· 01:01:31
Nitay (00:02.012)
Eric, thanks for joining us today on the pod. Great to have you here with Costas and I. Why don't we start with give the audience just a bit of your background and kind of work experience and a bit about yourself.
Eruj (00:13.349)
Thank you. Hey, it's nice to be here. Probably the first thing I would say about myself is I'm old. I feel old. I'm about to turn 61. And I was going to say that I started my tech career long before many of the people I work with were born, which is sort of odd. In fact, the people that I've even worked for are...
younger than my first career. started my career loading punch cards into computers that control data overnight. So the way software was written back then is an engineer would write some stuff. They'd bring me a box of punch cards, I at 7 PM, so whatever. We'd take a box of punch cards, run through a punch card machine, run the program as best I possibly could, and maybe they'd come back in the middle of the night or in the morning and get back the results of their software. So I've seen the industry change from me
know, loading cards or dropping a box of cards on the floor and having to put them back in order, which was, which is always pleasant to where we are today. So I this really interesting perspective of how computing has changed. I did my first startup. I was part of my first startup in, I think, 87 or so. At this point in time, the venture capital industry was five to 10 VCs on one road. You could hit them all in one day.
The industry was surprisingly similar. One interesting thing to talk about is how I think one of the industries that hasn't changed the most is the venture industry. So that process of raising money and working with VCs is similar to what it is today. But then my career span, startups too, I had very interesting jobs at Apple during the post Steve Jobs era, the Scully era.
And I was there for the Scully era. then when that went south, I jumped to a place called Taligent where it was a very interesting job. I was working on the desktop of the future with IBM, HP and Apple, but it was a joke and we can talk about that. It was a utter joke. And then I did a startup company or two. Ended up at Disney. I built go.com. I ran InfoSeek search engine for a while. I worked with Michael Eisner for a bit, which was pretty fucking cool.
Eruj (02:29.339)
Excuse my language, I have a potty talk. can bleep it out later. I have no filter. I did more startup companies. think the most financially successful one was Splunk, which was another thing to talk about because it was really designed to be successful. think prior to that, we were not building companies designed to be successful. And when you think about maybe starting a company that's designed to be successful, it's a little bit different than designing something that's fun, cool.
Interesting. know, this one thing was designed to be interesting to design to be successful and wise in the end I'm disappointed with the outcome. I wish I had stayed But I guess financially it was successful and now I'm backing on I've done a couple companies since bunk and I'm On another one, which I think is just the most fun in the world, but I'm 61 and I will say I have to go to bed early I feel old I'm no longer the person I was 20 years ago
And I want to be in my head, I still feel like I'm 23, but I'm just not. My brain, my body are not 23 anymore. Is that about it? I mean, I'm sure I missed some stuff along the way. I live in Wyoming. I rarely leave these days. What else is interesting? I'm into...
I can go with other things besides tech we could talk about as well, my hobbies and other stuff.
Nitay (03:58.914)
That's quite a wealth of background. That's very impressive. A lot of stories there. I'm curious in those early days after having kind of like you said, having done the Apple stint and a few others and starting early companies, you said something really interesting about designing to be successful or not. So what is it before even Splunk, kind of the early companies and initiatives that you were doing, what is it that you were doing? Did you consciously not design it to be successful and what does that even mean?
Eruj (04:25.979)
Yeah, think it's experience. I'll chalk it up to experience, but I think it's also a philosophical difference. think a lot of the companies that we... So we did a couple companies before Splunk, even that crew I did Splunk with. And you start to start with an idea.
Like you always start with some sort of idea because you're either fascinated with it and often you come out of a, like the company we did right out of Taligent was really an extension of the work we were doing at Taligent. Meaning I think a lot of startup companies is you're in a job, you're doing something in a job, you realize, there's a way better way to do this thing. And you go off and you do this thing. Right. And I think that's a lot of what we were doing is we had an idea, usually married with a piece of technology that's enabling it, right? You're sort of on the forefront of, we couldn't do this four years ago.
But we can do it now because of technology coming along. But it's something I'm familiar with, and it's an idea. And I think these things turn out typically pretty well because you have the experience, right? You firsthand knowledge. It's a problem I've been trying to solve for a while. Maybe I've been solving in a big company and I go to a littler company, or there's just some maybe 45 to 90 degree twist on it that's just new technology, new customers. But there's some vein or thread of
of it that you're coming out of. I think where the Splunk thing came along is if you start with a company saying, hey, what are all the things, I'd call it the physics of a business, not the product. But if you start with the physics of what the business needs to do and how it needs to operate, and you write down all the physics of the things that are required to make it successful, and some of them are a choice. Like, I don't know, we want to...
have something that's only, it's very expensive when we very few customers and we're gonna charge a million bucks. But nonetheless, there's a physics to the business that's sort of independent of what it is. But it doesn't include how you go to market and everything else. But if you write down the physics, and Splunk was more of an example of writing down the physics.
Eruj (06:29.921)
and trying to form fit an idea or product into the physics, which is to say you sort of scaffold up like this is what gravity is and this is what our magnetism is. And then you come along and you start thinking about what really operates best in this world of physics.
And because what typically happens is you build a startup company that's cool, but it's a cat that floats, right? It sounds super cool. It's neat. But then you gather and you go, shit, we just can't make a cat that floats. There's some physics problem about your go-to-market, your pricing, how your inside sales team works, like the marketing channels of how do I talk to a customer? How do I even get a hold of them? And how are they going to hear me? How do I differentiate?
Like how do I deposition my customers? Who are my customers? So you start with like saying, I don't care what the idea is. I don't give a shit. And you start with just the physics of all the things the company needs to do to be successful. And then you start taking ideas and you ideas have to be, you have to be motivated by it. can't be something completely random. start, so in the case of Splunk, a lot of this is just an example of form-fitting.
concepts into a physics to see if it fits the physics or not. And you'd be surprised how many things, and sounds stupid, but you'd be surprised how many come along and you're like, well crap, that's cool.
But I really can't reach my customers effectively or scalably. Or I don't have an inside sales team that effectively works. I don't have one of the fun ones is, OK, my product has to be free to 60 % of my customers. But then I also have to charge seven figures to at least 10 of them in my year three. And that's the physics of the product, completely independent of what it is you're doing.
Eruj (08:08.067)
And so you start to take an idea and say, look, the only way I'm going to be successful is if my business operates successfully. And you start with the model and not with the idea, you typically end up with a product that's going to be successful because you already know it meets the physics of a successful business. Does this make sense? Because too many times you just start with an idea and you get out there and you're like, shit, we hadn't thought.
Nitay (08:23.232)
It does.
Eruj (08:27.003)
of how we're going to deposition the customers and how we're different. So it's like all thinking about, just imagine the company's ultimately successful, remove the product completely. What does it look like? What is the ratio? How does my inside say, like I like enterprise software. I didn't do, sorry.
What, you know, my ratio, how do I scale out my field? What does that look like? Okay. Before the products even designed, how are they talking? Do I want PS? Do I not like PS? Like just, it's all of that modeling. then, and then you take the idea and surprisingly it narrows down what's going to work pretty well. It's kind of an interesting fitting function that
The prior company is because we didn't have the experience. I'd never built that an inside sales team, right? I'd never thought about it when I was starting the company. So I'm like, setting the company, then you get out there and you're like, shit, how's inside sales going to work? And then you fall on your face and then you end up selling the company. was like a lot. have a bunch of acquisitions just because the go-to-market physics wasn't, we didn't have the experience in the go-to-market physics of understanding how to get a seven figure deal.
and also have free product, how to talk to that customer. We just never had the experience, but we were really strong technically and product people. So you could give us an idea and we could chase that. We could try and get a peg that flies, but we couldn't make it fly. We made a killer peg, but we couldn't make it fly. It's like, well, crap. Why didn't we think about that? So that's why I say, I was in Splunk, I was probably in my 40s, right? We're close to it. We just failed enough times. We could just write down a list of everything that made us fail.
and all the things that hurt and sucked and start with the hurt and suck and fail and then come back to the idea. And the ideas are things you wanted, in the end they're ideas you want to do, not like, but, and they had to be things you understand. Like we, you know, we really wanted to cure cancer, but we weren't going to do it. It's not going to happen. Right. So does that make sense?
Nitay (10:19.774)
How do you know it does it and I love that, you know, one of the things you were saying here, there's a very crisp and interesting way to put it, which is that the physics of the business, because I think to a lot of especially technical founders, right, they know how to like iterate and just start writing code and iterate on product. Like that's a clear workflow. That's a clear thing for me. I can sit, can do, and I can get more done and I can look at the end of the day and say, okay, I'm farther than I was at the beginning of the day.
Eruj (10:34.181)
Huh?
Nitay (10:45.302)
But when I think about, like you're saying, the go-to-market, the business model, essentially, the pricing, the sales motions, et cetera, all that kind of stuff, I think to a lot of people that is like this amorphous art somewhere over there. And specifying and describing it as a physics makes it a science. And so how do you go about it as a science? How do you iterate on this in the early days?
Eruj (10:50.306)
huh,
Eruj (11:02.49)
Yeah.
Eruj (11:05.915)
It's literally, it's exactly like, I mean, in a funny way, it is a fitting function. It's just a fitting function, right? What you do is, you, I'm sure if someone mathematically wanted to come along and model it, they probably could, but it is fitting, right? You just, at every, and it's interesting because it, along the decision points, you have to hold a decision against the physics, if that makes sense. Will this violate, like if this was the space we're building a company and we're coming along, we're doing something, you would test,
the direction against the physics to make sure it fits the physics, right? Because it can't, whatever you're doing at any point in time cannot violate the It's like some, like probably online video games, right? You can't go floating through walls. It just doesn't work, right? So there's this, there's this model which you always are upholding a decision against the physics of the business and how you want it to scale, you know, year five, let's just say you model it at hundred million dollars, right? In ARR. You just go model your business. How do you want it to look? What is it supposed to look like?
and you think about how am I communicating with my customers? What's my support infrastructure look like? These are the things that make businesses fail, right? It's not a killer idea. I want to tell you the best companies we ever did had fucking killer ideas, but they didn't make sense. Again, they're pig the flies. It's cool. It's awesome. When you look at it, I can give a demo, your eyes are going to go round and round, but guess what? It's never going to make $100 million, right? Just because there's some breakdown in the physics of it.
And often those things just come from experience, right? It's like you have to have the foresight to go, okay, I've run a field, I've run a marketing organization, I've built brands, I've run support organizations, I know what support costs look like. You have to have enough of that to sort of scaffold up this picture. then that's why, ultimately, Smog more successful because it was our umpteenth company and we had done them together. There's other things aside, there's a ton of luck. I'm not going to say it.
Look, it's almost all luck. But the parts that weren't luck that were different from the prior one, the prior company that we did that was just super cool, violated probably half the physics. Like it was the funnest thing we'd ever done. But it just violated so many things. In the end, we went to go ship and build and launch it. We're like, God, none of this is going to work, is it? It's like, God, too bad, because the product was ridiculous.
Eruj (13:24.507)
And I think a lot of people luck their way, you so many startups just sort of luck their way into physics that work, right? And we were never lucky enough. So I think when we did Splunk, we were just tired of having no money. I know I shouldn't say that, we did okay. But we wanted one to be successful. So we're like, oh shit, we gotta make one that's successful.
Kostas (13:37.838)
You
Kostas (13:42.632)
I think that's a great way of describing the journey of building something successful, where success can be many different things. That's what I love from what you say. Success can be, as you said, my god, this technology just feels amazing, right? It does crazy things. Now it might fail monetarily, and that's another way of putting success there. But I like this mental model of physics, and I want to ask you...
Eruj (13:44.795)
Thank
Kostas (14:11.662)
Is it always possible to engineer yourself through this physics? Or in some ideas, there are, let's say, interesting reasons that they will never work. I'll give an example what I mean. Many times you hear the term, we were too early to market, right? So you build something, sure it's amazing and like doesn't scale to the point that it could scale
Eruj (14:31.813)
Mm-hmm.
Kostas (14:41.07)
15 years later, think like, Hiroku is like a great example of that, right? Amazing technology when it came out. Okay. Salesforce all happened to them, but still like it does feel like too, it was like kind of like a too early and time. Usually it's not something you can time travel, right? Like that's not a law of physics that you can break. so tell us a little bit about that. and
how you build intuition around these things. If at the end, know, like someone is out there and thinks, hey, I want to start something, I have this idea, is the physics going to work or not? Like, how do you reason about these things?
Eruj (15:21.122)
That's a great question. Well, first of all, we have to acknowledge that you probably can't, right? I mean, you're going to do the best you can. Like some things we can aim for and nail, right? And then other things I think you're going to have to have some luck and intuition. So a lot of it's luck, right? mean, it's just the time. So one of the things I've always said about a startup company is you stay really small. What your job is is you're like a...
perhaps even a predator, right? You're trying to stay small and lean until the time's right. You want to stay in business long enough for your time component to work, right? Because let's just say that you don't know if it's five or 15. That's a bit extreme. But nonetheless, as a startup company, when you're really small, your job's to stay. You're trying to stay on the runway as long as you can, right? And this is a difference I have, I think, with a lot of VCs, which is I don't really like fail fast in the traditional model. I like staying on the runway.
as long as I can, right? And run the business on the runway and wait and wait and wait and then pull up when it's time. A lot of VCs just say, want you to build something and ship it, build something and ship it. I'm telling you that's, this is not my opinion. I think it works fine, but I'm of the opinion time is that time is the thing, right?
And you're really waiting. You're trying to be around long enough, thinking about it and working on these physics and then tell the times right and then pull the stick. Right. And my current company is very much an example of that. I've just, over the years, I've learned that, look, I'm happier when I'm very small. It's a very lean team. We can change direction pretty quickly. We can stay on the ground. And then when the time's right, you just yank on the stick. And because you pull too early, right? Too many things can go wrong. So think the timing one's interesting. You just don't know. Right.
You either get lucky or you don't. You either time it right or you don't. I will give an example of the Splunk. So the Splunk, the only thing I will say around the Splunk thing was, I don't know how this applies to answering your question. So the Splunk thing that sort of hit home once we get through the physics was the hypothesis, we were doing Google for machine data. So there's at this point in time,
Eruj (17:29.915)
I had briefly run the search engine at InfoSeq, was a of early competitor, a precursor to Google. So Google comes along, and it's obvious that you need this way of accessing this ever-growing amount of human-generated content. And for us, the timing on Splunk was, holy crap, there's at least 10 times as much machine-generated content. And I have crap. Right?
And so our timing hypothesis was that is we could look at this wall coming up, holy crap. There's way more machine generated content than we can deal with. It's growing. It's It's going two orders of magnitude faster than human generated content, right? Because as you know, a data plumber, I type a character and God knows what happens, right? So our hypothesis there was we could see the coming of hitting a wall. Whether that wall, know, spunk was probably a little early.
but not too early, the big data movement. We were just right on the cusp of big data. And that was, you could just see it coming. But some of that was also staying on the ground a little bit long enough. Because if we had had our solution when we had thought of it, if I could wave a magic wand in year one, it would have been too early. We wouldn't have been to sell to customers. So I think some of this is, again, that timing of I can see the wall approaching.
stay in business as long as I can, scale appropriately at the speed which the business can grow at. But that one was pretty obvious. You could just see, we had Grep, we had Google, we had human-generated content, and we had machine-generated content. End of story, right? was pretty obvious at that point.
Kostas (19:08.46)
Yeah. So you said that Splunk was not your first attempt to build something, right? So I'm sure there were physics that you discovered before and hopefully you applied, but not like to move faster and like probably make less mistakes when it comes to make the company financially successful. What was something though?
like a new law of physics that you learn through Splunk experience because you always learn something.
Eruj (19:40.985)
Yeah, totally, totally. Well, there are a bunch of them. I think the one that I take away, I think we knew going in, but didn't know, is the importance of brand. So I've become a huge fan of brand. The biggest lever, it's sort of a meta lever, right? It's not like this little lever. It's this huge lever you have as brand. So I will say that think the biggest lever in a business is brand. And what Splunk did early on,
Because one of the things we realized going into Splunk is this, how do you get attention in an industry that's kind of noisy? And the IT space is pretty noisy. And one of the physics of Splunk that was really important is, look, we need to build something. The person who sees it at this time installs it, uses it, and pays for it.
that can't be anything other than one person. So some of us have operated in the past, the person who sees it is different, the person who has to buy it is different, the person who has to install it, right? So we just said, look, the physics of our business is there's one person, sees, uses, right? So we're like, how do you get in front of this person? And we just realized, I think it was brand that just did this. And if you remember the early Sprunk brand.
We were a t-shirt company. I've never seen our t-shirt video. We just thought of ourselves as t-shirts. How do we think of our... Our t-shirts never had our logo. We just had really stupid sayings like, take the SH out of IT and stuff. So we were consciously trying to build a brand that would be an enigma that our customers...
wanted to be associated with, if that makes sense. We wanted to be more like a consumer brand in our eyes of our customers. They wanted to wear our t-shirts. They wanted to come to our trade. It wasn't the job. was just this. So I the biggest lever that I learned there, and it was really hard work. It was probably the hardest, like just from how the hell do we do it, was this lever of brand and being really good at
Eruj (21:40.923)
whatever that thing is that makes your customers want to be part of your movement, success, whatever. And I know that sounds stupid because we're just flipping IT software. But I we did a pretty good job with the sort of the Kilt-Wearing dudes in the basement, right? I think we just really landed that, know, landed a brand that spoke to them very clearly.
I think over time it is hard to scale that. The one lesson there is I've just become a huge fan of brand is my biggest lever. It's the biggest gear I have. I can't change it once I have. It's the hardest one. I can throw my code base over and start over again. I can't start my brand over again. I just can't. So there's this narrative to brand that is probably as complicated as a product launch and a product roadmap that goes from first word
Kostas (22:13.57)
Yeah.
Kostas (22:20.014)
100%.
Kostas (22:30.798)
Mm-hmm.
Eruj (22:34.989)
color to fan, if you will, that has to be considered very carefully. And it's not an afterthought, right? It's just not. It's part of this because you can't change that path or journey someone's on. I you have to start over again, right? You just have to throw it out.
Kostas (22:44.141)
Yeah.
Eruj (22:53.463)
you might as well just reset and start another company because I can change my UI and lots of things. I can't change my brand. So be super cognizant of it and spend a lot of time on it. And it's not a lot of time in the sense like coatings, you're sitting in front of your computer for hours on end. It's more of this long running wave that you have to be super meticulous about and careful and understanding and then consciously understanding what for your customers, what that brand
is supposed to do. It's going be really hard to explain. Honestly, it's just, I think, the most interesting and biggest lever for go-to-market.
Kostas (23:31.214)
Yeah, a hundred percent. And I'm super happy that you mentioned brand actually and branding because it's, in my experience is also like extremely important as you said, but it's one of these things that I think it's also extremely hard for people, especially like engineers to understand just because it sounds very abstract in a way or build on intuition.
Eruj (23:43.002)
Okay.
Kostas (24:00.322)
But there has to be a systematic way to approach it, right? And by the way, just to give like an example of how important brand is, is one of the most, you're an early stage company, of course, like one of the things that you want to do is like as you get like design partners or your first customers, you want to put your logo out there, right? Because you don't have a brand, so you want their brand to reflect on you, right?
And usually when you start getting these legal documents from the serious big companies with big brands, one of the most difficult things to remove from there is about using their logo. And it sounds silly, right? Because the logo is like you search on Google and you can find anywhere, but they're like super, super strict about that. And I think it's like a good evidence of like how important and how valuable and how well guarded it is.
I think like one of the extreme cases of that is Apple, right? Which even like getting someone from Apple to give a talk somewhere, it's, I don't know like how many permissions they have to get from that, right? But what I like from what you say is that you kind of give like a systematic approach of defining and working with brands. So tell us a little bit more about that because it's super, super fascinating and extremely, extremely useful.
Eruj (25:18.095)
Yes.
I think your point is good. you know what's funny is when we, and this is a change in the industry, so when we started, we were doing startup companies in the 90s.
You know, we were just nerds building stuff. It was a very different world back then, right? There was no Silicon Valley in the sense that people knew what it was. Like there was pre-Facebook and the movies and all that crap, right? So it was just a nerdy place. The brand didn't make it. It wasn't brand, whatever. Right? I think, and even though I worked at Apple and Disney, I think even the startup companies were. But I think that in this day and age with, it's just different. There's so much info. Like if I were to go to market today and try and compete with an IT company,
I don't care how good your shit is, right? I don't care how good it is. It's just nearly impossible because it's just, there's too much. There's, you know, and you're in this industry, like there's supply and demand. There's too many VCs and there's too many startups and there's too many ideas and there's too much stuff. So it's like, okay, your biggest problem, your number one problem is like, okay, let's just assume you built something cool. Now what? What you gonna do? Right? If you can't answer that question for me,
why are you doing this? Because I don't care. It just doesn't matter. this is where I brand comes into it is brand is this journey you go on.
Eruj (26:40.207)
that helps you understand how to differentiate, get in front of, and build sort of this mental picture of how your thing is interesting or different. It's not shouting, right? It's a very complex way of taking something and say, look, you need to stand out from something else in a very unique way. And if you can't do that, then stop.
because you're going to build something that'll just be one of a thousand things that you can't differentiate on. And I think there's the go-to-market side of how you get there. Like, okay, well, we're going to communicate through podcasts or whatever. But I think brand is the thing that sits above it. That's like, look, if you can't tell me in a picture of branding, how this thing is uniquely different from a feeling standpoint, it's not just words, right? We have too many words. Like I tell you, you can go read every website in a particular category of business and they're all going to sound the same, right?
they're not going to be any differentiated. So what is the thing that's uniquely different between companies that sort of get some traction early on? it's, just gonna, I use brand in quotes. It's a bigger, I use brand bigger than I think most people do, but it's that whole mental, it's how your customers or everybody perceives this feeling of what you're doing, why you're doing your mission. It's a very complicated thing and it's critical.
It's the thing, I think, that is your test if you should do this or not. again, there are very few products that are, I if it's too different, then it's questionable whether the product is valuable. Let's just assume for the moment you have competition. It can be the legacy. I typically it's legacy, but then let's just assume you have some other startups.
Tell me how you different. And by the way, that's going to be the beginning of brand and that better come down to a feeling and not just words. Because if you're telling me your differences on words, I'm going to pull up a competitor and say your words are the same as those words. So you're not going to Is your technology better? Sure. I don't care. But they've got a field of 7,000 sales folks. Doesn't matter. It's going to take too long for you to get into that. You may have a better product. I don't care. It's just going to take too long. And you're going to run out of money.
Eruj (28:59.487)
So it's again, it's thinking about that to me is just the crux of the problem, right? Because I can come up with five ideas today that I think are killer ideas. I can, and you can too, right? But they're all gonna be challenged with this problem. Again, it goes back to sort of the physics, but this is that sort of meta-letter of physics that says, just, what is this thing that we're doing from a feeling standpoint that's gonna attract?
somebody and it's very complicated. But if I were to do one thing, that's just the one thing I would spend my time on. And that's where you like our current company is just spending a lot of time on how do you picture this thing in a way that's uniquely different because you don't. We just have too many startup companies. We got too many VCs. We got too many startup companies. We got too much noise. Right. We just it's like, what are you going to do? And other than luck. Right. I sometimes there's a lucky company that just just nails it. And you're like, holy shit, how'd they do that? And I'm like,
Look, maybe. don't I don't And also, think, you know, I will say that it is a side note for another podcast is how to use LLMs for brand. think LLMs have made for an engineer this brand thing a little more interesting. It's not for today, but I do think there are tools now where you would have to have an agency. have to go hire Google. We have to go hire somebody. think now technologies probably have some tools available to them to help.
Kostas (29:58.011)
I love that.
Eruj (30:24.603)
do a lot of this work, or at least have the framework, they didn't have before. Because think LLMs are actually pretty good at decomposing brands and what makes brands successful, and how to identify it, and how does yours compare. Like I'm just saying, it's one thing I think they're not bad at. There's enough framework there that you can have a dialogue with an LLM that's a nerd about what all this means, and how do I fit into this, and how am I different. And if it comes back and says, you're no different, you're like, oh, shit, OK, I'm no different.
Kostas (30:53.762)
No, that's amazing. By the way, it just clicked in my mind. Last night, I saw that OpenAI announced GPT 4.5, which by the way, the price is ridiculous. The very, very interesting thing is that if you go to the product page of this new model, one of the things that they sell about it is that it has more empathy. And they have an example where
Eruj (31:05.936)
Yeah.
Eruj (31:18.192)
Yeah.
Kostas (31:22.158)
Someone puts a prompt there. says I had a really bad day today and the previous one would be that sounds Unfortunate. Let me give you a list of things you can do to improve, right? And then they have the new one which is more empathetic. It's like I feel you I'm and I think And the reason I'm saying that's because you use What I love is that it boils down like to a feeling as you said
Eruj (31:39.109)
Yeah.
Kostas (31:52.078)
And I think there's tremendous amount of both wisdom and importance in that when you try to build, even in B2B, because I think people traditionally think that, okay, brand and feelings and all that stuff is like, things of B2C, Apple type of companies, da da da da. But I think it's...
Eruj (32:15.615)
Splunk sold to people, I mean, think about it. Who would ever want to apply Splunk to a bunch of IT nerds? Like they're using Grep. What does brand have anything to do with it? They wear a kilt, they're in the basement. It's critical. I think you're right. think the giving technical folks the tools, I think we can all, some people love LLMs and some people don't. I think one of the great things about why we...
Kostas (32:26.509)
Yeah.
Eruj (32:42.971)
as technologists love, is it begins to help us through some of those things, right, that we're not particularly good at, right? I have some deficiencies as a human. I don't know, I'm, whatever, I'm less sensitive. Like I have some shortcomings, right? I think an LLM can help me overcome some of those shortcomings. It can help me appreciate the difference, right? So I think on the brand side, I think an LLM tool is a great way to get, yeah, maybe empathy's the right word. Like it's just really digging into.
how is what I'm building fit into this massive world of all this stuff? And I think before you would have a hard time. You'd have to hire some agency that did some, you know, come in and do therapy on the organization for six weeks. Whereas now I think you can probably muddle your way through a bit. So I would encourage you to, know, companies that are doing this now, just be, just as you have a product roadmap, just as you have a vision for your product, just to spend a bunch of time and thinking about brand and go-to-market and customer and...
Kostas (33:25.198)
you
Eruj (33:41.007)
from day one, because the odds are that when you look at that, it needs to change the product. Does that make sense? Like the physics, this goes back to the physics question. There's probably something about the physics, I don't clue brand, that mean you have to change your product. So things like getting started experience, like, you know, no one ever thinks about how I get started with something. They build the version one and then like, how the hell are we gonna get people started on something? Wrong, right? I'm not.
Kostas (34:03.054)
Yeah. One last question from me and then I want to give the microphone to Nitei because I feel a little bit bad. monopolized the conversation.
Eruj (34:11.809)
I'll talk all day long, too, by the way. You're talking to me. It's like, we do it. We can do an eight hour podcast. I don't care.
Kostas (34:19.246)
one of the things that, I always, it always comes in my mind when I think about Splank, is actually two things. all the, long as I remember myself interacting with like Silicon Valley and this is here is everyone was always trying to kill Splank somehow. Like it had to happen. Like a new company had to come out there and somehow dethrone Splank. It never happened at least like.
If it happened, happens not through building new companies that they did that.
Eruj (34:54.235)
There's companies that have done a good job. By the way, I have a pretty good answer for that question back to the physics. So I do have an answer for the physics on that.
Kostas (35:00.822)
Uh-huh. And just to add something to that, there is a quote from an actual user of Splunk that I keep remembering. And the guy was saying, well, you know what? I hate how expensive Splunk is, but there's no product with an experience like Splunk.
Eruj (35:23.589)
ding ding ding ding ding ding ding ding. So physics again, physics again, right? I mean, that's super important. Part of the physics is like, I like to be expensive. Expensive is the wrong word. I want to be expensive.
Kostas (35:25.567)
yeah.
Kostas (35:35.886)
So how do you engineer that? Because to be honest, don't know any other, you're like, developers are like grumpy people. They like to complain about stuff. So for someone to sing, like make a statement like that, I, I find it very unique and like important. So please share the secret sauce. How we can do that.
Eruj (35:55.195)
Totally. I don't think there's a secret sauce to this. think there is the... Look, one of the physics is you want to... I think you want to be expensive. So there was a whole... Again, we could spend an hour on open source versus expensive prices versus... I mean, we could spend... On pricing alone, we could spend an entire podcast. In the case of Splunk, and again, what we were trying to optimize for, which is different than...
Like he said, if I want to optimize doing something cool in tech, I'm going to make it open source. just am. If my goal is to, and I think there's a real good reason to do that, that motivates me, I'll work hard, I'll spend my nights on it. If your goal is to maximize sort of revenue, it's a different question. And I think in the case of Splunk, it was really maximizing revenue, right? And one those things was to really live on that borderline of too expensive. You're trying to pin that.
I'm too expensive. You go too far. It's not going to work. You're too cheap. I it's just a simple supply and demand pricing, for us, it's OK to be expensive. I will say the answer to your question around why it was... So when you build a company, you want to say, look, I have... It's going to take me five years to build. It should take you five years to build what you're building. OK?
And the trick you have is at the end of five years, you want it to be the core of what you're doing, sufficiently challenging enough that someone can't come along in six months and do it. Back to the basics. So is what I'm doing incrementally getting better and better and more difficult such that when I am successful, $100 million in revenue, $1 billion in revenue, that it isn't something easy to do?
And I think in the case of Splunk, that's a conscious, what we call sort of a swim lane. This is like, want to build a swim lane where, look, I can get to version one and version two and version three and version four. But when I get to version five, it's nearly impossible for someone to come along quickly and rebuild it. And that's because we're constantly investing in IP that is more and more challenging and hard.
Eruj (37:55.971)
And part of the defense is like there's lots of customers and lots of data. I think that's a different one, which is like land grab, lots of customers, big field. It's hard to compete against a big field, right? It just is. They're going to outsell every startup company. But I think on the technology side, what technology people should do is look, you want to have a very clear roadmap that says, I am building something. It's going to take five years because that's how long it's going take you be successful, right? You're not going to be successful before then. But if I am successful at year five, it should be really fucking hard to go build this.
such that someone can't come along in six months and take me out. And in particular, a startup company, don't think companies or whatever. But you really want to end up a place that someone, and I think you can, like new technology can come along, but you can't foresee that. So when you're building your sort of technical roadmap, it's incriminating, it should be hard and you should be building stuff that is a technical mode. You have data modes and customer modes and everything, but I like the technical mode. Otherwise, you're right. If you're successful, someone's going to come along, you're to have 10 startup companies in a week. It's like, oh, those guys made a hundred million dollars this quarter. Okay, great.
But it should be so hard that like crap, I can't do that. And I think Splunk was a bunch of that, right? was like, it's just, there's a constant evolution of a technology that makes it more and more difficult to compete. And if you come to me and give me your business plan and you can't show me that roadmap to where in year five, this is really hard to compete against, then go back to drawing board because it's violating one of the physics, right? Sorry.
mean, there's a ton of other physics in there, but that one in particular is critical, because otherwise you're right, you'll be successful and you'll have 10 competitors immediately. It's a good question though.
Nitay (39:33.767)
What was a, no, that's really interesting insight. I'm curious, you mentioned kind of the physics and clearly you guys thought a lot about brand and thought a lot about, you know, we want to be expensive, but we're gonna own it and we're gonna, people are gonna pay for it because they're gonna get so much value. What was it maybe in some of the early days, either it's Blanc or even some of the starters before some big physics or brand mistake that you made that you think back?
And you're like, okay, at that point something became clear, at that point we had to pivot or whatever.
Eruj (40:09.883)
Totally, think...
I think one of the mistakes I made, because I did have time at Disney and Apple, which were mature companies, I sort of misunderstood. Godfrey Sullivan, who was the CEO of Spongebob for a while, had really interested me in putting this, which I think is great, is you need to be a tool before you can be a solution, before you can be a platform. So make a hammer. Just make a fucking hammer.
Right? So I'm coming out of Apple and Disney, and I'm thinking platform. Right? I'm just thinking platform. I'm thinking I'm going to build a platform. Because guess what? When you're Apple and Disney, you can't. Right? So we go build platform and look at platforms just too hard. Right? So I think the big takeaway was, look, just go build a hammer. If you can get people to use your hammer and love your hammer, right, you can eventually get to a tool chest. Right? But.
Don't start with tool chest. So the lesson, the biggest physics lesson, I think, was really that journey from tool to solution. Don't even think of solution too soon, right? Because even that, I mean, I think you should have it in your physics or your roadmap. Like I need to get to platform, but I build a, what's the stupidest tool I can give away first. Like just, again, if you say I'm just building a tool and you don't have the long-term, here's how I'm gonna be a platform in five years, also go back to the drawing board.
But you got to start with a hammer. Go from a hammer. Show me how you're to go from a hammer to an entire garage, but I want to know the hammer first. And by the way, you can even start with a look. I don't know if it's the hammer or the screwdriver or the wrench, but I'm going to test these three and we're just going to pick one. It's either going to be the hammer or the tool. I'm not going to get all three. I'm just not going to happen. So I that was one of the big physics ones, because we were just aiming. We'd come out of...
Eruj (41:58.679)
organizations where they were big, right? So you're like, okay, think when, I think the takeaway in the end was what's the smallest, simplest.
part of the journey. Again, you have to map that journey. can't ignore the how I'm going to get back to the brand again. I can't ignore my five-year plan, but I got to start with something stupid simple. That was one of them. I think the getting started motion for me is probably, besides the brand piece, which I think we covered, which is like, I can't even talk to my customer or they're going to see me. It's a non-starter, but we talked about that one. The other one is just, I think getting started is with something is just.
Fascinating, right? And if you take a look at which things work really well for, I don't care if it's enterprise or consumer, right? It's just that first experience. And what's funny is I think often people will build the hammer. Maybe this analogy is not gonna work, right? They're gonna build something and then come back and go, okay, we need to figure out how people get started with it. It's like, no, no, no, no. Start with how they get started with it.
Because the odds are your product will have to change when you think about how they get started. Does that make sense? It's like you build something and go, OK, it's done. And then you'll stick it in front someone that's like, well, how do I get started with this? And you're like, oh, shit, I need to go rethink a bunch of things. So I think there's this, let's call it time 0 to time 1, where time 1 is they've actually started to use it. It's that 0 to 1 thing that's probably the most fascinating. Because in all cases, I think, that we had, when we looked at 0 to 1, it changed 1. You're just like, oh, shit, we need to rethink.
And because you spend your entire day with whatever it is you're building, you totally get it. I know how to set it up. I know how to use it. I know what it's for. But there's that zero to one part of that process that likely impacts the way you should think of your product and how you should design it. So those are the two things. One is, look, it should just be a tool. If you come out of a big company, just build a tool. You'll have time. But then map. Tell me how you can get the platform.
Eruj (44:00.027)
But then importantly, it's that zero to one from first word color picture to first use and really being mindful. then the odds are, like I said, it's going to change your product. If it isn't, you're either lucky or you're doing it wrong. I don't think there's a way to have a product that's getting started. I mean, it can't be too different from using the thing. You can't have some setup thing.
And then how does that first use? What's the first thing that happens? The k-value is another one. Oh, never do a business that doesn't have a k-value, right? If it doesn't have one, just go back to the drawing board. So that means that for every person I get to use it, I need 1.1 more, preferably 1.3, right? So I don't care if it's invitation, I don't care it's word of mouth, I don't care if it's collaboration, I don't care if it's social.
If you can't show me a 1.1 or better on the k-value, start over again. Just because it doesn't work. can't. Or let's put this way. It gets a lot easier if a person who uses or buys your thing gets another person to use it or buy it. So that's another lesson we learned. It's one-to-one. If I have to get every customer one-to-one through my mechanism, that sucks. If whatever I'm building has an inherent
K value built into it, it gets way easier. So I wouldn't do a product that didn't have a significant K. I just wouldn't. It's just too hard. Does that make sense? I mean, so.
Nitay (45:35.972)
Yeah, absolutely. And there's some gold anecdotes there. I remember even in my own journey, I've heard different versions of that as well. So for example, my early days, I remember my own company, we pitched to VCs, we get funding, everything's amazing. We're going to build this beautiful platform, et cetera. I walked into the board meeting. It's like three months after post investment, we've started building a bunch of stuff. I give this big presentation and my investor goes, stop.
Eruj (45:42.267)
Thank
Nitay (46:05.38)
just stop right there. I'm like, what's wrong? He's like, stop saying the word platform. I don't want to hear that word anymore. And I look around and I a little dumbfounded like, and I literally said, you know what you invested in, right? Like we had this conversation. he looks at me he says, you know, you're a platform. I know you're a platform. Nobody else should know that yet. He's like, we're going to build here is a solution. We're going to build. We're going to solve one problem. Let them wake up two years from now and realize that they bought a platform.
Eruj (46:09.645)
Thank
Eruj (46:28.303)
Yeah.
Nitay (46:34.113)
Let them come to you in time.
Eruj (46:34.435)
I'd go back to tool. I still think tool before solution. think even solution is a bit, because the solution again has a promise that you likely can't deliver in day one. I'm just, honest, right? Which is you ship stuff. You remember your first customers, it didn't work. It just doesn't work. It doesn't work for a while. So the benefits of having a hammer that doesn't work is very different than having a garage that doesn't work, right? It's like, okay, I can kind of noodle. So I think you're absolutely right. A platform is when someone tells you platform and you design a platform. My point is just like, it's even simpler.
Nitay (46:38.347)
or in tool, exactly.
Nitay (46:43.959)
Mm-hmm.
Eruj (47:04.495)
You need to build something that has a promise that's very simple, because you're probably not going to get it right. It's going to be version. I still think it takes a long time to get it right. So just build something stupid. again, with mindful of that thing. But you're right. Yeah, don't use the word platform. You can.
Nitay (47:19.907)
As a question I have here, the other bit I just want to touch on, we had actually another guest on a different recording that touched on it too, particularly where he runs a DevTools newsletter and he reviews thousands of DevTools. And we asked him, what's the number one thing that tells you whether this DevTool is good versus that one? And he said, I can tell in about, what was it, like 30 seconds. And I said, how? He said, I just look at the documentation. That's it.
So just the onboarding experience, as you said, how good, how much are you actually going to walk me through? How do I start to use this? So the question I have to you is, so you're building this tool, you have this kind of first thing, this MVP first level product I'm going to build. But at the same time, you're thinking about the physics, you're thinking about the brand, you're thinking about the onboarding experience, you're thinking, want to charge a ton. But all you've got is this tool.
How do you marry between those two, right? Those first few customers that you want to turn into those, you know, six, seven, eight figure deals. You want to represent your brand, but all you've got is this wrench in front of them.
Eruj (48:28.266)
So I don't think it's that complicated, but I may not be able to describe it. I think it's again, it's a it's just a roadmap and a journey you need to get through. I actually don't think that's again, if you can see how your. Let's particularly pick on I think the hardest parts in all of this again, come back to.
your go-to-market. So there's product market fit, which is a journey in its own, right? And I think product market fit is a well understood, look, we just need to go do this. But I think the go-to-market is different than product market fit. I can have really good product market fit and still fail, right? Because I can't get in front of some. It's like, if you can't get in front of them, like you can't run Facebook ads, you can't run Google ads, it's not going to work. So I guess the one where I'm interested, like where I find your...
this thing particularly interesting is really being thinking about how do I go from three people who love me to nine people who love me, right? To 27 people who love me and where I'm doing as little work as possible to have that happen, right? Because I think the trick there is you can't, I think a lot of that is outside of your control. Like you can't force, I can't push on that rope, right?
And people have that product-led growth. I think that's fine. It's a bit boring. But again, your product can't be a rope you push on. It just can't. It needs to be a rope that gets pulled. So what's pulling the rope? I can put lots of rope out there, and I can try and do a lot. But something's got to pull. So what is the pull that happens? And I think there's a lot of ways to skin that one.
mean, there's a ton of ways to scale it, but you have to have an answer. I don't know exactly answering your question. just saying, think there's a, what to be mindful of in that phase is, look, if I do get a happy customer, if I get one happy customer, how do I get more? And then again, this is probably pretty well understood. I just think there's things you can do within the product that help make it a lot better. We already, we know this a lot, right? We want product-led growth. So the product needs to grow.
Nitay (50:42.188)
What was the, to that point, what was, if you remember, the kind of first tool you guys built at Splunk and how did you weave in that K value and growth and so forth into it?
Eruj (50:53.989)
So I the K value, we didn't. That's one of the ones I learned after the fact. So the thing the tool was was I think the, and I give you an example for that company, and I think we can just sort of distill that into something simple is in the case of Splunk, in its stupidest form is as you would say, take a file and 10 seconds later, you can Google it.
I think the tool here that was the... We wanted to go do a million things, but what we did is you said, okay, give me a file and I'll put a Google front end on it.
And you can search for an IP address, and we'll highlight an IP address. Simple hammer. That was the simplest form of a hammer. Did we want it to do more? Yes. we have, did actually build it to do more? Probably. But it had to be constrained to, it had to be so narrowly tooled down. So we have a picture that is this, every sort of journey of a startup company goes through this, a good one, goes through on the time left-hand side.
It starts with a pretty wide funnel of ideas and concepts and stuff. And your goal is to narrow that funnel as much as you possibly can. And then the funnel opens back up the other side. So it kind of goes wide funnel to skinny pipe to wide funnel. And the trick with your company is you're going through to a launch as you're trying to get that, you're trying to find that narrow funnel. And for us, in the case of Splunk, was, we didn't think of.
a lot of things until we went through that funneling process to try and drive it down to the stupidest, simplest, and it became troubleshooting for very specific use cases was the thing that we funneled it down to. Even though we started with, it's for marketing and sales and all this other shit, right? Because that's when you ask people, why were they looking at log files, right? It's, well, I'm trying to get my sales numbers are volume. It was a million reasons. But in the end, the thing that we funneled down to was very specific.
Eruj (52:48.763)
I don't know if I'm answering your question, but I think in that case, was just you're going from really wide down to really narrow. And then you go back out wide again, and you're going back out wide again as fast as the business can allow for widening. Some companies never have to get that wide. They can go $100 million on a very simple little pipe. Other things have to get wider pretty quick.
Nitay (53:16.384)
Oh, that makes a lot of sense. No, no, I think that makes a lot of sense. I think you did answer it. think that narrowing of use case and that narrowing of the problem that you're solving and kind of walking between building a tool to a solution to a platform while at the same time really, really honing in on the use case and the capability and the customer too that you're solving a problem for.
Eruj (53:16.676)
I don't remember the question, sorry.
Eruj (53:41.435)
Three to nine to 27. mean, how do you get, it's the hardest part is, again, you can build the coolest thing in the world. Do know how much cool shit's getting built today that's never gonna go anywhere? Right, it's just because it hasn't thought about, well, shit, how do I get, how do I go, like, because you can't brute force it. You don't have the money to brute force it, right? You don't have the marketing revenue, I the marketing spend.
You know, of course, if you're Apple, like I can build really crappy goggles, right? And get them on market, right? But how the hell do do that if you're like, okay, your marketing budget's 50 grand. What are you going to do? And again, exactly. And, know, and that's why you have a lot of feelings. Certainly the enterprise world goes to like the elephants right away. It's like, if you see something that's got like, I got seven customers all paying me a hundred grand.
Nitay (54:17.512)
You've only got so many shots on goals as a startup, you've got to use them wisely.
Eruj (54:31.611)
It's like, no, no, no, no, God, this is just not going to work. It's just, I mean, it may work, but the physics of it is just flawed. There's too many, you just can't make that work. So you want to see a physics that's very different than that. Because I can always find people. Yeah, go ahead. Sorry.
Nitay (54:48.608)
We're good, we're good, sorry.
We can certainly do an entire episode just on that. But I'd love to maybe fast forward a bit. Tell us about, given all of these great lessons that you've had over the years, what are you building now? What's the cool stuff you're working on? And tell us some of the physics of it.
Eruj (55:07.259)
Oh, I can't tell you. Let's see how to describe this. So we've been working on this project for a while. It's called Bestimer. There's no public information that's accurate. I don't even know if there's internal information that's accurate. It's kind of the Splunk crew. I think we're taking a lot of the lessons we learned at Splunk and applying them to hopefully a bigger problem and bigger market. think, what's funny is we found Splunk to be disappointing in its success.
So I think we're trying to find the physics of something that is significantly more interesting. It's taking us a long time. And I'm again, I'm in no hurry. I want to keep the plane on the ground as long as I can. We're about 14 people. We're building lots of really cool IP. We're beginning that roadmap of defensible IP such that someone can't come in and do what I'm doing. But we're sort of replaying this book with a bit of a twist. One, think AI is in a place where
I think we started the call before the podcast saying, look, I couldn't be on the sidelines right now. I'm too old. I shouldn't be doing this. I don't need to do this. I couldn't sit on the sidelines right now. Right. This is, it's just fucking nuts. So I think what's, what's, what we're capable of doing right now is mind blowing. And I think we're just applying it to a particular problem. the, the physics of one of our problems is we're trying to do something that hasn't really changed much in 30 years.
And we're trying to do something very data heavy. So sort of a very large volume of data against a problem set of incumbent tools, which all kind of haven't changed much. But doing it for a larger audience than Splunk. But it is a very Splunk-like play in the end. I tell you more. mean, can, it's sort of, we're borrowing.
Some, if anything, I would say at this point, we just confuse people. know, in another four or five months, I'll have more information on what we're doing. We're a ways from, we're probably two or three months away from having something we're launching. But it is the Splunk crew redoing something with all the Splunk lessons in modern AI applied to a larger audience. Borrowing from some different technologies like genomics.
Eruj (57:29.051)
I'm trying to blend some stuff. think AI is just, it's right, I think what's interesting right now is it's a really right time to take what, let's just say these LLMs do, and they're just the foundation for shit, right? They're just the foundation. It's just concrete, right? It's not the thing. And everyone types in chats and does shit, but it's just great. It's concrete. It's not the thing. For me, right? I love, trust me. But.
Nitay (57:55.516)
It's that enabler that allows you to build the skyscraper.
Eruj (57:58.107)
It is. So you're just looking at going, holy crap, I could have like, I think a lot of what you see today is like, it would have taken me years to build. I would probably couldn't, you couldn't have built it. Sorry. So there's just no way you could, there's no way a company of 14 people could have done it. Maybe Google could have done it, but we could have never done it. But now all of sudden, thanks to the ability of where, quote, just AI in general is going, is that we can do stuff that you could have never done at a kind of a cost price.
Which is reasonable, right? mean, it's, it's the price of using this stuff is going down and down and down and down. And we're getting open source for you. It's just like, it's, it's just changing to the point where, wow, four years ago, I could have never done this ever. But, but now I can do it and I can do it with a pretty small team and keep the team small. And then just really figuring out what's the IP relative to, I think the hard part right now is, there's so much going on. It's like, okay, where is everyone else going?
Nitay (58:41.309)
Absolutely.
Eruj (58:57.211)
But we have a very unique angle. Yeah, it's like, it's just, in some of this is a crap shoot. Honestly, if I were an investor, it's like, you can't predict where all this stuff is going right now. So you just, you're going to have to just kind of go with it. But in this particular case, we're, we're building is whacked. It's cool. It seems to fit all the physics of everything we need. I'm in better, much better physics than, than Splunk in some cases. Some of the things I think we learned about Splunk, about that K value, right. Is, is, has been fixed in this model.
Nitay (58:57.499)
Ready, focus.
Eruj (59:27.147)
I think the defensibility of the moat's a little bit stronger in this too.
I would say more, but I think anything I would say would be wrong and or not helpful. But I think four months. Four months we'll have a public launch and people will get it. And again, it'll look like a tool at first. Again, remember this conversation. It'll look like a hammer. But we see an entire garage, right? We're just giving you the hammer.
Nitay (59:55.431)
Well, we look forward to seeing the hammer and we'll definitely have you back on for five months to continue. If I may, one last question, because you said something really interesting there that I think would surprise a lot of people. then clearly there's a lot of inside baseball there, which is you said we were disappointed in the success of Splunk. What does that mean?
Eruj (01:00:16.473)
Huh.
I, so that's a really complicated question. I think the disappointment comes from.
Probably the number one disappointment is, is I think a lot of the the zeitgeist of the team for many reasons either left or was displaced. This goes back to that, you know, like what were we talking about a couple of weeks ago about, you know, founders, CEOs staying on board, what was the whole, you know, that whole thread we had about, you know, founders, CEOs staying on board longer and yeah, we...
Nitay (01:00:53.917)
for a long time as the company scales.
Eruj (01:00:56.495)
Yeah, forget the whole thing. For like three weeks there, that's all we could talk about. The company just sort of lost its way in bringing in a whole bunch of new people. I think there's a lot of reasons we could go into this is I think being in public company is a very challenging thing. I'm just going to admit, I think public companies have a really hard time. Right? You have to go quarter to quarter. And there's times, at times you're doing very unnatural things to go quarter to quarter.
And it takes a lot of courage to do the right thing long-term in spite of that. One thing that you learn really quickly is the tail wags the dog and the field wags the dog. It just does. The sales organization in the end is completely driving decision-making because you have to make the quarter and you have to make the year. And what ends up happening in there is that without really strong
visionary product leadership, the business model begins to drive it all. And I think in this case, Splunk when it ended up happening, is yeah, it started making a lot of money. And it grew and grew and grew and grew. And the ability for the product organization to keep true to what the long-term thing would have enabled maybe a hundred billion dollar business instead of whatever they sold at became constrained by sort of a local maxima. It's constantly the local maxima of the quarter.
driving decision-making and the inability for me and other people to fight that, right? You just, the local maximum drove everything in the end and to the point where it did.
Nitay (01:02:34.159)
The tactics takes over the strategy and the vision and innovators dilemma to kicks in.
Eruj (01:02:38.403)
Absolutely, same thing. it's just in the end, I wish that we had found a way to keep, like the roadmap, it just ended. It ended, like whatever's flown got to version eight and it just ended, right? At that point in time, the weight of the revenue generating arm completely took over the product.
And then they're just down to M &A activity. At this point, the only way to make progress is through M &A, which is a very, the sort of inorganic method there is very complicated. As you know, you've been through acquisitions, right? It's very complicated. And typically then you're back to the brand problem and everything else. It's just a really impossible problem. But if you need to grow, you're doing M &A. So this is a cycle of death where we didn't need to do that. I think we just, it was disappointing.
Right, it could have been, it just ended in chapter seven, or it ended at mile seven of a marathon. Right, it just ended. It hit a wall when it could have been a lot more. And I, you know, I take some blame, but there's probably nothing I could have done, right? At this point in time, you have a public company. It's like, what am going to do? I can lay my body on the tracks and those run right over me. Entire dev team could lay their own body on the tracks and they're to run right over you. And so you just kind of throw your arms in the air saying, okay, well, whatever. I'm on to the next thing.
Nitay (01:03:53.82)
At some point the machine takes over and it starts powering itself essentially.
Eruj (01:04:00.199)
That happens with most, I mean, there's very few companies that find their way through that. I mean, we could spend an episode on decomposing which companies seem to have done really well. And some have, right? And often they just require strong leadership, right? I will say it is without Mark Zuckerberg, what would happen, right? It just wouldn't. No one would fucking invest in all that stuff. In Satya, same thing, right? It just...
Nitay (01:04:19.983)
Yeah, what Satya has done to Microsoft has been incredible.
Eruj (01:04:24.653)
It's stunning, stunning, right? And then without that, if you leave it up to the field and the quarterly numbers, just will never, it'll never solve for the right thing. So maybe when they're big, maybe it becomes easy when they're big enough again. It's like there's this middle point. Cause when you're really big, maybe you just, like Google's got so much money, they can just do it. But I think when you're in that, you know, I am, yes, I'm a public company trying to become profitable, trying to show growth.
I need to show a certain amount of growth. I don't have so much cash in the bank. can just do whatever the fuck I want. So maybe looking at Zuckerberg's the wrong case. Maybe you need to find more that are public companies that are valued in the low billions and say, OK, which ones are seemingly capable of maintaining their long-term vision while still growing? I'm sure there's a ton of them, by the way. It would be really interesting analysis. And then sort of decomposing, well, what is that thing that's going on there in terms of leadership and technology?
no, it's the whole podcast there.
Nitay (01:05:26.393)
Yeah, all right, so I think we'll have to leave it at that, but I have a whole list of podcasts that we could do in future. I think we have at the very least a pricing one, we have the physics of Vestimur, we have some public market analysis, so plenty more to talk about. So with that, thank you very much, Eric. It's been an absolute pleasure. We look forward to having you on again and continuing the conversations.
Eruj (01:05:36.397)
yeah.
Eruj (01:05:45.659)
Yep.
Eruj (01:05:49.595)
Cool. Guys, take care. Thanks for the opportunity.
Kostas (01:05:52.878)
Thank you so much, that was amazing.
Eruj (01:05:55.471)
Great, talk to you guys later.
Listen to Tech on the Rocks using one of many popular podcasting apps or directories.