Into the Forge Podcast: Synthesis Questions Seven and Eight
As we get ready to launch our third season of “Into the Forge,” the Lemnos podcast, we are taking a moment to look back at the wisdom shared from the first two seasons. We ask each founder the same 10 questions, and their answers shine a light on the diversity, frustrations, and joys of the hardware startup experience.
Question Seven – What has been the most surprising thing about this process of bringing your idea to the market?
Question Eight – What’s the hardest part of being a hardware startup founder?
In each Into The Forge podcast, we ask each founder about the most surprising and/or hardest part of being a hardware startup founder. As you can imagine, the answers are varied as each team faces different market and product challenges, but one theme emerged in the podcasts.
Scaling your business takes more time than your already conservatively estimate.
The surprise and difficulty of scaling a business was unexpected even for the more experienced founders. Built, Fieldvision, Privacy Labs, FarmWise, Marble, and Sproutling’s founders all focused on this challenge. Noah Ready-Campbell had successfully scaled a software startup before starting Built Robotics, but still commented on how daunting scaling a hardware startup is.
The Privacy Labs (now Helm) founders, also experienced software entrepreneurs, commented that iterations, no matter how small, take more time to realize than they expected. The pace of hardware development does not always match the speed of software innovation, and when mixing the two streams, things get even more complicated. The FarmWise and Revolve founders commented on the importance and difficulty of mixing hardware and software efforts, and the importance of keeping the two efforts fully integrated through the development cycle.
Hiring was mentioned by a few founders, including Seriforge and Local Motion, as the hardest part of the job, and that challenge has only gotten worse since these podcasts were recorded. Managing cash flow as a young startup was an interesting challenge exposed by Flybrix, especially if you haven’t raised significant outside capital. It’s also interesting to read Nima and Elroy Air’s comments about the challenges in taking care of yourself as a founder, something that we’ll dive into more detail in with our next question synthesis!
Below you’ll find the origin stories from the founders interviewed for Into The Forge. You see many different paths explained throughout the first two seasons of the podcast. Enjoy, and get ready for more stories in Season Three, coming soon!
From Season Two:
Noah Ready-Campbell, Built Robotics
There’s a lot to learn in construction. I’ve been making all the dumb mistakes and learning a lot. I would say that even though I saw the high-level market numbers, going on to some of these bigger construction sites, the scale is really emotionally impressive. A few weeks ago, I was on a high-rise foundation on the Peninsula in Sunnyvale. Just this giant pit in the ground—50-foot deep pit, 300 feet long, 100 feet wide.
Honestly, maybe it’s the logistics of finding office space. We’ve got a five-ton machine that’s driving around autonomously. Being able to be close to your hardware, it reduces your cycle time for iterating. I think that’s tremendously important. It’s one of the reasons we’ve been able to move pretty fast.
Kavodel Ohiomoba and Andy Kittler, FieldVision
Andy: There’s always going to be a lot of surprises, and things like technical surprises are just a part of the game. I think learning about what types of things come up from a hardware perspective was new for Kav and me. We kind of knew what a software development roadmap and timeline looked like, but you run through an entirely different set of constraints and parameters when you’re dealing with hardware.
Kav: I think the most surprising thing is how much is involved in bringing a hardware product to market. That may sound a little naive, but the amount of information you have to process, capture, keep in mind, and keep organized, especially if you’re building a team out, distributing units, and thinking about inventory—all things that I hadn’t thought about before—those were the most surprising things.
Andy: I think all entrepreneurs are optimists to a certain degree and they tend to want to say, “We can do this and we can this in X amount of months.” When you’re dealing with a new field, which was hardware for us, there was a learning curve in realizing that the timelines were going to be extended because of the nature of hardware development.
Jon Hollander and Eric Gregory, Seriforge
Eric: I can give you a positive one. When we made our very first sample parts and we didn’t have any in-house testing capability, we had to send it off to a customer. Our customer was going test something before we could, and it tested really well.
Jon: This is more of a business thing, but when I started this company, we were going to sell machines. We were going be a 3D printer company but for composites or a machine tool company. When we brought on our VP of Business Development, he really pushed us to examine this objectively, to look at our business model and ask, “Is selling machines really the right way to go?” If you asked me that four years ago, I would have been, “Absolutely. There’s no way we’re going do anything different.” Then when we were in business, it’s like wow, selling parts off the machines that you make, as opposed to selling the machine, is a way better business model. Current revenue, lowest barrier to entry. You don’t have to convince someone to buy and invest and learn how to use this incredibly complicated machine. You just give them what they actually want, which are the parts.
Eric: Hiring. We have high standards. It goes back to those values and we filter for that. We’re fairly particular. The hard part isn’t finding people. What’s hard is setting yourself up that way means that you’re inevitably going to be in the situation where you haven’t hit your staffing plan. Now you have candidates you’re interviewing that aren’t meeting the standards you set. You’re now in that position of “Do I compromise my values to hit an objective on paper?” That’s hard, meaning it’s that emotionally hard thing to do, making those decisions.
Having these cash flow crises happen, scrambling, and getting the money you need to do production. However, actually navigating it was a good growth experience. There’s wisdom gained there to tell other founders to stand by for angel investor credit, to have that built into your assumptions. The other surprise was just how well received the product was by the market. I’d never had a product that on launch did hundreds of thousands of dollars a day.
I would say it’s the laws of physics. Physical product actually requires work; you actually have to make something. What that means is you have to reframe what you think about funding—air freight, all these other things that you have to account for. In hardware, especially consumer hardware, your cash flow is limited by how mu31ch credit you can get to make inventory and then move that through a sales pipeline. So it’s this acknowledgement that you are going to be in debt.
Dave Merrill and Clint Cope of Elroy Air
Dave: I think we’ve started to realize that we are becoming known in our industry, even though we feel like we’re pretty early stage. That’s a pleasant surprise. Somebody e-mailed me the other day saying they’d been following us for a long time, and maybe interested in investing. I was boggled by that—how could somebody be following us for a long time—because we had this very low profile web presence.
Dave: There’s a lot of unknowns, and I think for me it’s just kind of self-managing that internal state. I think there’s this amazing blog post that Brad Feld wrote about how every founder carries around, I think he called it, “a bag of despair with them at all times.” Even if you talk to a founder, and it sounds great, and they’re doing awesome and they’re killing it, it’s guaranteed that there’s something that’s worrying them that they’re not talking about.
Clint: I’d say that it’s probably self-care. I think sometimes you are going so hard at a problem that sections of just being a basic human start to break down. So I think it is important to remind yourself that if you don’t take care of yourself, not only do you suffer, but also the company suffers.
Sébastien Boyer and Thomas Palomares of FarmWise
Sébastien: One of the most surprising things when we started was to see how many people were willing to help us, like potential customers from the farming industry that were interested in getting a product like that. But also people outside of this field. That kind of was surprising at least to me, because when you start your first company, you can feel like you’re jumping alone into the very unknown area. But the first thing that you realize quickly, is that you’re actually not alone, and if you believe in your vision, you can get a lot of people to help you.
Thomas: For me the most surprising was how naïve we were and how easy we thought the challenges would be. There has been a lot of realization, but I’m very happy that we’re naïve and optimistic in the beginning, because a lot of stuff we would never have dared to try. Probably they just ended up being more difficult, but we always managed to make them happen.
Sébastien: I guess one of the key differences between hardware and software startups is the time that it takes to grow, both to scale revenues and number of products. So matching investor’s expectations with what you can actually achieve in six months is one very challenging thing. It’s feasible for a software startup to have 500% growth month over month. It’s quite challenging to have that for hardware startup.
Thomas: I would say also that with hardware or a startup that combines hardware and software it’s always pretty hard for everyone on the team to understand the challenges of the other people on the team, like the software understanding the hardware or hardware understanding the software. Every hardware decision we’ve made is going to impact the way we do software and vice-versa. It is really important to have everyone aware of what’s happening. That was fairly easy when we were four people, when we could have everyone present at most of the meetings. It starts to get harder as we get bigger.
Josh Ouellette and Randy Shults of RavenOps
Josh: The most surprising thing for me has been the fact that this idea of swarm robotics and getting multiple robots to coordinate on tasks has been around for as long as it has, but how little of it actually exists out in the real world. We’ve got some stuff going on in defense, a few people nibbling around the edges, some pretty light shows, but that’s kind of it. It keeps surprising me, particularly given the business advantages for us and the engineering advantages around not having to build one robot that does everything. We can use lots of simple robots. We can buy other people’s robots and integrate them into this other system. And then from the customer perspective, the real benefit is around speed. Ten robots will always be faster than one robot, and when something needs to be done quickly, this is an obvious idea.
Randy: I think it’s this mismatch of a lot of mechanical engineering talent centered around NASA, centered around the Houston area, and then a lot of self-driving car talent located out here in the Bay area, and nobody’s crossed that divide yet.
Randy: So going back to the Web 2.0 again, the testing iteration cycle. I used to complain and think it was a really hard thing when I had eighteen different models of Android phones laid out on a desk, all plugged into our continuous integration testing suite. But now we’re looking at like, “How do I, in San Francisco, find the real estate to host a 20-foot diameter by 20-foot tall carbon steel tank to actually test our drones with it?” And that’s like the smallest that we could envision because some of the ones that we’re going to work in in the real world are the size of a football field. So by far, the hardest thing is that testing iteration cycle where we have to plan trips to Houston or to the North Bay just to take our systems into the field.
Sankarshan Murthy, Prahlad Athreya, and Garrett Rayner of Bumblebee Spaces
Prahlad: One of the surprising things that I have learned as we’ve gone through the process of pitching our idea to various people is that a lot of people seem to want to invest in incremental known ideas. When something is fundamentally new, you actually have to do quite a bit of work. We almost feel like we’re creating a new category. We feel like we have to do quite a bit of work to find the right person to move forward.
Sankarshan: For me, the biggest surprise has been who finds value in space. Even though people are struggling day-to-day, paying half their income monthly, residents don’t value space. The people who value space are the people who are selling the space: the builders, the developers, and the operators of real estate. That switch in seeing who values that most and how it changes our business model significantly [has been surprising].
Prahlad: I think the hardest thing is you know how Reed Hoffman says, “Ship even if you’re embarrassed about it.” I’m misquoting him, I’m sure. You can’t do that with hardware. You can be embarrassed about it, but the threshold for embarrassment for hardware is much higher than software. With hardware it can actually hurt somebody if we don’t have the right threshold of development done before we ship it. It takes initially more effort to get it out into the world, and to test it.
Garrett: One of the challenges that I’ve had a lot of time to wrestle with, and don’t know that I’ve actually even come to a conclusion on yet, is when do you spend the money on the tool? When do we go from building our stuff out of totally off-the-shelf components to forking up the 10, 20, 50, 100 grand that we might need for a production tool at the scale we would need?
Giri Sreenivas and Dirk Sigurdson of Privacy Labs (now Helm)
Dirk: We outsourced some of our hardware development, we did it through a very reputable partner, and we based the product on something that had already been built. So I had expected that things from a hardware perspective would be very easy. This is a small iteration on what they’ve done in the past, but we did run into hardware electronics problems. It wasn’t something I thought of as being risky, but we encountered it.
Giri: The other thing that has surprised me has been the conversations that I’ve had with people who have heard little bits and pieces of what we’re doing and really want to understand deeply what it is we’re working on. We basically closed our seed almost a year ago, so it’s been surprising given how little we actually put out there about ourselves. It’s encouraging.
Dirk: For me, it’s basically the hardware aspect. We had done a software startup before, but the hardware aspect of this has been exponentially more complex. I still am doing software development, and so it’s been difficult to take on the additional role of managing the hardware. I’m still writing code, I tend to get into solving problems, and I really want to focus on solving that problem. That makes it difficult to juggle all these other things.
Giri: You have to context switch a lot. And you have to be able to do that without much overhead. It’s manageable, but at times can feel overwhelming because many times you’re context switching into areas where you’re learning things for the very first time. And you have to go to advisors or you have to go to peers, peer companies, and say, “Hey, you talked to this test fixture supplier and they gave you this sort of advice, this is what we’re hearing from ours, what should we be thinking about this?” And I think this is one of the benefits of being in the Lemnos portfolio is that there are other hardware startups that we can go talk to, get feedback from, and get advice from. That helps manage the challenges around those context switches.
Matt Delaney, Kevin Peterson, and Jason Calaiaro of Marble
Matt Delaney: We built a lot of robots and tried them out in academia or in a testing zone. But putting robots out in the middle of the city is fascinating to watch how a random person, who never expected a robot to be there, interacts with them. Some of the interactions have been people saying, “Oh, hey, robot, let me help you along this sidewalk.” It’s like, “What?” So I’m delighted by some of the surprises that we get to witness, which are a lot of humanity’s first interaction with robots in everyday life. That’s something that in history books, it hasn’t happened before.
Jason Calaiaro: I think one of the most surprising things for me is how humbling it was. There are so many things that you have ideas about, you have assumptions, you may carry them as convictions, and I watched time and time again the world step up to you and challenge you in ways that you didn’t imagine. I think it’s difficult for anyone to go through this process and not to be humbled by it.
Matt Delaney: Safety issues. We have to think about policy, because robot laws aren’t written yet, and everything is multi-disciplinary. It’s so integrated.
Kevin Peterson: This concept of the lean software as a service startup, where you can just scale it infinitely, is very different than what we do in hardware. Where there’s a capital expenditure, it might cost you tens of thousands of dollars to do an experiment, or more. Also time, because you have to build a prototype, you have to build another prototype, and only after a few of those are you going to get it right. The expectations of speed to get something out to market are different with hardware than they are with software.
From Season One:
One of the hardest things is, with software, I can work all night and have something. I don’t need to rely on any outside vendor. The hard things are dealing with shipping, like this particular chip because I need a board assembled and that chip happens to not be available, but it’s going to be available soon, so getting that. I feel like I’m waiting on components and other things. In a perfect world, I’d be able to just push a button and print out a PCB board and then have all the stuff on it.
Ilya Polyakov of Revolve Robotics
Software. I’m a mechanical engineer by degree and I’m a designer kind of. I’m a hardware guy. Realistically, hardware nowadays is software wrapped in plastic. At Revolve, we have one person helping me on hardware and six contractors working on software. The software stuff has been really challenging, and navigating all the different ecosystems we have to operate within. We do web-ops, we do server ops, we do front-end web development, we do push notification, we work within iOS, and we do Bluetooth. There are just so many components outside of the actual hardware. By comparison, the hardware is the easy part.
I think the hardest part of being a founder is dealing with people saying, “This is completely stupid. It will never work. It’s impossible.” If you’re a founder, it’s basically like disqualifying you as a person. And dealing with that, not taking it personally, and actually saying, “Okay, tell me why it is stupid.” I’ll ask the question seven times, “Why? Why? Why?” because nine out of 10 times, there is something that you can learn. It could be that you are presenting your idea the wrong way, that it is actually a wrong idea, or that you’re talking to the wrong people. Use it to your benefit, learn, and grow from it.
Art Tkachenko of Pantry (now Byte Foods)
To have a narrow focus on exactly what needs to be done, what’s the best thing that could be done for the company. We get inquiries continuously. Narrowing the ones that are really important, the ones that need to be pursued, and being consistent on dropping the ones that are secondary are probably the hardest challenges. I don’t think I’ve got it figured out yet.
The hardest part up until a few weeks ago was knowing that I made a commitment to these 2,500 customers that believed in us so deeply and spent their social capital and their money on us. Up until the moment it ships, you’re not really sure it’s going to ship. We had some big manufacturing problems as early as December, when we were like, “Oh, crap.” There’s that responsibility that I think weighs on you. On the backend, now that we’re shipping, it’s just more expensive to make things better and fix problems—everything costs money. I can’t ship a product without paying for a mailing label. The financial hurdle is a little bit higher.
John Stanfield of Local Motion
Hiring salespeople is the hardest thing that I have done to date. I would much rather solve a complex mechanism or a technical problem. I really had to learn how to hire sales and marketing people because I’ve never done it before. I had to create a way that I could look at sales roles and sales hiring from a technical perspective. We had to come up with a real, concrete solution-based set of questions.
Adam Ellsworth, Founder of 8-Bit Lit
The most difficult moment was actually launching the crowdfunding campaign because it was first time that I’d ever been responsible to thousands of people. It was higher stress than I had experienced before.
Shireen Yates and Scott Sundvor from Nima
Shireen: Turning my head off to not think about the business, especially when I go home and spend time with my husband. It is so obvious when I’m physically sitting there and not totally there.
Scott: Also utilizing the resources that you have available and knowing when to connect with the higher leverage resources you have. As alumni, we have connections to a bunch of just fantastic resources, and through our own networks and the MIT network, we have these great connections. So, it’s figuring out when and how best to use those resources. It is not always as intuitive as we would hope.