Embedded - 42: Blocks of Gold with LCD Displays

Episode Date: March 12, 2014

Christopher White (@stoneymonster) and Elecia talk about the failed startups (and projects) they've been through, focusing on identifying how to discern the end is nigh.  A nice collection of startups introspecting their failure. Wonderful, in-depth Everpix post-mortem. If you liked this episode, try 24: I AM A TOTAL FRAUD.

Welcome to Making Embedded Systems, the show for people who love gadgets. This week, Christopher White will be joining me to discuss failure, in particular, startup failures. Hi, Chris. Thank you for joining me. Yep. It's great to be back, especially talking about something so exciting and fun and uplifting is... Well, I'm really glad you agreed to turn on your mic so that I don't have to discuss only my own failures. I think I have more than you do, so... So we're going to compete as to who can fail the most? Well, I've got a list of five here.

And one of them may not count, but... Gosh, I hope he isn't talking about our business. I'm not predicting the future. The things have been really good lately. Um, the podcast is growing. Thank you. Amp hour and our business is doing reasonably well or a little consulting company and it makes me nervous. Uh, so I wanted to, to talk about failure, um, so that my head doesn't get too big to fit into my hat. Is this like a ward against failure? As long as I repeat its name three times, it won't appear? It's like a reverse Beetlejuice? Well, I tried to look this up, but I couldn't find where I originally heard it. But in Rome, if you came back from a successful military strategy,

everybody insulted you to make it so that you didn't get too confident. And so I was kind of in that vein. That's like saying break a leg for actors. Very much. But when I suggested the idea for this podcast, having a podcast about failure, you were immediately into it. Well, that's because I've had so many and I like to complain. I see. That's true. I do think my focus on the active failure is going to be tempered somewhat

by your idea of how we identify failure. Well, not just how we identify failure, but well, I guess, yeah, how we identify failure. Well, not just how we identify failure, but, well, I guess, yeah, how we identify failure and how you can notice it when it's about to occur, you know, and try to catch it before it happens. Because it's no fun to be riding a company down. My recent robotics failure is kind of one of those cases that's odd. I liked it. Good technology. Excellent application. They even had a reasonable product plan and reasonable amount of money to get what they needed.

But we did a big demo and it was supposed to get funding. It was for investors. I think it actually showed the engineers how far, how very far the system was from being a product. And so I'm just a hired gun. I've been there for a few months. But the three-year veteran felt really, really bad about pouring his life into this neat product that clearly wasn't going to ship on any timescale that he thought. Yeah. Well, that's one of the things too, is your connection to a company and how you feel about it. Because, you know, as a contractor, I've tried to distance myself a little bit. That's one of the reasons to be a contractor.

Well, it's a positive and a negative sometimes, but, uh, you know, not get emotionally attached to, Oh, this company better succeed or else, you know, it's going to reflect badly on me or, you know, or feeling impotent to change what's going on at a company because you are, you know, one of many people who are doing engineering work, not necessarily driving the direction of the company. Um, so yeah, I mean, there is that difference of you're a hired gun versus this guy who's put three years of his life into it. And probably because he believed in it at some point. I think he still does very much so. And I appreciate that. But as the company is starting to fold, he, he was saying that he felt guilty for not sticking it out even further.

And I'm sitting over here laughing like, yeah, there was no way I was going to stick it out more than six months. And I wonder if my lack of perseverance is not good. But I'm up front with it. I mean, I told them in the beginning, you only have me until March, and here we are in March. Well, a lack of perseverance, I mean, it depends on the situation, right? Because if the company is clearly screwed up,

I don't feel like you should be under any obligation to put yourself through stress to try to save a sinking ship. And a lot of times these places are sinking ships and they sink for years. That's always a shock to me is how slowly I, you know, when I first heard about startups, they all seemed like, well, you know, one in 10 fail, but those, i'm sorry one in 10 succeed much different that is much different and and nine fail but you think nine fail like like spectacular flame outs spectacular flame outs or they just one day close the door and you don't get paid and that's it but a lot of them just kind of no it's years and years layoff layoff after layoff, one every three or four months.

And, well, we've got enough money to do this. And if you work for half pay, we'll give you stock. And, you know, in six months it'll get better. We're sure it's going to get better. Usually at least, you know, two or three of the companies that I've been involved with that have quote failed. Uh, it's been, you know, dwindling and dwindling and dwindling until they finally get sold for, you know, a fraction of whatever their max value was. Um, and you know, nobody, nobody does well, except maybe some of the investors who got in very late, you know, get their money out first,

but, uh, yeah, it's's usually not so it's hard because you think well i'm gonna go to the startup i'm gonna work for these guys i'm gonna put my heart into it and at least i'll know that if it fails it'll fail quickly and i'll move on to the next thing but that's not the case you know you sometimes you're just there and you're going well should i leave should i be the first to leave because I leave? Should I be the first to leave? Because it's always hard to be the first to leave. But sometimes when the first person leaves... You feel like you're dooming the product. It's your fault they're going to fail. Right. But sometimes when the first person leaves, that's just, that uncorks it and

everybody else says, well, finally somebody did it for me and now I can safely get out of here without being the first person. That's happened a number of times. Okay, so I want to make sure we can talk about these failures, but not necessarily name names. Yeah, I'm not going to name any names. Although maybe one because it doesn't exist at all anymore. So my robotics company is going to be called G, and I'm going to try to remember that.

Here, I'll put it over here. G equals robotics. And it, it definitely is on the track to failure. And I've already gotten the, well, we're going to go talk to VCs. We're going to get some more money. Will you be around? Will you wait for us? And I'm like, no, no, if I'm available when you're ready, that's fine. But I'm not putting my career on hold waiting for you to get... It's not just your career. You're an hourly contract worker. So it's basically asking you... I'm not putting my lunch money on hold. It's basically asking you to forego income. Yeah.

On their say so. Okay. So what was the first of your failures? The first of my failures, and I'm going to name names, was a company called Procket Networks. Procket Networks. And this was back in 99. Yes. I came out of there from Cisco.

A bunch of people who I'd worked with at Cisco disappeared one day, basically. And then you went and you interviewed, and you saw some posters that you had called from somebody else's office they weren't supposed to be there but um so yeah so i ended up leaving as well from from cisco and going to this startup and it was really cool it was very cool it was uh you know uh for the time they were trying to build some technology that was extremely advanced uh basically you know I characterize it as kind of six space shuttle kind of complexity projects at once. They were doing custom silicon for high-end routers, which nobody was really doing at the time. They were just doing, you know, ASICs, which are custom or FPGAs, but this was on the CPU production model. So these

were full custom down at the cell. So, you know, it's really interesting stuff. I was there, I was employee 19, I think. So I got to be involved with a lot of interesting software, you know, kind of define how the operating system was going to be architected and, you know, write core pieces of all of it. So it was a learning experience. It was fantastic because I'd been at Cisco, I think four years at that point. And I'd kind of moved from junior engineer to, I don't know, mid-level at that point. But this really boosted me from mid-level

to somebody who could do all kinds of things because in that startup environment, you know, it's, well, today you're writing a routing protocol. Oops, today you're looking at the scheduler and the operating system. Oh, we need a shared memory architecture so we can do some quick message passing stuff.

So it was really all over the map in just a couple of years. But they were kind of one of these hyped up stories of the dot-com boom, even though they weren't a dot-com. Oh my goodness, I remember them looking for property and spending a hideous amount in rent, but they were going to lock the rent in for five years. Yeah, which't a.com. Oh my goodness, I remember them looking for property and spending a hideous amount in rent, but they were going to lock the rent in for five years.

Yeah, which is a huge mistake. The next year it just tanked. Yeah. So they took a huge amount of VC funding. They had a huge number of shiny stars in the networking world. Yeah, a lot of big names in the networking world worked there. And I think at the end, they burned through 350 million of VC capital.

And at one point, they were valued at several billion. But they didn't take Cisco's offer to buy them for like one billion or whatever. Well, that was always a rumor. We never knew the truth of that. So yeah, we got through all that. We actually built the damn thing. But it took a lot longer than expected.

It had a lot of problems along the way. And they made some mistakes with the marketing, I think, and how big this was all going to be and who was going to be willing to buy something that wasn't from Cisco. Nobody ever lost their job buying from Cisco. That's been the quote because you don't, it's not going out on a limb, but you're doing something that everybody agrees is probably right.

Yeah. So yeah, long, I think I can't remember when they got bought. I think it was 2004. So they were around for about five years and many, many layoffs after I left. You were there for three and a half.

I think I left in 2002. so yeah, three and a half. And yeah, I think they'd gone through, I think I left around the fifth round of funding, which is kind of my rule for startups. If you see the fifth round of funding, it's probably time to leave. Round E? Yeah, I think they may have been on round F.

So what was the fail there that was a tricky one because uh there were a lot of moving parts there definitely was enough money at some point definitely bit off more than they could chew technology wise it took a lot longer than they expected and a lot of the architecture had to change radically as reality kind of imposed itself well now there's all this talk of mvps and minimum of minimum viable products do you think The architecture had to change radically as reality kind of imposed itself. Well, now there's all this talk of MVPs and minimum viable products. Do you think they didn't, that wasn't a term? No, that wasn't really a term and that wasn't, you know, they had good specifications for what they wanted to build. And there was a lot of good stuff in terms of having requirements and things like that. But I think they spent too much money.

Spent way too much money on people. Spent way too much money on space. Dinner. Dinners, lunches, free stuff for employees, although it was peanuts compared to $350 million. But I think it was a confluence of things where the dot-com boom busted busted yeah and you know the market for core routers kind of went away juniper and cisco were making things that were okay they were good enough yeah so uh

you know it was just a tricky tricky situation that one was tougher to kind of define um but it was late too it was very late i remember we were signing lots of pledges to maintain certain schedules and they had t-shirts that everybody signed and we're going to ship by this date we missed it by we missed it by two years uh so two years late in the company that was around for well two years late from five years late from the date they wanted not two years late from the date they wanted, not two years late from what was really possible. Yeah.

Well, I think that unreasonable schedules is definitely a theme. Although I don't know that that's a theme only in companies that fail. I see a lot of unreasonable schedules at companies that do just fine. It's just completely unreasonable. And at a big company, if you have an unreasonable schedule and you're two years late, there's enough slop to cover it, maybe. At a little company where you're rolling the dice, you're betting your whole VC-funded capital on shipping by this date. Well, being late means you missed. Well, you know, they had some other problems too.

They did another classic mistake, which I've seen in multiple places, where they get so excited that they decide to hire a VP of sales. Really early. Really early, before you have anything to sell. Because people come to you and say, oh, you're making large routers. I'm really interested.

You should come talk to my people. Do you have a sales force? And you feel dumb saying, no, we're just a bunch of engineers. And so you hire a sales force and then they all play solitaire for the next three years. Which is what actually happened.

They hired a massive sales force with no product. And I remember when you came home and said, I walked through there and they were all playing Solitaire. Yeah. As though they were the only approved game. Well, speaks of a certain lack of creativity. Send your emails from salespeople to me, Alicia. Well, I mean, I feel bad for Pruckett

because why would you handicap your salespeople like that? Why would you hire them? Why would you have them? If you were just going to bore them. It's pointless. But there were lots of, just lots of mistakes like that. And then there was lots of political machinations with who's the founder, who's the CEO. You know, founders getting pushed out by some point.

And then they bring in some CEO that they found, you know, from some search who knows nothing about what the company's been doing. And, you know, just lots of classic stuff that, you know, sometimes you just can't make that leap from, I don't know, from an engineering company to a product company. I'm not sure I'm using the right words exactly, but that leap from being able to build and sell something instead of build the first one. Right, right. Well, my robotics company, after a month, I said, you guys aren't building a product, you're building a science project.

And I don't think this is a good idea. This is not what a company should be. You can't build a science project. In fact, I had some people come over for the demo and somebody said, this would be great at Maker Faire. Yes, because it's a science project. It's not a product. And I think that's a huge, huge fail point. Sure, you have to build the first one. You have to build the prototype. But is the prototype where it's going to end? do you see beyond that? do you see the path from here to productization?

so Procket had a little bit of that that they were building one or they were building a small they didn't have that problem exactly one of the biggest problems was the complexity i mean they wanted to build you know it started out as two cpu class things and they were big they were pushing the limits of what was possible in fab technology at that time and you know it's they had to split some of them to get the size down, which meant more communication links, which meant just more complexity, and it just took a lot of time.

And that doesn't count the hideous complexity of networking protocols on their own. I think you can start out a company trying to do one impossible thing. Yeah. But if you try to do three or four or five or six impossible things simultaneously, you're either going to be super late or you're going to just run into the limits of physics. Let's quit bagging on Procket because I think one of their problems

was that they were just unlucky. There was a lot of bad luck involved. If they had started two years earlier, they would have, they might have succeeded. Maybe, but I don't think the CPU process technology was available then, so. don't think the CPU process technology was available then. So what's the next one? Oh, we're just going to make this about me?

The next one. No, I go next. You don't have as many fans. No, I do. The next one was a company that made medical devices. I went there after Procket. And a little bit different. You got a master's degree after Procket. I'm there after Procket. A little bit different. You got a master's degree after Procket.

I'm just talking about companies. It's not like my master's degree failed. No, no. I'm just trying to figure out which company you're talking about. Okay. Medical company. R. Right. And that was fun. That was another departure for me. I got to work on lots of interesting stuff with lasers

and motion control and all kinds of stuff. So that was my first real experience with, you know, kind of embedded systems where you've got lots of little peripherals you're talking to and you have to write drivers all the time and stuff like that. So that was fun, but they couldn't decide what they were doing. Yeah, they had multiple paths and they kept trying to travel both. Yeah. Or they tried to make the product flexible enough that they could decide later what they wanted to do and that was a huge mistake and so that one that one was a case of being late as well uh although for one of the products that we did we actually pretty much hit the schedule but they didn't sell it for like six

months because they couldn't figure out what configuration they wanted from their flexible system. Build me a car that can do anything. It can go on the water. It can go on the road. Oh, now we have a car that can do anything. It can go on the water. It can go on the road. Yeah, I guess it can also fly, but how are we going to sell this? So we had to do some really sophisticated stuff with motion control and optics to get it to be configurable in the way they wanted, which ended up making the timeline very long and also ended up making it so that it was really expensive. Yeah. So the profit margins were lowered by this crazy stuff that we could only buy from like Switzerland and one company.

Flexibility has cost in lots of ways. There's actual physical component cost, which this definitely hit. And then there's cost in SKUs. Then if you have more than one product starting out, I mean, you start going out with 15 products. It gets really complicated. And then there's the software complexity and just the mental complexity of having a product that can be on the road or in the water, maybe fly. And they wasted a lot of money too. There was a ton of rework and manufacturing, quality issues.

And a lot of pie in the sky thinking from upper management like well if you guys can just push this out our stock price will be worth x and you know they'd say push it out in six months and it would be an 18 month project you know you'd go back of the envelope calculate the schedule and it's like there's just no way this is going to take less than 18 months because you know some of the lead times for the parts were six months so uh it's just a lot of crazy magical thinking um and they were another one that uh uh they got sold for a fraction of of their uh vc you know what they capitalized in the vc fund i think was exactly the same as as a rocket actually so that was that was fun to do that twice yeah they actually tried

to go public but failed which is always amusing they went on their road show but the road show told them not to go public basically well let's see i guess it's my turn um worked for a company which i guess i'm and they did a lot of interesting, cool, nifty things, but we were making a generalized product and selling it to specific people. And each specific person wanted their own set of features. They would have been okay with a generalized product. So, for example, let's say we were making a spoon and everybody wanted the spoon to be to the dimensions of their

mouth. If we'd had a good sales team, they could have sold it as a generic spoon and it's better for being generic because it's more robust, it's more stable, and you know you can buy a whole bunch at a cheaper price. But this company just chased down too many paths that didn't have profit. It was a place where actually, somebody actually said to me, we will make it up in volume when I pointed out that we were losing money on every unit. And it was such a Dilbert moment of, oh my God, can you even hear what you're saying?

Yeah, there was some of that too because the product consisted of two pieces there was the console which was big expensive part and then these plastic tips do you want to go back and mark that you said yeah we can leave this part in though i'll bleep that uh they had these plastic tips which were you know i think they were like 20 or 30 bucks to make there was a 11 sapphire window and then some plastic and a counterfeiting chip uh anti-counterfeiting chip a counterfeiting chip would have been for a cooler uh yeah and so they were selling those for like four hundred dollars and so that was where they were trying to make up their money but the sales people just kept giving them away for free

to get people to to use the system so it totally backfired and they they lost a ton of money anyway i'm sorry i interrupted your story about c no that was kind of it you i think one way to have a startup fail is um not sell your product for enough money well it's the idea that you know oh we'll get exposure it's the same thing as like a music industry we'll get exposure we'll we'll put our product out for free even though it's hardware and and we'll see if people eventually start using it enough that we can start charging them. Little known fact, when you start charging people who you haven't been charging for something,

they start to get angry. Well, and they don't exactly cough up the money joyfully, for sure. That was also the company where I learned that the reason the military has $300 ashtrays is because those ashtrays cost $300 to make. We weren't making money on any of our military contracts. And they would just go into them and say, people are like, military, there's so much money to be had here.

There's so much waste and corruption we can take advantage of. There's so much rent we can seek. Sorry. But the actual process to get qualified as a military device was so enormously expensive. I think you've got to be one of the big military industrial complex corporations to kind of take advantage of that kind of graft. I think if you're a startup, you know, it's a windy road to get yourself noticed or used. When you've got, you know, the lockheed and northrup and people like that out there probably you know well i i think that lockheed and northrup and those folks definitely

have an in but i think they are they are not using i don't think there's a lot of graft i think there is really truly that much additional engineering and quality and verification of engineering and quality and verification of verification sure it seems kind of fluffy that you need all that stuff but i think there's no question that regulation make in particular military regulation makes stuff more expensive because you have to navigate the whole path. But there's a lot of no-bid contracts out there and stuff like that. And you're not going to get one of those

if you're a tiny 20-person startup. No, because they're pretty sure you can't do it. If there's any risk at all that you can't do it, they don't want to do it. And another company that I'm not going to call as a failure because I'm still waiting, there was a military contract and it was clear that the military wasn't going to choose us. So we worked with another one of those names we just named,

the big ones. And basically the big company was kind of our guarantor. If we failed, they would basically redo our stuff. Oh, redo your stuff. So they wouldn't acquire you or help you. They would just, we'll do it our stuff. Oh, redo your stuff. So they wouldn't, they wouldn't acquire you or help you. They would just, we'll do it our way. It was unclear as to what would happen, but basically the government would contract with one of the big ones and then they would sub with us. And if we failed, then they would still be on the hook with the government, but the government wouldn't deal with us directly. It's a weird situation,

but I think if you want to have a startup fail, definitely go right after that military money. It's just as easy as it looks. Not. Not. I noticed on the Amp Hour, I actually listened to that one, you can't tell when I'm kidding.

It's like I need some sort of music in the background to say when I'm sarcastic. I can tell when you're kidding, so I don't care. That's a good thing. Most of the time. Okay, so after R, there was... I went back to Cisco. Well, yeah, that wasn't really a failure, was it? A total failure.

Cisco didn't fail. No, because... The project failed. Okay. And that, briefly, was it? A total failure. Cisco didn't fail. No, because... The project failed. Okay. And that, briefly, that was Procket version 2. Cisco actually bought Procket, the remains of Procket. For the IP, the people, and the parts. Yeah, not even the parts.

IP, maybe some IP. It was mostly the people. And not many of them stayed for very long. So this idea was to redo some of the chip work within Cisco. And it just, this was a different kind of failure than I was used to. And I'll talk a little bit about what I call founderitis later. This wasn't that. This was just complete apathy. I was with a bunch of people I'd worked with before and it just, there was no cohesiveness to the team. There was no plan.

It was, we're going to build this chip. And that was actually the whole plan. So. Well, that's what I'm saying is my, is the root failure of my robotics thing was nobody there's no commitment nobody loves the product and I don't think you have to love a product enough to give up all your weekends and and work eight hour weeks but you have to to be committed someone has to to care you have to be committed enough to make some difficult decisions sometimes and maybe put in some time

that you normally wouldn't have and to really think about where is this going and what are the trade-offs and putting the whole company in your head is what is the best thing for this as though the company was a little child that you have to think for it and nobody's willing to do this i'm just supposed to do software and the few times i've gone beyond that i they've been glad but then gently pushed me back to the software which is fine with me um and i think i you know i think if i loved it i could take it over but i don't want to love somebody else's company i want to love my own company and you're expected that's the that's the hard thing about startups is oftentimes you're expected to love them with the same intensity that the founder does. Yeah. And that can be really tough,

especially if the founder's in a, uh, a management role or something where they're, they're in contact with you often and, you know, kind of reviewing your work or, or at least looking over your shoulder. And, you know, the intensity with some founders that I've known is incredible and it's to their credit, but expecting the rank and file to have that same kind of drive is a mistake because sometimes if they do it, they burn out. Yeah. Oh, yeah. And sometimes if you ask them too much or more than they're willing to give, they'll start being resentful.

And resentment was what this company came, my robotics thing came to, is lots of resentment. Not really for me, but for some of the folks around. And it's just too much. You have to, if everyone around you says you have to care, you have to give up your life for this, then you're like, well, now I have two options. Well, and it sounds selfish,

but next question is what's in it for me? And for the founders, they're talking 30, 40% stock. Yeah. And if you're employee 19, you're not getting 5%. You're not even getting 1% at that point. Yeah. So the upside may look large, but it's just,

it's hard to balance all that. And so what are you calling foundritus? So all startups are founded by somebody and they're usually founded by people who are very passionate about what they're trying to build. Because when you found a startup, you recognize that this is going to suck your life. But a lot of times the founder ends up being CEO for quite a while. Because he's the person who cares the most. He's the person who's willing to talk about it 24 hours a day.

But they're often not the person most skilled in running a company and building a team. Particularly not some of these engineering-based companies. So founderitis is when you've had a founder in charge for far too long and they're starting to make terrible decisions. Because they love the product and they're an engineer instead of they are a CEO and a marketer and a business person.

Is that? Yeah, that's fair. And they can't be pushed out for whatever reason. No, because they have 30% of the stock. Sometimes they can. Sometimes they have 51%. There's nothing you can do about it. And sometimes it's really painful when you push them out

because you're losing somebody. You're losing that visionary. Driving force. Right? And sometimes the company doesn't survive that. So it's a tricky position to try to convince somebody to take a step back instead of stepping out. Well, I think founders are a special breed of people.

Even startup folks are kind of special. I've said that I prefer companies under 20, really under 50. And it's because I like to do all of the different things. You were talking about having to do operating systems and networking protocols, you know, every day. And that many hats approach, I really enjoy it. And when it starts getting too big and I start getting pushed into one role, I get a little antsy. Not quite bored, because software definitely

interests me. That's what happened to me at Procket. Eventually, the OS was done and the work I got to do was less and less, but harder and harder, because there was such a machinery built up that changes became more more difficult and it it you know that's that's just universally true there's nothing you can do

about that but yeah it was less interesting because okay today i'm gonna tweak these five lines of code to try to get this bug fixed instead of today i'm gonna architect this complicated system and then try to implement you know 500 lines of code and some tests and see if I can get it to work by the afternoon. There was just, you know, those opportunities were gone and because we were done. But I know some people who've been at big companies for a long time and they're happy and they're good at that bug fixing and getting things going and keeping on to the next product and keeping their eyes open for what can be better. And it's a different skill set. And so when you lose your founders,

sometimes you lose some of those engineers who were in it from the beginning because they're, they're getting pushed into pigeonholes. Yeah. And the founder is being pushed into a pigeonhole of CEO when at one point he was CEO, marketing, chief bottle washer, engineer. And he was the guy who closed the garage door every night. It used to be lots and now it's one thing. I think those people who are good at a broad range of skills may not... It's a different set of skills than people who are good at a deep range of skills.

No, and that's why a lot of people rotate in startups. You start out with one crew and then you end up with another crew that's maybe a slightly different skill set or more suited to a mature company instead of a company that's trying to piece something together. Yeah, I was listening to someone else talk about a completely different company. And their problem was they had gone through three sets

of engineering teams, and each team kind of started over because they would bel would be little to use the previous guy's work obviously they didn't know what they were doing other way to be done right now and all of the if all the engineers leave at the same time you're really hosed and you probably should start over because you're not going to have the understanding of whatever they did for good or ill yeah you need more overlap more overlap. Yeah. But that's a, that's a people skills problem to figure out how to convince people that, okay,

your best buddy's leaving, your job's getting smaller. Just hang on for, you know, long enough to transfer your, your information, your empire over to somebody else. The other part of founderitis is,

um, what I call what I did before worked. And this is when you've got somebody who's founded a few startups. And maybe one or two of them were sold or went IPO, and he's a billionaire or a multimillionaire. But often the engineers and folks didn't make any money off of these things. Sometimes they did. or made a little. Sometimes it did well. Sometimes it was a great success for everybody, which is worse. Well,

not worse, but it's worse when they're coming into that next company and, well, you know, I did this before. So obviously doing that again is going to be successful. And they start to attribute all these random actions they made at that previous company to success when it may have been timing, it may have been luck, it may have been some confluence of events that can't happen again. And you're trying to make that same scenario exactly the same with the next company when it probably needs different things done. And I've seen that a couple times. That happened at a recent company and at the R company where it was like, well, we did this.

Actually, at Pocket to a certain extent as well. It's like, well, we'll just run this playbook and it'll work exactly the same. And it's like, well, it's not the same company. It's not the same people. It's not the same year. It's not the same technology. You don't have the same experience anymore. You know, it's just these things just don't apply.

You know, if you could do the exact same thing, why aren't you making the exact same company? Well, that would be silly. Right. But that's the only time you could even conceivably say this might work again. I don't think, you know, I've read a lot about random actions and luck and how much luck really plays a part in our world. Define luck, though.

I mean, it's not some supernatural. No, no, no. Not fate. Just statistical anomalies. Yeah. The person who sees the tweet for your job opening on their way, on a day when traffic is terrible and realizes they want to quit their job, as opposed to any other day you would have tweeted that,

it wouldn't have mattered. The newspaper article that's written because this person just needed something to fill their page with today and suddenly it's popular. But if you hadn't, you know, been dating the person who was talking to this guy at lunch. Right, right. Yeah. It's. Yeah.

No, there's a lot of random chance involved with everything. And I think that's what people lose sight of sometimes because they get this notion about themselves that they're, you know, I'm a serial entrepreneur. Or, you know, I look at my bank account. That's a billion dollars right there. That means I'm a genius, right? I must be right.

I must be right. I must be right. People are paying me. But that filters down to everything that happens at the company. It's not just decisions they make about funding. It's decisions they make about schedules, decisions they make about product mix and requirements. It's decisions they make about

how they want their engineers to work. There were a couple of times where, you know, a founder at a company I was at would lay down law about how you should be working, you working, when you should get there and when you should leave. Because idiots who wanted you to start at 7 a.m. each morning. Yeah. And you were trying to hire software engineers. And I think they were thinking back to, well, hmm, you know, when I was at this other company, we had these rules and it was successful.

So we'll try that. And you have to wonder, did they really have those rules or did they just have a small group of engineers who liked early morning working? Who knows? I mean, you know, at one of these companies, I got laid off, the C company,

I got laid off. While you were writing your resignation, I need to interject that. That's true. You were printing it out when they called you over and laid you off best chance use of the printer being broken yes you almost missed out on severance by 45 seconds

um but this company i forgot what I was saying before the whole. They laid you off? Oh, they laid me off. And I heard, you know, I still had friends there. And I heard through the grapevine that they had hired another embedded software engineer when they realized that this position actually wasn't optional. And her name, you know, embedded software engineering women is so rare. Her name was Elizabeth and she had red hair.

I'm like, trying to replace me much? Yeah, sometimes life is almost a sitcom, but it's not quite that funny. My life's usually a sitcom. Okay, do you want me to go next, or is it your turn? I had a thought on the other topic, but it's now fled me. Yes. All right. Okay. So I worked at a little company that was trying to improve the credit card market. Improve the credit card market? Well, they were trying to prevent fraud. Oh, okay.

And it was pretty cool. I mean, the technology was very cool. Right. Okay. I remember this now. And their intellectual property position was the strongest they had. I mean, that was the one big thing. That's always a red flag, really, really, really is.

No, it should have been. Intellectual property is not a product. Well, they had the IP, and it was definitely, this area is one that is full of IP, and if you don't have it, you're going to get sued. That's, you know. I guess everything's like that. The technology was okay for what they were doing and they had plans for better. So I was happy with that. And I'd worked with some of the people before, so it was a pretty good team.

I liked that. But they fell for this guy who every month, who discouraged them from VCc funding which was i generally agree with discouraged them talking to other angels for one reason or another and every month this guy would say next month i'll invest and if you just do this good night wesley good work today i was like i to kill you. I was likely to kill you in the morning. And the guy never invested. I mean, he lured these people into certain doom.

They overextended their personal finances. Oh, that's... They ran out of money. They ran out of options. They didn't... I mean, they still owe me money. I was part of their bankruptcy settlement. I'm sure you'll be getting that any day now.

Yes, I know, 32 cents. But it was, I don't know why somebody would drag a little company along. Why would a little company go along with that? Well, and that was odd because there was one meeting where the engineers met the investor and the engineers laughed because the investor was so comically... Stereotypical.

Evil wizard laugh evil. He was just not somebody I ever would want to have dinner with. And I think of investors... I've met some investors that would want to have dinner with. And I think of investors, I've met some investors that I would happily have dinner with. And those are the people who I would want to invest in my company. Yeah. Was his name Snidely Whiplash? Snidely?

Snidely? I can never remember which that is. Yeah. Well, that's, I mean, that's another, I mean,

that probably falls under founderitis to some extent because you still have people running the company who don't understand business or investment. They were engineers, and they really did not understand the liabilities that they could fall under personally. Well, not just the liabilities, but how to know when you're being swindled.

Swindled, yeah. And sometimes there isn't a good reason for being swindled. I mean, I think the IP was the thing that the guy wanted, and he was trying to drag the company into bankruptcy so that he could get a big chunk on the IP. How'd that work out for him? I think that actually, I don't know yet, but I mean, that's still all behind the bankruptcy veil.

That's the plan. It was the only logical thing I could attribute to him. This is all crazy because you think, I'm an engineer. I want to make this artifact, this thing. I want to make this new gadget. And I want to put it out there for people to enjoy, and I want to get compensated for my work. But when you've got a company, the whole company part doesn't care about the product.

It's, it's, it's swimming in this different ocean of politics and money and people's, people's schemes and things. And once you get to that level of, you've got a company that can produce something and sell something, you've actually got to work with those other those other discipline i don't want to say disciplines guys giving you too much credit but uh those other kinds of you know very personal and very political and very uh you know machination kind of things that engineers most of the time aren't that great at and if they are they're those one in a million people, you know, who are running apples and Googles and Tesla's and things. Um, they have those, you know, they have that skill to be able to see when something's right and go for it and also see when somebody or something is wrong and, and something before it becomes a disaster. Okay, so I definitely am starting to learn what it feels like

when I've been around a place for like a month and I can identify some things that make me think, yeah, this is going nowhere. The whole, this feels like a science project and not a product is a big red flag for me. Not being able to trust the management. Because if I don't want to hang out with them, and I don't necessarily need to actually hang out with them. But if I don't want to have dinner with them.

If there are people you'd turn around and run the other way if you met them in a dark alley, you know. I'm, you know, I'm more meal based than alley based, but definitely like for LinkedIn, some of, some of you listeners have invited me to connect on LinkedIn. And I'm sorry, that was such a weird sentence. I'm still trying to parse it. No, I'm, I'm going somewhere with this. I, and I may actually remember where I'm going, but, um, on LinkedIn, the bar is if I'll have lunch with you, I will connect with you. And if I don't know you well enough that you call me up and ask me to have lunch, then I won't connect with you. Because that means that my group, my connections, my contacts, whatever they're called, I really do know all of them, at least a little.

And I feel that way about my CEOs and my director of engineering, even my director of sales. I want to be able to converse with them. I think that's too stringent. Yeah? Yeah, I think there are a lot of good CEOs out there who I'd want running a company who I wouldn't want to spend a lot of time with or have lunch with. A, because I wouldn't want to expose myself to their eye sometimes. And B, because sometimes you need people that you wouldn't necessarily like in positions that you don't want to have. That's my personal feeling.

Yeah, that's going to be a personal difference. You know. But yeah, I think... I think if you're in a company that's only 20 people, you've got to like every one of them. That's true. I mean, sure, the 500-people company. Yeah, I think... I think if you're in a company that's only 20 people, you've got to like every one of them. That's true, that's true. I mean, sure, the 500-people company... I can't say that I've disliked most of the CEOs.

You know, you come to that point where after a few years, you think, wow, there's so many weird decisions. What are you doing? And you get mad or you get disappointed and frustrated. And it's sometimes hard to separate that from the person and realize that, okay, this is somebody out of their depth and not necessarily somebody being malicious. Because they certainly don't want to see the company fail. And they're trying everything they can.

But I think you were asking, so what does it look like when the first inklings of failure? Yeah, what do you... A lot of things. Start seeing weird turnover at the top. You know, VPs getting replaced a lot. And not just replaced once. Other C-level executives getting replaced a lot. Those guys getting replaced

and suddenly there's a whole bunch of new people who they brought in as their cronies. Whispers in the hallway. I mean, you can always tell the health of a company by just how much gossip is going around. I remember the time I quit a company and the director of engineering who I quit to asked me if he wanted me to send,

if I was going to send an email, did he want, should he send an email? We were just, I was like. The announcement email. The announcement email. And I said, no an email, did he want, should he send an email? We were just, I was like. The announcement email. The announcement email. And I said, no, no, in two weeks, we'll send the lunch email. But in the meantime, I want to see how gossip flows.

Yeah. And it was true. The people who are connected enough to walk up to me and say, oh, you're leaving. They clearly were more connected to the people than the people who Friday morning came over and said, your goodbye lunch is today. I didn't know you were leaving. They clearly were more connected to the people than the people who Friday morning came over and said, your goodbye lunch was today? I didn't know you were leaving.

I think the whispers in the hall, I don't want to say be a gossip, but definitely listen to the gossip. It's worth a couple of minutes a day to make sure to say hello and how are you. Gossip can be healthy. It could be positive gossip i remember that but you know if you're hearing things about finances from other engineers yeah there was one company i wish i had had listened to some of those rumors and instead

i listened to the official report yeah um and i really should have listened to those rumors. What else? Just thrashing. Thrashing on what are we making? I believe it's called pivoting now. No, pivoting is larger. Like, you know, we're going from making core routers to being a dog food internet ordering company. But this is more like, you know,

just medium-sized requirements changes all the time late in a project. Sometimes it's okay because you're realizing, you know, you need to address a different market or this isn't going to work. But if it happens too much, you never finish. And, you know, there's been other companies that ship crap. And that's a warning sign to me that things aren't going to go well because management is willing to just get anything out there as soon as possible to start selling something. And that's happened twice to me where we put together basically a lab prototype and minutes after powering it on, it was being put in a crate

and shipped somewhere basically. And then they built a manufacturing line and they, you know, they build these things out of, you know, point to point wires and, and, you know, and just super labor intensive, uh, uh, manufacturing with parts sourced, you know, from Digi-Key and things in the low quantities and just, Hey, we have a product yes but it's costing us two thousand dollars every time we make one when you know if they've taken a little bit more time and designed for manufacturing and things like that you know and that's happened twice and i think that actually probably helped to kill both those companies because they spent so long

not doing the right things because they thought they were done oh yeah because the science project was on the table therefore ship it and and there's there's a difference between science project and product it worked once oh yeah hire the sales team it's the definition of amateur and professional. An amateur is finished when it works once and a professional is finished when it never doesn't work. I mangled that. I'm sure there's an actual quote about that. Works forever.

It was by the Rush drummer is where I heard the quote first. What? Neil Peart. All right, well, he's typing that. Like, I don't know who Neil Peart is. No, I figured you were typing the whole... Have you seen our music collection? Yes.

Have you seen the adjacent room? That was how Christopher fell in love with me. I knew all the words to at least one Rush song. Remember it that way. Okay, anything else on the list of things that just you can kind of spend less than a month and realize your startup's going to fail? Yeah, I think notice what other people are doing.

Notice how hard they're working. And I'm not saying people should be coming in at eight and leaving at seven. I'm saying, are they doing anything? And there's been recently, I've been to places where there's people who are just there. At a startup? Yeah, and brought in by somebody

and they're kind of under their wing and they are net negatives. They'll produce almost nothing, but they'll be in meetings defending their turf, where their turf is defined as keeping them from doing anything. And just be disruptive. And that's happened a lot where... I don't understand those people.

God, I'd be so bored. They never get fired and it's so weird and they never get fired because they're friends with somebody who protects them i'm not really against cronyism because i certainly have my own go-to hardware engineers and that's different than cronyism but cronyism is usually when you're bringing your buddy over not somebody who you respect and have seen their work and they can be your buddy, but if they're just your buddy. Well, okay. Okay. I learned the new definition of the word. Well, I don't know if that's accurate, but I mean, I think cronyism has a definite negative connotation.

And if you know somebody good, you know, most of the best teams I've worked in has been people who know each other and trust each other. So, I mean, that's not really cronyism. But when a VP brings in an engineer, and it's not a VP of engineering, it's a VP of operations or something, and they bring in an engineer and that guy is put in charge of a bunch of stuff and nothing ever happens and he doesn't ever get fired. You have to wonder if the company's serious. When you're in a startup and you see, I mean, everybody thinks nobody, every engineer I know has felt that they are the only one

who actually produces something, and everybody else doesn't get jacked done. But in a startup company, if that is really honestly true and not just some small view of the world of what's in front of you, that's really bad. I mean, 20 people to have a product succeed means that you're pulling a lot more weight

if one or two of them is slacking off. And money is tight in startups, and each one of those engineers is costing hundreds of thousands of dollars a year, fully burdened. And that's just going down the tubes. That's holding a space for somebody good and wasting money that could be used elsewhere. So it's a double whammy.

Yeah, I think, you know, this is going to sound terrible, but I think sometimes a sign of a healthy company is one that knows when to let people go individually, not layoffs, although sometimes layoffs are necessary. Fire slow, fire fast. Fire fast, yeah. And somebody's not working out for whatever reason. They may be a perfectly good engineer, but startups are weird. They have their own cultures.

They have their own ways of doing things that develop organically. And some of them are just wacky. And I've been in ones where I'm not culturally suited. And I've been in ones where I was. And it's a different thing. And sometimes you can't even put your finger on what's wrong. But, you know, some people just don't work out in one situation. And sometimes they're just terrible, terrible, terrible, terrible people.

Not usually true. Just not in the right job right now. So I'm going to say these words to you, and then I'm going to watch your face turn mottled red and glare of anger appear. So I'm sorry that you listeners can't hear it. Synergistic leveraging. Fail fast.

God. I heard one of your startups really had that as a. A recent one. That was their mantra. That was their mantra. That was their mantra. Fail fast. And what they meant was screw up often and then never reevaluate the thing you screwed up

to see if it could have worked with some changes. Just move on to a completely new thing. So it's fail fast and learn nothing. Pretty much. Okay. Because I think the corollary to fail fast is supposed to be learn from your mistakes. might be true this they they forgot that part yeah there is there's there's a graveyard of uh designs somewhere in somebody's notebook that will never be looked at again that

probably could have been made to work and you know because they failed on the first option they just buried well that goes back to perseverance. You do need a lot of perseverance. Well, and fail fast can be misconstrued in other ways. Fail fast and learn something is all well and good. But if you're moving too quickly because you want to fail fast, you want to try a lot of stuff,

you're going to sometimes pass the thing that you may need to see. Sometimes science is hard. Sometimes engineering is hard. Sometimes it's methodical or needs to be methodical. Sometimes it's slow. And sometimes things take a long time. I'm surprised at how many startups continually go nuts when you throw them a two-year product schedule from conception to shipping. And just about everything I've worked on from scratch is 18 months to two years of any reasonable complexity.

Yeah, add on products. Some consumer products you could probably knock out in a year if you're focused and you know exactly what you're building. But larger systems, you have to source parts, you have to build, you know. Well, the consumer products are hard because you have to set up the manufacturing. Or you have to pay a fortune. You have to do that with anything. Even if you're manufacturing it within your own company, you have to build the manufacturing

floor, you have to qualify, you have to hire the people, you have to, you know, you have to do all that stuff. It's just lots of things that have to happen. And there have been companies I've been at where you throw them a two-year schedule and they say that's not acceptable. And you fight and you fight and you fight and pretty soon five years have gone by

and you've finished. And you could have done it on the two-year schedule. You might have only been six months late. Could have done it on the two-year or two-year plus six schedule. But since there was so much indecisiveness while trying to get a schedule that was shorter than two years it ended up taking twice that long it's it's incredible to me it's

not it's no longer incredible to me i wish it was i suspect you wish it was too well you know you shouldn't happen it happens all the time you get these people who are running these companies and you think they know how this stuff works from experience. And it turns out a lot of them don't. They just, I mean, I have that sort of feeling about two years, 18 months because it just happened over and over and over again. And they should have gotten that too. And somehow it's just, either they're just,

they don't want to accept it because they need to meet that next funding milestone and they know they can't get there if it's two years or that they think the investors won't accept it. Or they haven't really been through this before. You know, it took two years before, but that was because you were a bunch of slackers or the previous team wasn't, you know, but we can do it in six months

because we're, you know, my new company. And it's, you know, kind of trying to ignore reality and the laws of, not physics exactly, but sometimes it is physics. Scheduling, the laws of scheduling. The laws of scheduling, yeah, yeah. You know, I want to go back to Founderitis for a second because there was one thing, although I want to go less to your Founderitis and more back to founderitis for a second because there was one thing although i want to go less to your founderitis and more back to mine um in that founders and early startup people tend to be different than those who can sustain a company up into the 500 employee level yeah um because

not only is it a different skill set, the breadth versus depth, there's also a different mindset. The small company people, the under 20 engineers sort of companies, or even worse,

the under five engineers sort of companies, get told you can't do that all the time. They get told that what they're doing is impossible. And so they hear those words and they don't listen. Because if they listen, then they wouldn't be doing this. And they wouldn't have succeeded in the last venture. And so they get used to not hearing, that's not a good idea.

They get used to hearing or to, to failing to hear, um, this is wrong. This is impossible. This is not what, and they, they get so used to following their own vision that they can't see anybody

else's has value. Yeah. And I think that has happened in a couple of my startups that it reaches that point where the founder can no longer be the central hub of all communications. And it all breaks apart because they won't listen to anybody else. I wonder if that's a common failing where you've had a founder that had a successful startup before

and that's actually bad. I mean, that goes back to what I was saying about what I did before worked. Yeah. But it's beyond that. It's more like what I did before worked and nobody listened to me. Therefore, I'm even more right because I did the impossible once before and of course I can do it again. And, you know, I don't have to take any advice this time at all.

Yeah, last time they don't realize they actually did take some advice and thought about some of those things people were saying in order to prove to themselves it was worthwhile. Yeah, I wonder if I would rather work for a startup where the founder had failed the time before. I remember interviewing a job and the founder had failed before and before. I remember interviewing a job and the founder had failed before

and I asked him about it. And he was very upfront and very personal about it. And that's one of the startups that I'm not really it may still succeed, it may not. Well, let's see. What else?

Yeah, we're about out of time um i did look up some other info um you read about everpix they they published all of their information about their failure they did we should have talked about them i should have read that and then talked about them well maybe we'll do failure failure 2.0 at some point. Failure 2.0. Yeah. The faility. They put up their entire startup Bible, all their funding stuff, all the pitches. I'll track that down for the show notes.

That was really hard for me because I liked Everpix. I really liked their product and I was willing to pay for it. Yeah. And then I also, as I did a little bit of research, uh, I came across a business insider.com of, uh, basically a bibliography of startups that died and why they failed. And they're all written from the startups, uh, perspective. So that was kind of cool. Um, it does get depressing to read about all the fails.

I think you can read about successes and learn something. You can read about fails and learn something. But then the overlap is strangely high. The reasons some people succeed are the reasons other people fail. That's where it goes back to luck or just serendipity. And I often wonder,

all these companies that fail, some of them are doing sophisticated stuff. And I wonder how many times, like the fail fast situation, how much really interesting technology is kind of on the bottom of the ocean and nobody's looking back and saying, hey, you know, we can probably make this work now. Beyond IP and patents that are lying fallow, I bet there's stuff that just wasn't patented

and that company failed and it's just gone. Well, that's kind of sad. Okay, we're not going to end on that for sure. There's transporters out there somewhere. A replicator. Okay, so I haven't figured out what makes the company succeed but if you did we wouldn't be doing a podcast might be doing a podcast we do this for fun remember oh we'd be doing a podcast all about our money and about about how our successful startup succeeded because we were

in charge doing only the things that we knew how to do because it obviously succeeded. So therefore, everything we did was right. Making Embedded Systems, the show for people who love gold-plated gadgets. Oh, screw gold-plated. Gold gadgets. Fully gold gadgets. Like electronics made out of gold. Just blocks of gold.

With LCD displays on top. We say that, but we're pretty lucky with what we have. Okay, so what one or two or three things do you think are required for making a company succeed? We've talked about failure. Let's try to identify what we care about. I think an open mind from the people who are running it and open eyes is number one.

And I don't know how you can quantify that or figure out if they have it, but, you know, people who really listen to, listen, they don't have to act, but listen to what their employees are saying, have an ear to the ground of what's happening in the company and are open with sharing with the company, you know, some of the gory details sometimes. Transparency is really nice. They're, you know, sugarcoating things in emails and, you know, everything's always great when you're looking around and the place is in flames. You know, there's a lot, I've been in so many companies where rearranging deck chairs on the Titanic became, you know, an after meeting catchphrase. Um, so that's one. Um, two is, is kind of being able to pivot and not, I hate that word and I hate, but,

but being able to know quickly, and I don't mean this from fail fast, because it's not necessarily a failure, but to know quickly when what your plan currently is isn't going to work, for whatever reason. It could be you can't do the technology, it could be the market's changed in the six months since you were working on it, but keeping an eye on, you know, is this going to work? And if it isn't going to work, how can we make changes so that we still have a chance? Because a lot of times you get a plan in 2000 and whatever, 2005, and by 2008 you've got a product and you look around and it's not the same planet you started on. I've been replaced by an iPhone app.

Your customers are gone. It's a quickly changing world. I've been replaced by an iPhone app. Your customers are gone. Yeah. It's a quickly changing world. That's one of the things that makes it really hard to do a hardware startup. I mean, you have physical things they take time to build. Yeah. And I think you do need,

and I think number three thing is I think you do need passionate people. That's my number one. Everybody doesn't have to be passionate, but you do need somebody up there That's my number one. Everybody doesn't have to be passionate, but you do need somebody. You need somebody up there who's willing to say, this is a beautiful thing that I want to make. And a visionary, somebody who's really,

this is where I want to go. I want you all to come with me. And it doesn't have to be the CEO. A lot of times the vision is, this is the IPO I want to have. The product, it's a means to get to the CEO. A lot of times the vision is this is the IPO I want to have. The product, it's a means to get to the IPO. But that usually doesn't make for a good product. No, I really would rather they cared about what they were making.

But that's partially because I always want to care about what I'm making. All right. Well, that's the show, I think. All righty. If you liked this one and you're a new listener, go back and listen to episode 24, I Am a Total Fraud, in which Christopher and I discuss imposter syndrome amongst other strange things. That was a little more personal than this one, but it had some parallel themes. Christopher, thank you for producing and for helping puncture my little success bubble. Always glad to be of service.

And as always, thank you for listening. As usual, feel free to say hello or make suggestions, email us show at embedded.fm or hit the contact link on embedded.fm. A final thought, this one important to those of you who are staying at your sketchy startup. It's only hubris if I fail from Julius Caesar.