and secure enough as you can literally bring them into a social graph, interesting possibilities pop-up, and so the book Social Machines was really written from the sample and the things we are seeing and how we thought value could get created in IOT around that sort of vector.
David Kruse: That’s a wonderful vision, because there is always connected items but man as long as it’s hard to connect with them or you need to, they all have their own app, but if you could just print into something that your using and just hey, follow my car, follow my fridge, follow yes – oh man. Make it that easy and I can see where Dweet now is – that’s where it’s probably headed. We’ll that’s where you wanted to headed. Okay.
Peter Semmelhack: That’s where we want to go, yes.
David Kruse: Interesting that’s pretty sweet. How far away think we are from build a follow your car on Facebook or another…
Peter Semmelhack: Well, you can do it today.
David Kruse: You can.
Peter Semmelhack: I mean, well the technology exists. The problem is just, it’s just getting applications that are easy to do and easy to install and so on. But we are not that far, I don’t think we are that far. So say tuned.
David Kruse: Got you and maybe, my last question was I mean that kind of answer, a lot of things in the IOT world, lot of like really interesting cool ideas, but it’s hard to find that ROI, but you may have kind of spoke to that that well always its always about ROI. But I mean have seen any examples where there is – the ROIs really strong that you run across and maybe ROI we have to look at it in different way than just necessarily the economics, could be engagement which leads to better economics later on, but yes.
Peter Semmelhack: Well, I think if you take a subsection of Internet of Things, when you look at it from the standpoint of M2M like I mentioned earlier. I think from the very well established and proven ROIs in that world that have to do with remote monitoring, making sure that a machine really broken before you send a tech, if it’s an inventory issue making sure that you only restock the refrigerator, if its empty. So there are real, there are real numbers you could put against those types of business processes.

The flip side though, which is the making money one, those are just starting to become I think better understood and those have to do – in many cases they have to do with sort of fuzzy metrics which would be things like market awareness and market positioning, that’s not so easy to gauge. But there are things such as promotions within retail stores where they can literally say, you know this technology drove this many sales.

So there’s a funny example. There’s a vending machine in Argentina I believe and it was called the Rugbeer Machine and the idea was because Rugbeer is a big deal down there – the idea is that you would hit the machine as hard as you could and then depending how hard you hit it, you got extra amount of beer and so it became this huge macho thing. So the woman, if she said she wants a beer and some guy to come really smack the machine and get her a big beer, because he’s this macho guy. You can imagine what happens the more people drink, right. So the statistics were something like, any bar that had that machine sold 30% more beer.
David Kruse: What? Are you serious?
Peter Semmelhack: It was like an astonishing increase.
David Kruse: Oh! That’s a good example.
Peter Semmelhack: But there are things like that right, where the machine becomes part of a marketing or sales mixture and drives top line. So like I said, those are all fairly new, but I think again over the next couple of years you’ll see a lot more like that.
David Kruse: Interesting, ultimate right one. I have not heard of that, but that’s pretty brilliant. An extra 30%, that’s a big number for anybody.
Peter Semmelhack: Oh! Yeah, for sure.
David Kruse: All right, well. I think we’ve unfortunately come to the end here, but definitely I appreciate your time here and your thoughts and this is quite interesting and we learned a lot about hardware. That’s not – and numeral software, so this is – for me this is great, learning more from you.
Peter Semmelhack: Oh! Good. I am super happy to have been invited on your show and appreciate it and hopefully it was helpful.
David Kruse: Definitely. And thanks everyone else for listening to another episode of Flyover Labs and this is Dave and Peter signing off. Bye everyone.