Elliot Godzich joins us to share his tale of working for a startup through its acquisition and why he made the decision to leave the salaried life to start a new business in virtual reality software.
- what was Elliot’s background that afforded him to make the entrepeneurial leap (and why he would do it again)
- why you should fail fast
- how much money Elliot lost in the process
For the full text transcript see below the fold:
Max: Welcome, guys!
Today, we have the pleasure of Elliot Godzich joining us. Elliot is a senior software engineer at Juniper Square, a private equity software firm in San Francisco, California, USA.
First of all, I wanna share with our audience why you’re such an amazing guest to have on and why your experience is super interesting and relatable to virtually everybody in tech.
I think a lot of people have this idea in the back of their heads that they want to start their own startups eventually–even if it starts out as part-time as opposed to full-time.
And you have made that decision, started your own startup, and then you have made the really difficult decision to abandon that plan and re-enter the workforce as a full-time employee, salary and what not.
For our audience, that doesn’t know you personally, do you mind sharing for people your background, from college through to today?
Elliot: Sure. Coming out of college, I was working for a company called Originate. They were somewhere between a venture capital firm and a consulting firm. We did small gigs for companies in exchange for both cash and some sort of upside. I stayed there for about two years, I was working in Scala, which was really fun. But I left because I wanted to get deeper into projects.
I’m sure you can imagine, you get really good at starting projects and then you’re like, “What… I’m bored now.”
So I left and joined a small start-up. At the time it was 10 people doing e-commerce retail advertising on Facebook. I switched from being a full-stack engineer at that point in my career to being only a back-end engineer. I was doing this ads automation system.
It was a big change, as you can imagine, going from…at the time, we had grown to 24 people. Rakuten as a whole was a 15,000 person company, and Rakuten Marketing was 3,000 or something.
Max: Yeah, they’re an enormous Japanese conglomerate that are making their way, through acquisitions, into the United States tech market place.
Elliot: Yeah. Rakuten Marketing, specifically, is entirely composed of acquired companies, there is no actual Rakuten Marketing company.
As you can imagine, the cultural situation at Rakuten Marketing was a complete mess because they…there was no central culture to wrap around because every single one of these [acquired] companies were coming together with their own culture and then being forced into an office that none of them chose.
Max: You had worked at this startup in the e-commerce space for two years, you said?
Elliot: Yeah. About two years, and then I spent another six months at Rakuten Marketing.
Max: We’ll get to the topic of your startup and your decision to leave Rakuten, but do you mind sharing for people what the acquisition was like, working at a tech startup that got acquired by an enormous conglomerate?
Elliot: Sure, yeah. At first, we were…we started getting office visits from them, right? They were thinking about buying us. Of course this was a big deal, everyone was like, “Oh, what’s happening? What does this mean?”
And our bosses really couldn’t tell us anything, they said, “Well, we don’t know what it means yet, but we’ll let you know when we do.” Eventually, you have people coming in and doing technical due diligence. So my head of engineering was completely out of the picture for weeks on end because he was deep diving with someone technical on their end to make sure that there weren’t any dead skeletons in our closet, technically speaking. I’m sure the same sort of thing was happening on the business end.
Eventually they told us, “We’re acquiring you.” They took each of us into a room separately and put a bunch of paper on the table and said, “This is the terms of the acquisition for you,” because you’re essentially getting rehired at a new company now. “This is your new salary, this is your new stake in the new company, this is how much money you’re receiving,” just all at once, for getting acquired.
“This is where the new office is.” And they had literally printed out a special map for each person, like, “This is your new commute.”
Elliot: Yeah. I was going from having 20 minutes of walking outside as part of my commute, as part of my 40-minute commute, to now having to drive 45 minutes, at a bare minimum, but that’s what the map said because that’s the distance on Google Maps–with traffic, it was actually an hour and a half.
Max: Did you have to re-interview for your own job, like some people do?
Elliot: No, I didn’t have to do that. They just hired everyone. They did talk to each of us and, I guess, we all seemed decent enough people. So there was no intense vetting of people. But that sounds really scary, I can’t even imagine what that feels like having to re-interview to keep working here…
Max: Well, there were other aspects of that transition that you were telling me about, before we started recording, that we’re kind of grating, like you were just describing that there was no real culture there after acquisition?
Elliot: Yeah. And everything became a lot more complicated too.
So I was the head of back-end. When I decided I wanna build something I just built it. I would plan it and I would break it down into stories and fan pieces of it out to my engineers, and things would get done, it was great. If I needed to understand priorities, I would go talk to my CEO, I’d go talk to my head of product, magical.
Shifting over to Rakuten Marketing though, things got more complicated because there are all these kind of financial things that you had to do, they were trying to save a lot of money by capitalizing projects. I don’t know if you know what that means but the basic gist of it is, any software you write that is adding new functionality to your software and…anything that’s non-maintenance, basically, can be considered as a capital asset that you’re building. It’s taxed differently, it’s taxed as a depreciating asset.
So we had all these crazy requirements to time how long we spent coding “capitalizable” projects versus “non-capitalizable” projects. As you can imagine, as a software engineer, you’re switching between branches and suddenly you have to go to a time tracker and click all these things, and it was just…it was crazy.
My engineers basically mutinied on doing time tracking and so I was estimating all of this for them. To build things you have to get your roadmap approved. Whenever I came up with a thing, I had to go through this whole set of meetings, just to even start building it.
We were working with Facebook, right? They were our lifeblood and they were constantly building what we built with an 18-month lag. Our advantage was how agile we were and now we were getting weighted down and that 18-month gap is much smaller now. It was very un-ideal from a lot of situations, but yeah… It was not good.
Max: Are the accounting requirements that were imposed on you something that you found to be unique from your peers’ experiences post-acquisition, or working for larger companies?
This is the first I’ve ever heard of having to do time-tracking to service accounting priorities.
Elliot: Yeah. I don’t know, honestly. I think I’ve heard that Amazon does things like this sometimes.
The other thing that was definitely complicated was hiring. I had been doing a lot of hiring as the lead of the back-end team and at my startup, it was a pretty easy process. Hiring is never easy actually, but that’s a different topic. But I was in full control of the process–I just, I didn’t even think about it. I thought I was still in control of the process. So I was just still going on with my hiring and then about a month and a half into me working on hiring, suddenly I get very angry emails from HR like, “Oh, you can’t…we’re the people who do hiring.” And I was like, “Oh, I did not realize there’s a whole specialized department for this.”
The way we wanted to do hiring is very different from how the way they wanted to do hiring. We were pretty convinced that being able to talk engineer-to-engineer immediately was a great way to pull people in because a lot of people are turned off by the ads industry. I found that when talking to engineers, if you actually present the technical challenges to them first they get more excited about it. They were basically saying, “We get to be the first people to talk to them before you.” And since we were doing analytics on our entire hiring funnel, I can tell you that that definitely had a really bad effect at the top of the funnel. There’s lots of stuff like that–just a lot more process, a lot more people involved in any single thing you do. And I get it, it’s a big company, you kind of have to, but it was just…it was cultural whiplash going from 24 people to a 15,000-person company.
Max: So the bureaucratic red tape, the process changes, the cultural changes–by this point you had saved up a financial nest egg to be able to do without full-time employment for a while.
I think a lot of people will understand why you’d wanna leave, encountering all these roadblocks.
So how did you manage to break free of the golden handcuffs and break out on your own?
Elliot: I was, really unhappy, and I was thinking about leaving for a while. I spoke to my boss about it because we had a really good relationship and still do. He said, you know, “Jose, our head of product who left…” he had left like a few months prior, “he seems to be working on some sort of VR thing or thinking about working on it anyway, you should go have dinner with him.” So I went and had dinner with him and he tells me that he’s become really interested in the VR space and wants to build a company around a telecommuting app in VR.
I’m sitting here going, “Well, I wanna quit my job and that sounds way more fun than what I’m doing right now, which is essentially just being in meetings all day to get very little done.”
I said, “Yeah, I’m down, let’s do this. Give me a month to transition out of my job and I’ll join you. I’ll give this a shot.” A month later I go and find him in this tiny room that he’s renting, in a school for animation, right across the street from Pixar.
The room was smaller than this room, right? It’s maybe 10 feet across and six feet wide. We had a desk on each side and we were working in VR. It was a hilarious situation because we both had VR headsets and we’re both… I don’t know if you know how the HTC VIVE works, it has those light boxes that are that are emitting light into the room that are picked up by all the sensors.
So we’re sharing the light boxes but when we were both in VR at the same time, we couldn’t see where the other person was, so there was constant bumping into each other. But it was really cheap rent for office space that was less than 10 minutes from both of us, so we made it work.
But yeah, so I just decided I’m gonna try and start a startup for the first time. I’ve tried it before but not dedicating fulltime work to it and so we just went for it.
Max: Did you guys, from day one, concern yourselves with, obviously, costs but what about revenue? Did you guys have a calendar date by which you wanted to be signing your first customer or making your first money?
Elliot: Yeah. I think that was, honestly, our biggest mistake. If I had to retrospect on the whole experience, I think the biggest mistake we made was not getting customers first.
Between the startup experience and all of my interviewing that I did afterwards with startups—[the experience] gave me so much to talk about, I found that all the successful startups I spoke to were…almost none of them had a product before they had their first customer.
A lot of these really successful founders were going around and speaking to people in the industry they were trying to serve and understanding their problems deeply, and then getting buy-in from people to say, “Hey, we’re gonna build a product.”
They have mock-ups basically, here is what it would look like, but they have not built any of it yet. And they said, “Are you interested, would you pay for this?”
Okay, you will pay for it, sign this contract, and when we get it built, we’ll give it to you and start charging you for it. They get to the point where they have three or four customers signed on and it’s enough to pay rent for the office and give you food, basically.
If you can get to that kind of steady state, then everything else is much better from there. It is really morally draining to be in the situation I was in, which was building the product first and then trying to find customers later because you’ve invested a lot more at that point.
And it turns out, if no one wants what you’re building, there was no reason to do it in the first place. But it turns out you can find out if people want what you’re building before you build it.
And that never really occurred to me before, and I think that’s definitely the way to do it now.
Max: This might be specific to the startup that you guys were working on, but what impediments do you think there are to finding businesses that are willing to pay for what you guys were trying to build? Separately, what exactly was it that you guys struck out to build?
Elliot: We were trying to build telecommuting software in virtual reality. The motivation behind it was that there’s a special thing that happens when you’re in person with someone, as far as communication bandwidth goes.
There’s facial expressions, there’s body language, there’s all this stuff that video chat doesn’t actually capture that well for me. I don’t know if you’ve ever tried whiteboard collaboration with someone over video chat, it’s not great.
Max: There’s an amazing article that was written on the topic of bandwidth and latency in communication, and what happens when bandwidth is reduced in communication.
For example, you’re not able to see the person’s face in high-res or you’re not able to see their body language. And latency means delays in communication, so if any of you guys have ever been on a phone call or screen sharing and there’s a delay…
Elliot: Oh, it’s awful. You’re always interrupting each other like in the middle of a sentence without realizing it, it’s terrible.
Max: So the reason why a lot of employees haven’t embraced working from home and working remote–the author argues that it’s a prisoner’s dilemma situation, where if you have two employees who report to the same manager, it’s important to signal to your manager your loyalty, your output. If you can communicate your loyalty and output with higher bandwidth and lower latency…
Elliot: Right, the cooperation is to stay home but the defect move is to go into the office.
Max: Exactly, exactly. It’s a brutally accurate analysis and really undermines what you guys might have been trying to do, which is push for more remote-work technology.
What did you guys find in trying to build it?
Elliot: We wanted to create a situation where you put the VR helmet on and you have a full-body, and the other person tied in has a full-body, so you actually get body language and voice and it feels like you’re actually in a room with someone.
We wanted to make a collaboration space that was more powerful than real life… We built things like whiteboards that you could pick up and produce them in midair, draw on them, toss them over to your friend. Weightless whiteboards, you can copy/paste them, you can undo all this stuff. And it was great.
We spent about two months prototyping the first version. It was a lot of learning because I had never done 3D programming, I had never worked in VR. My co-founder was doing all the graphics, he had never built 3D models.
t was really funny because we would be really productive for two days doing a task and then we’d hit some new concept like, “Oh, how does the lighting work?” And then you have to just read the whole internet to understand how lighting works. And then you’re like, “Okay, now I can set up lighting.” And then you get there and you’re like, “Oh, the lighting has slowed everything down.” Now you have to figure out how to do performance tuning for lighting, and performance tuning is really really important in VR because if you’re under 90 frames a second in VR, you can actually…you feel the world not quite tracking you correctly and it’s very nauseating.
Imagine if you turned your head and the world followed afterwards “chunkily”, it’s not a great feeling.
Max: Do you think that the immaturity of some of the software you were building on top of contributed to your guys’s problems or was the primary problem not finding paying customers fast enough?
Elliot: So, okay, yeah. Let me get to how the business didn’t work out. We spent two months building prototype, and at that point we decided to go shop for investors.
This was a very new experience for me because I’ve done engineering before, so the first two months…which, by the way, were the most productive two months of engineering I’ve ever done in my life, despite all of the learning barriers. It’s amazing what you can accomplish when your head’s down, focused on just engineering and you have no other problems. I highly recommend that experience.
So we reached out to our personal networks and said, “Who knows VC’s?” and got a bunch of meetings with them. We actually demoed the product to them, put them in VR and engaged, had the meeting in our app sometimes, which was fun. And everyone was really really compelled by the experience, they said, “This is amazing. This is really cool. I think this would be a wonderful tool that the world wants, but…” But, as you may or may not know, the HTC VIVE is pretty expensive, it’s…at the time, it was $800.
Max: We had a video recording a week or two back, where we did a price comparison about your different VR options.
Elliot: Oh yeah, I saw it.
Max: All in, I think it’s two grand for the computer with the right GPU and VIVE headset.
Elliot: So this is exactly what [the VCs] brought up. They said, “right now, to use your application, the barrier to entry is $2,000, before they even start paying for your application.” And not only that, but it’s bound to a room, I can’t use it anywhere else. For all of the investors that was a non-starter.
They said, “There’s no way you’re going to make it in the enterprise space with those kinds of barriers to entry on top of already being an experimental concept.” These meetings were both encouraging, in terms of people shared excitement for our vision, but also discouraging, in terms of “the implementation’s not really working out.”
We brainstormed for a bit and then said, “You know what we can do, to work around this, is port the application to Google Daydream.”
Google Daydream is a VR system where you have this $25 headset that you slot your phone into, and then you put it on so it’s portable. Most people already have phones so the barrier to entry is reduced at $25 and it’s mobile, so you can work if you’re traveling and still teleconference into your meetings.
Porting it to mobile was its own technical challenge. We found we were able to get away with, it was like one tenth of the polygons in a scene, so we’re already reducing the bandwidth of the communication.
The graphics I don’t think were the most important part of the application, it was having the body language and the voice. Those things, those were some of what was lost.
With HTC VIVE, you have a headset and two wands and using inverse kinematics you could figure out where the person’s arms should be, where the person’s body should be, as a result of that. We had full body language in the VIVE. The Daydream doesn’t have any hands. It has a remote but the remote SDK only gives you where it is pointing, it doesn’t give you a position in space.
Max: So you don’t have any accelerometer?
Elliot: Yeah. I don’t know what technology they use in it, but it’s something strange.
Max: Point being: porting from VIVE to Daydream killed a lot of the attractive features of what you guys were trying to build?
Elliot: Exactly. The headset also lacked translational tracking, so it could tell you where you were “planing” in space but couldn’t tell you where you are in space. So from a first-person perspective, when you were in the Daydream application, if you moved your head then the room moved with you which was not great. But more importantly, from a second person perspective, if I was interacting with you… Before, if you had jumped for joy or done a little dance it would have looked like a very happy little avatar. Now it was just a bobblehead–the body doesn’t move at all and you’re just this head on a swivel.
And there’s no way to get around it, the platform doesn’t give us enough data to actually re-create your body language.
Between all of these factors, it killed the experience. We, even as founders who were excited about it, could no longer tell ourselves that this was better than video chat. So we were stuck between a rock and a hard place because there are a business requirement on this side and a technical requirement on this side that couldn’t meet to make our vision a reality.
Max: At what point did it dawn on both you and your co-founder, or just you, that this wasn’t gonna shake out business-wise and career-wise for you?
Elliot: It was a slower death than I would have liked. Here’s how it down. We were porting the app and I think we were slowly realizing, in the back of our heads but hadn’t made it up to my consciousness yet, that this was a failure. At this point, we were no longer in the same office. As soon as we got the first prototype working, we got rid of the office and started only communicating through our app.
Elliot: Yeah, because it was cheaper. And we’re big believers in dogfooding, like using your own product because basically you get rid of the most annoying things for your app first, you smooth out the user experience to please yourself.
I noticed that my productivity was slipping compared…it was some of the worst productivity weeks I’ve ever had in my life. It was taking me two weeks to accomplish what, during that first month, was taking me half a day. That was near the end but it was slowly kind of dying in motivation. And I was feeling guilty, you know, I was like, “Oh, I’m letting my founder down…my co-founder down.”
But the same thing, I noticed, was happening to him, he was also really, really slowing down. And we weren’t talking about it for a while, and then, finally, I called him up and I’m like, “Hey, I think something’s wrong,” And he’s like, “Yeah, it’s not…this is not good. I think, like, I don’t think there’s anything…anywhere for this to go.”
We were actually in complete agreement, we just hadn’t come out and said it yet. We agreed to shelve the project. You can never…you never really destroy a software startup, it’s still a repo, a repository somewhere. But we said, you know, I think we need to stop spending time on this because we’re both just feeling bad and I don’t think there’s any way… We couldn’t really figure out any way to make it work so time to cut the losses and continue on.
Max: I know this is a little bit of nitty-gritty detail for our audience, but I think for our audience members who are serious about wanting to start a business, did you guys incorporate as a business?
Elliot: We did.
Max: Did you have to dis-incorporate as a business, or unincorporate?
Elliot: I actually don’t know what happened there. Jose was handling that side of things near the beginning. I think, he got in touch with the same lawyers that we were using at Manifest to handle that stuff, to incorporate. But basically you, by default, dis-incorporate if you don’t pay your reincorporation fees every year. So I think we just kind of let that process work itself out.
Max: I know this may be sensitive information, but for our audience who are considering doing something like this themselves, and it differs depending on your business model and what you’re setting out to do and how quickly you get to paying customers, but all-in, how much money did you forego?
Elliot: Oh, sure.
Max: What was the opportunity cost, the cost?
Elliot: Yeah, I haven’t talked about that. Having been acquired recently, both Jose and I had made a lot of money to sit on and so we were like “we can just burn our own cash for a while as an investment” basically.
Our plan was, when we get an investor, then we can pay ourselves the bare minimum to cover our own costs like food and rent. So I…actually this is a little detail left out, but we actually ended up finding someone who was willing to give us $25,000 near the end, but we decided not to take it because we were stopping the project and we were just like that’s…we don’t wanna do that. But I think over… I think it was over the course of the six months, from October to March, I burned through 25-30k.
Max: Six months. Is that right?
Elliot: Yeah. I think it was less than that. Anyway, I don’t remember.
Max: So something like 20k?
Elliot: Yeah, between 20 and 30k. And you know, rent is killer.
Max: For sure. And the opportunity cost of the salary that you might have otherwise been…
Elliot: Yeah. And you know that money could be accruing interest, you know all of that. But I would do it again.
Coming out of that experience, I felt so much more self-motivated to work on projects long term. And going into interviews for full-time positions after that, I had so much to talk about.
Especially when interviewing at startups. When I spoke to the founders, I found that I actually related to them a lot more now. I could talk about the business side of things, I could talk about what was it like during the early days, all of that. I found that that made me much more engaging in my interviews than I had before. It wasn’t all focused around me, as an engineer–I was a more well-rounded individual to bring into their early-stage startup.
Max: It makes a lot of sense.
Dude, this has been freaking awesome. Your story is, I think, very compelling to the vast majority of our audience.
I want to welcome people to check out the show notes because we’ll include a lot of links to the types of stuff we were just talking about. But also, if you guys have any questions for Elliot, get in touch with us, through our website, there’s a contact page or the comments section.
Elliot: Yeah. I’m gonna answer any questions.
Max: Absolutely. Dude, pleasure. We’ll do this again very soon.
Elliot: I hope!