Transcript: Yair Miranda | Cloudinary | How Do You Bring 3D Capabilities to E-commerce?

Intro: Cloudinary, from bootstrap to unicorn

Eyal David: [00:00] Hi friends, it's Eyal. You've reached Product Builder, where you can listen to the most interesting product conversations. My journey began in 2012, when I learned what product is at my own startup. Since then I've worked in the industry, and since 2017 my company has been advising a wide range of companies. My drive is to help all of us get better as product people. And before we start, if you enjoy listening to Product Builder, please rate us five stars on Spotify. It gives me great feedback that the podcast brings you value, and it would make me very happy, so yalla, let's start. Hi Yair, how's it going? What's up? Great. Joining us today, on the first candle of Hanukkah, is Yair Miranda, GM Dimensions at Cloudinary? That's how it looks. I said it right. So okay, tell us a bit about Cloudinary, let's get to know you a bit, and then I want to hear a bit about what Dimensions is.

Yair Miranda: [00:54] So Cloudinary is an amazing company that was founded completely bootstrapped, more than 10 years ago, in what's called "from bootstrap to unicorn, ask me how." The company was founded to solve a problem that website developers had back then, around everything related to media, media delivery and media optimization. They discovered the problem hands-on; they discovered it as a team doing projects and felt it firsthand,

Eyal David: [01:24] And then they founded a company around it, the rest is history. I really remember it, image optimization was a thing, and suddenly it was a game changer, they really owned that category, everyone has used them to this day, right? Totally. And that's actually the core business. That's where it started. I remember it firsthand even; I didn't know Cloudinary as a founder, when I was building SDKs for apps and had to crop images to the different sizes and make sure they uploaded and all those things. And then when I got to Cloudinary, I said, oh, really?

What Dimensions is: Cloudinary's 3D

Yair Miranda: [01:59] It existed and I didn't, it existed and I didn't know. A simple API, here's a link, amazing. Yes, a piece of cake. And today the company has more than 2 million registered developers. Nice. Tens of thousands arrive every month, something like that, really a very well-oiled machine, over 100 million dollars ARR, a unicorn, all that... all the... all of it. Um... And I joined about two years ago to lead a new activity called "Dimensions." So what is "Dimensions"? So what is "Dimensions"? Yes, great question. "Dimensions" is actually a new offering; in the simplest words, it's Cloudinary's 3D. Okay? So Cloudinary does everything it does for images, does it for movies, for videos, and it does all of this for developers, and it's also advancing and providing services to marketers, or to places that are, say, people who manage content, people who create content. And I actually joined to add another layer. Another layer, a 3D layer, both at the level of being able to experience 3D experiences, augmented reality and viewing products in 3D, but also to disrupt the way commerce organizations create content for their websites, or brands in general do it. Today they either have to take some sample, send it to a studio, go out and shoot, do touch-ups, and then say, wait, actually no, it's not good, we need to go back, and other things that cost a lot of money and time. And sometimes it even means you have a product and you can't launch it because you don't have images, you don't have creatives. Or you have a designer sit on all kinds of Adobe software and create this content manually. And we're coming to disrupt that, because if you actually have a 3D model of the shoe, then now this shoe can be in a sports hall, in the gym, and be photographed whenever you want, however you want. Those are the things we're actually working on, and we're at the stage in the process of getting to the place where the vision becomes a product, and we're also coming out of beta soon, so there'll be an announcement here at the end of the show.

Eyal David: [04:19] Wow, wow, so wait, let's take a step back for a second, okay. Just so you know, my wife is a photographer, product photography among other things, so I also know the field well. So basically what you're saying is: we have a company that's existed for over 10 years, is insanely successful in its field, knows how to market to a certain audience, knows how to create a lot of value and a lot of money from it, and decided to enter a completely different field that targets a different market. Because basically what, why did they actually go there when they even thought about it?

Emerging Business: an incubator inside the company

Yair Miranda: [04:47] So here I'll take a step back and say that a company, and I've seen this in all kinds of companies, a company eventually reaches a place where things are good, it's growing, it's doing something excellent in a certain field, and it wants to grow. And then you can grow in all kinds of ways, and Cloudinary came to the conclusion that in order to grow you also need to invest. Again, a bootstrapped company, so it's not like raising money from somewhere and throwing it, no offense to anyone, at some risky investment; rather, you want to try ideas and very quickly do a turnaround: it works, great, we continue; it doesn't work, let's sunset it and move on. So we founded, I say we founded because I'm part of it, I'm a partner, but it was founded before me, an organization called Emerging Business, which is a kind of incubator, you could say. There's an idea, you start, you take two entrepreneurs, let them run, and try as fast as possible to get signs of oil: there is, we continue; there isn't, we shut it down and move to the next idea. That's what happens, and into that place I'm actually doing the 3D, Dimensions, as a pod that's hopefully a growth engine for the company in the future.

Eyal David: [06:02] Stunning. So it's a bit similar to Wix and how they spun out, for example, monday, in a similar way; there they really also used the tool, I remember, they knew how to provide value there. How does Cloudinary's core business actually provide value or information or data to all the emerging businesses growing there?

Yair Miranda: [06:18] So that's a good question, because there's actually a trade-off here, I think, in general for product: if you want to run fast, sometimes you'll break things, and you have very, very powerful machines sitting there, not machines exactly, but a skilled machine that knows how to do everything Cloudinary does. But sometimes, in order to use it, you'll have to slow down. And then sometimes we just very consciously say, let's, just like if I were now a startup founder I'd say okay, after the Round B we'll reopen this, so the same thing. Let's do it the fastest way possible, and when we reach the right stage we know how to scale. We'll open it up and rearrange it. So we actually use Cloudinary when we do the go-to-market too. So I have my own agency, but there's a sales organization that knows how to sell; if I want to approach enterprise, it's the same thing. If I want to access all the traffic that comes in, in the end there's also very large infrastructure that is ready for some additional system, some additional product to run on top. And that's why it's easier to be a founder inside Cloudinary, inside an organization like this, than to be in some incubator where, I don't know, you got some credits on some cloud but you're alone.

Eyal David: [07:41] Of course. So okay, basically, let me take this for a moment. So okay. We decided to set out, Cloudinary decided to set out. They brought you in. It's also important to say, the advantage here, I don't know about advantage, but basically you take an image, you'll tell us about that, and you know how to model it as you described, and you called these, basically, creating, like in 3D printing, right? That's what you described: I have a model of the thing, and now I can start projecting it and producing images, media, videos too from it.

Yair Miranda: [08:07] So yes, exactly. You have a model, a 3D model in a 3D model format, it doesn't matter, GLB, glTF, or, we won't get into all the buzzwords, but just like there's MP4 and there's JPEG or PNG or whatever, the same thing: there's a model, and this model can be displayed in space, and it can also be placed in a scene, and it can be photographed, and with a cell phone you can see it next to you on the floor or on the table. There are many things that exist and many things that we're building.

Eyal David: [08:42] I'm really curious to know, and we don't have to touch on this, but you're actually also targeting the traditional advertising market, blah blah blah, as it is today, and you're also actually offering new media that can be used, like you said, I can suddenly point with my phone and actually see the product next to me, I can put it on my foot, just saying.

Yair Miranda: [09:00] So there's an infinity of experiences, and I think part of the challenge, we're here on a product podcast, is ultimately to choose. Because there's the Apple Vision Pro, and there are augmented reality and virtual reality glasses, and so on and so forth that the big companies talk about and launch. And as a founder, I work at Cloudinary, but a founder has to choose where I'm betting there'll be a significant business, and not something that will only happen in five years. So honestly, say, Apple Vision Pro is insanely cool; I didn't invest more than an hour looking at the first presentation, because it's so far from me. I'm not a research organization, and I don't want to build some pretty demo to show off; I want to sell, so we passed on it. Where the big money is, that's actually the question, and once I find it, I have ideas, so that's where I'll put the resources of this startup.

Eyal David: [10:08] Which is also different, as I said, from a regular startup, because your goal really is first to crack it as fast as possible, to show there's oil, and then you race toward the broad vision, as you say. Versus a traditional startup that would now have to think fifteen years ahead, present that story, and also operate on the track of converting customers, generating customers, blah blah, but also keep telling the story to its investors all the time.

Yair Miranda: [10:31] So you have more focus and more quiet in that respect. So here, actually, I'll tell you a bit about how we work. We try as much as possible to stick to the startup story, inside a company that's also a startup, yes, but a startup inside the organization. So for that matter, I'm a founder, I'm independent, I have a board, I have investment, I need to hit milestones in order to raise the next round. I sometimes ask myself what I need to do to raise the next round. Meaning, if I only present a vision, that's not enough. I think in real life too, if I only present a vision and there's no progress, then after a year and a half the money runs out; it's pretty similar. So here too, nobody would have started this process and this journey of each of the pods if they didn't see a vision of inventing a new Cloudinary and building something at least as big as the existing business. But, what they call show me the money, I need to show that I'm progressing and selling.

Eyal David: [11:37] I just think you have less convincing to do, because it's not necessarily that you built the vision/the milestones, still, you optimize them or improve them, but unlike a startup you have less convincing to do with those investors, that's what I'm saying. I hope so, I know them much better, maybe. So maybe tell us for a second what your day-to-day looks like. Okay,

Yair Miranda: [11:57] I think it's not fundamentally different from a founder's day-to-day. At the stage we're at, we're after raising sums that aren't trivial, but we're not yet at a Round A, okay. Again, we're in '24, I'm not sure anymore what a Round A is and what a seed is these days, but yes. And in the end I'm running in every direction like a startup person. So day to day, it could be that right now I'm dealing with sales and really with conversion, and building a funnel and getting on calls with salespeople. Before that we were more focused on self-serve and PLG, so I was much more there, in fine-tuning that. I have a product hat, so I also work a lot with the team on what to develop and how to develop it, with UX and so on. I also work with the organization, which is a bit of working with the board and the investors and creating mindshare about where we're going, and making sure the next board meeting is a successful meeting and that there's a raise. So overall it's not fundamentally different, it's just from nicer offices, and there's a Hanukkah party, and we were also at the company's gathering in September, which is a truly amazing event where the whole company travels and meets once a year for bonding and team-building, but also to work on ideas and create the company's improvement for the future. All those things we didn't have as startup people.

Eyal David: [13:41] But I think my day-to-day isn't fundamentally different from if I were a CEO outside. So take us to your challenges: what do your challenges look like as Yair at Cloudinary? What is this initiative actually called?

Toward product-market fit, and MVP mode in a team used to scale

Yair Miranda: [13:54] Dimensions, which is my pod, it's the product, yes, and Emerging Business is the umbrella for what we call invest; it's where we invest in order to grow. So what are Yair's challenges at Dimensions? We're at a stage where, I hope, we're getting very close, a step away, from what's called product-market fit, but we're not there yet. So right now I have the conviction and I have the data, and I need to prove that we're really doing the right thing. So our challenge is actually to crack product-market fit, to do the minimal things needed to prove it, before we already think about scale, which can come easily in an organization like Cloudinary. Meaning, we serve some of the biggest companies in the world, not Dimensions yet. And then the thinking is, how do you do it in order to prove it, but the day after, one of those giant companies will come, and you'll be able to serve them very quickly too. So there's this tension all the time, which obviously exists more in a company like this, because everyone around is working at scale, and it's a kind of muscle. Let me say this, maybe let's open a parenthesis for a moment. The team that works with me is a team made up mostly of Cloudinary people, some more veteran, some less, who know how to work at scale. I had to get them used to saying, guys, it's okay, it'll break, but that's fine. It took time until we got used to thinking in MVP mode, and not everyone around is used to that mode; that's also a process that's a challenge in itself. Let's stay in that parenthesis for a second: so how did you instill this zero-to-one, this approach, a lot of slip-ups where you, what does it look like? Look, in the end I think it's less processes and more managing by example. Meaning, in the end I'm the GM, I sit with the team, my development team reports to me, so it's not that I tell them what to do at that level, but the more I demonstrate tolerance for failure and tolerance for... it's okay, let's see if customers eat it and then we'll improve, let's take the... let's cut this corner here, it's fine. We understand what we're doing, we understand the risk, we manage the risk. So that's ingrained, and today I already... I think the team knows it on its own. Meaning, when we're talking about a new idea right now, I think nobody will come to me with how to do it for a billion or tens of billions, but rather how to start and we'll progress from there, and how to do hacks, and how to use all kinds of things, Intercom, instead of developing. Even though at first it was, I'll say for me too, very hard to make decisions like that, because it looks less good, but okay, never mind. No offense to anyone at Intercom, just as an example. So it's something that does come from experience. Simply risk management and personal example, at the end of the day.

Eyal David: [17:13] Exactly, exactly. Okay, we've closed the parenthesis, we're back to your challenges. I'll take you back to the beginning of this journey: so when you arrive,

SMB vs. enterprise: where to start

Yair Miranda: [17:24] I think anywhere, I think it's true not only for a founder or a VP Product, it applies anywhere, whether you arrive as product or in any role: first learn which tools you can use, what works in your favor and what works less well or isn't optimal for the start. So Cloudinary is a company that, it's bootstrapped, yes, it was founded in this place of customers, people who sign up with Gmail and have no business behind them, okay, they come on every day, enter the app every day, really crazy amounts. And on the other side it serves the Ubers and the Teslas and the Nikes of the world, and you have to choose where to start. So the choice was, let's really start with the small ones, reduce risk, and there you can also, you know, what can you do, sometimes practice on them a bit, be faster and less stable and other things of that kind. And as time went by we came to the conclusion, it was very clear to me but I also needed to bring data to convince, because again, I have a board, so that's part of it, that's fine, that we basically wouldn't find the oil there. Also because 3D is a bit more sophisticated: they don't have it, they need things from you, I don't know, connections to Shopify, that you don't want to provide. We're developer-ready first and foremost; they need to have a development team, these and all kinds of other questions of that kind. And it's also a different kind of story to do both. And enterprise we have; Cloudinary has tens of thousands of paying customers, thousands of enterprises, it's truly an amazing story. And if I can be in touch with whoever will later become my customer, then why not. So that's actually a process we went through and I went through. Over time, basically starting from, let's sell to these mom-and-pops, and gradually moving to selling to larger companies with bigger requirements, but also with security, and also with really more advanced capabilities, more ready for things that will disrupt them, plus all kinds of other things I learned about the SMB market: they like to pay less, a free tier, 25 dollars, that's, that's something, and it's harder to build a business that way. Very logical. So now you've completely dropped those SMBs, you're focusing on enterprise, and you've actually finished all this shifting, you're launching this product, it's actually going to change. The product is focused on that, even though anyone can still come, sign up and start using it, but don't tell anyone, it's not my priority, let's say, to funnel the funnel right now, so I deal with it less.

Eyal David: [20:20] I'll deal with it here before we want to move to the next stage, probably, because that's the Cloudinary way. Yes, we have customers who started as developers, tried it and then brought in the enterprise, so there's a lot of review in that. The well-oiled machine is actually even better, it's worth investing in, no? Yes, yes, totally. Stunning, okay, that was a great overview. So let's talk a bit about Yair. Tell us for a moment a bit about your background. So right now you're at Cloudinary; take us a few roles

Yair's background: from development to product

Yair Miranda: [20:48] Back, and let's understand a bit about you for a second. So you actually started as a developer, right? I started as a developer, absolutely, quite a few years ago, but I think throughout my career I'm technical, and that's part of who I am, I don't say it negatively. But I think in my whole career I was something like five years really in development, developer, development manager. I started at Radvision. A company where, when I joined, we were fewer than 30 people, and we went through an IPO on NASDAQ, and the company grew, and one day it was also sold to Avaya. So, you know, I grew together with the company, and there were a lot of opportunities, and I took one and made the move, and from developer I made the move toward product until I was fully product, and since then I haven't looked back. And is that something you actually aimed for, product? How did it roll into that? I don't think that when I studied math and computer science and when I started as an engineer I thought about it; honestly, there was no product, come on. There was no product; one day product arrived, and I remember that as developers we took her to lunch and tried to interview her about what this is, what it means; I didn't understand back then. I think today I understand, but back then I didn't, it was very hard, and also it really wasn't mainstream to have product; development did what it thought. But as time went by, I really felt I needed to get out of the lab and see the world and understand why I'm doing what I'm doing. And there was an opportunity, so I took it. What was the opportunity? First of all, there was unpaid leave. Always good. The woman who's now my wife and I went traveling, we went to the East, to China and all kinds of cool places. And then when I came back, I picked up the phone on some Tuesday, right, Wednesday was the Passover Seder, and my former boss, because I'd gone on unpaid leave, I said okay, we'll talk when we're back, told me, listen, on Sunday you're getting on a flight to Japan. Okay, so on Saturday we went to buy a suit, I took a laptop bag without a laptop just so I wouldn't look strange to the Japanese, and I started managing a project that actually took the company into the field of mobile and service providers. So I managed it both with the customer and I managed the development too. And then when I delivered the project, we sat down and said, wait, what is this thing, maybe there's an option here to build a new business, and the answer was yes. And then naturally I took the whole product area of that business unit and let go of development. And that was the first time, really, and from there.

Eyal David: [23:35] Oh, that's lovely, our stories are very similar. I was also a developer, and at some point I took unpaid leave and flew to China with my wife. I also suggested experiments to you, but I came back and started a startup, so very similar, really. So okay, take us from there: Japan, also amazing, you came back here, really, I'm looking at LinkedIn, you've been through quite a lot, and you actually got to Cloudinary in the founder role, stunning. So let's go back to Cloudinary for a second. You know what, before that I'll ask you a question: I see you've been through many, many companies. A rich career. I want to know, in your eyes, what are the most important product skills in a product manager. Male or female.

Product skills: the CEO of the product

Yair Miranda: [24:13] The million-dollar question. I think it's actually good, because you look at it with several eyes, I like that, both as, so to speak, a founder, a founder who needs the zero-to-one, and on the other hand with a broader view. So I think what I need from product, whether it's me back then or from my product people: that they see the big picture and are curious, and don't become technicians who pass requirements to development. That's the least interesting thing in my eyes, like managing development or trying to keep development doing or not doing something. You're there to move that rock, and nobody else will move the rock. Development will run even if you write a few fewer lines in the spec. Nobody else will interview the customers, nobody else will think about the next stage, nobody else will be the person there when they're needed. And if you fill this role that people call the CEO of the product, I don't know, there are all kinds of clichés that are sometimes a bit triggering for people, and I strongly agree with it: think of yourself as the CEO of the product and go for it. And that's the most important thing for me as a manager, I said, I don't always have time, often I don't have time, but I have no choice but to deal with the details. But I always want to know I have someone to talk to about strategy, and that there's someone else I can send to talk with a customer and it isn't only me. And also in a larger organization, where ownership sometimes gets a bit lost: who's the owner of this product? He sits here, he sits in sales, he sits in development, there's nobody else, it's product. So product needs to understand that and be on it and embrace this charter, this mission of theirs. I always look at it this way. By the way, I think it's very clichéd, everyone talks about it, I even read something about it before I came, someone wrote about it recently, I really think so.

Eyal David: [26:34] So actually, when we talked about it before, you also presented it as the thing that drives you, and today that's also what you really do: you're the CEO of the thing. So let's really go back for a second, I'll keep calling it Dimensions, you'll keep correcting me, which is funny. So as the CEO of Dimensions and the person with the most accountability who pushes it the most, take us for a minute to that moment where, okay, you came in. Describe for us for a second what it's like to come into the thing, not day-to-day, but coming into this role, and what the last two years actually looked like. Take us a bit through your journey, it's really intriguing.

Yair Miranda: [27:09] So one of the things I set for myself was to talk to as many customers as possible, potential ones or just people from the industry. So I started from my network. I simply went on LinkedIn and contacted people and said, you work in e-commerce, let's talk. That was still when you were actually targeting both SMBs and big companies in parallel. Yes, but even today I want to talk to everyone who's at a certain point today and was at an earlier point in another role, and everyone gives perspective to understand how this ecosystem works, what people's KPIs look like, what their titles are, who reports to whom, who does what. Those are very important things. And Cloudinary is a world champion at knowing developers; we're all developers, some more, some less, we know them excellently, but I need to know the others: the buyers, and the users, and the operators, and the budget owners.

Eyal David: [28:17] So from day one you actually have a list of, I assume already at the level of hundreds, probably in the range of hundreds, not the upper range, but

Yair Miranda: [28:29] That you talked to, some prospects, some people who today just, like, let's take an example: me and the UX person on my team, we simply interview people for pay. There's a platform called PlaybookUX that's usually used for user testing, and if you pay a bit more, you can talk to a persona with the profile you want without selling to them. Just come and ask them, okay, who, what, what do you do, how does it work, describe to me a product launch, the choice of the platform you work with. And people, very surprisingly, are very open and talk and are happy to do it. And sometimes a lead even comes out of it. Amazing. Not that expensive, by the way, to reach enterprise this way. I need to do you again. But, so that's the first thing: knowing that around me there are technical people who'd know how to develop anything I ask of them. There are salespeople and there's a skilled sales organization, so we don't want to bother it. Let it make its ARR; we're selling on our own right now. But the most important thing was actually to come and condense all the basic assumptions of the product's initiation deck into, okay, what of this works, what of this doesn't work

Eyal David: [29:50] And let's already create the first experiments in order to go out. So that's where I started. I'll ask you, then, within this exploration journey we didn't talk about competitors at all. I don't know what this industry looks like, so can you describe a bit for us, because at this stage you surely got to know a few names, who the players are.

Yair Miranda: [30:06] So there are quite a few players in this area of taking a model, displaying it and optimizing it; it's quite a red ocean. Lots and lots of small players. And the reason it's like that is that there weren't models. So every company that started a while ago and raised money on the basis of the metaverse, of here, this is going to explode, suddenly found that, great, but nobody has a model of their product. And then many of these companies simply became companies that produce models, and then they're service companies, and that's less interesting to us, to me, less scalable for us as a SaaS company. So there are many companies, and we chose not to take that angle. I know how to help them if they need help with customers, but it's absolutely something we don't do. So that's why we have a lot of competition in this field, and for better or worse they do more than us and have been on the road much longer than us. But I think we have an advantage in that we understand developers very well. Because in the end there's an initiative, there's a vision, there's a need in an e-commerce organization, say, that I sell to, but in the end the developer, someone gave him a task to go and solve the problem: go create this experience for me on the website. And then he knows us and knows to come to us and solve the problem, and we speak to him as developer to developers. So that's, let's say, some initial differentiation we have. And beyond that, there's no way around it, in the end if we go to enterprise, enterprises want to know that, like they once bought IBM, they're coming with Cloudinary and scale is guaranteed; maybe they're already our customers, so it's not an issue at all.

Eyal David: [31:57] And probably also by choosing enterprise from the start you made a clear cut on many things you don't need to deal with, so it simplified the competition for you and maybe even positioned you better.

Yair Miranda: [32:08] Cloudinary has a DAM, a digital asset manager, one of the best. If I were a startup doing what I do, I'd provide something minimal like that, and there aren't many companies that have this for 3D and for images and for video and for all the metadata and everything a product like this needs to do. So if you're asking where I started, one of the first things I did was to say, okay, what are my advantages as a startup inside Cloudinary, what are the liabilities, those may be places that stop me; there are, that also happens. But once I identified that, okay, these things can serve me, we doubled down on those capabilities.

Eyal David: [32:48] Excellent. So you went on some exploration journey, came back with recommendations about what the direction is going to be, right, if I understand correctly, and out of that you also did some scoping already at this stage, you started defining what the first version of the thing would be, when it would come out.

Yair Miranda: [33:03] So it's less orderly than that, because specifically, when I arrived there were already a number of developers who had started working before me, and there was the initiation deck, okay, and there were recordings from the product folks who simply did excellent discovery calls. Yes, I sat and listened to all the calls, all in Gong, so it was all easy. And then basically we needed to, it wasn't some kind of, here, let's sit for two months on a deck and start, but rather, okay, let's take what was developed and think as fast as possible about what we, how we can get to an alpha, how we can get to something we show a customer, potential or real, and that they'll give us feedback on it. Like, so we'd know how to validate. None of us is a visual person, meaning we look at things as engineers: wow, looks good, look at this shoe, it looks great. Two years later I can already say this shoe looks a bit computer-graphics, it wouldn't pass on a website, and certainly not in something that's a hero image on some site. So I recruited, as one of the contractors, my consultant: he's a studio, a studio that works with me, and I brought him in pretty early, because he knows what looks good and what doesn't, how you take that product to a studio today, you set up lighting, you position it, it's a whole discipline in itself. None of us knows it.

Eyal David: [34:34] So you also actually started working with the team, and in parallel you set up another envelope of consulting to bring in this finesse, to bring this specialty inside. And I'm waiting for the moment you say there was a difference between reality and the initiation deck. Or there wasn't, I don't know. It's not clear there was, I think there isn't, there isn't. And did that come up after this exploration process you did, or did you actually start throwing things at the wall and getting people to see them? So we have principles at Emerging Business, one of which

Data eats conviction: alpha and private beta

Yair Miranda: [35:07] Says that data eats conviction for breakfast. Which I really love. It means, okay, you can argue in the room a lot, but let's see what actually catches on. So as fast as possible we wanted to go out to what we call alpha, and then to private beta, and from each of these stages things changed fundamentally. What does that actually mean? Okay, take me to the alpha, what happened in it, and after that we'd love to hear what private beta actually is.

Eyal David: [35:35] I think at this stage the metric isn't sales, right? What's the metric, like, what measures success at a stage like this?

Yair Miranda: [35:43] Okay, so the question is great, because it really reminds me what happened in the alpha. In the alpha I wanted to know that what we're producing, I wanted to know that the potential customers I'm talking to have models, 3D models, yes, at a level and quality I can work with, and that what I'm producing interests them. So I actually didn't want it to be self-serve. I wanted to get on a call with them where I do the onboarding for them. I wanted all those things, I don't deal with them, I just want to get to a state where the most basic thing is: give me a 3D model, get a code snippet or some image from me, and you give me a thumbs up. So that's what the alpha was. If I'm honest about the alpha, they gave me a thumbs up.

Eyal David: [36:30] But today I know it wasn't good enough. Like, whoa, whoa, yes, you said it wasn't enough, but thanks, they were nice, that's part of it, and by the way, right. So what would you do differently today, in your opinion, to get a more decisive answer?

Yair Miranda: [36:46] I think I'd take that same expert, that studio person, that creative person, and join forces with him earlier. Because it's actually a bit like, I don't know, selling to a persona something you don't really understand from day one. So, like product managers at defense companies are, I don't know, former military people, and astronauts sell to astronauts, and so on, so I needed... but we thought we were selling only to developers, and therefore developer talking to developer is fine. But I think in hindsight I needed that creative person much earlier. And then he arrived, and then we really started to understand more. And did they really have models, when you started to understand more, like, end customers? They have models. Models are something that's much easier to get today than two years ago, and certainly than, I don't know, five years ago. For companies founded five years ago it's still something that costs a bit, not a lot, it costs money: there's from 15 dollars or 20 dollars on Fiverr, or more, up to... When designers do it, it costs them in the hundreds and even more, with some

Eyal David: [38:04] Very sophisticated scanner, everyone chooses their level, but it exists. And are they doing it today, like, actually the use case, because they're going to produce, their thinking is to really produce, let's say, the images like you offer, like they're preparing for a magazine, or do they actually have something else they do with

Yair Miranda: [38:22] These models? So that's also a good question, because here too, the world is complex. So if you sell large products, you find that okay, as a furniture store you can't photograph all the sofas with all the fabrics, and you need to work with something digital of this kind. And if you sell other products, or if you're a marketplace and you're just importing from China or selling all kinds of brands and nobody shares their models with you, then there are all kinds of approaches to that. There's also, in the end, if I'm the brand and I manufacture a product, I have the CAD model it's manufactured with, so now you can do something with it, maybe not perfect for what I need, but. And there's also the matter, in large organizations, of taking it from whoever has the model and bringing it to marketing, right? That's also something to research, whether it's easy or not at all. There are endless questions and challenges here, which sometimes work in our favor, because it's easy for us to come in through someone who's in pain, even though maybe he needs to work afterward to bring approval or bring the model. Sometimes it's impossible: we have a customer, I won't say the name, of Cloudinary, who's really the top of the top, and he's not willing at all for the model to reach the client, meaning everything has to be through some kind of streaming, so that, God forbid, nobody steals the model and produces something similar to his. So over time you learn to work around it. That's good. So I'm bringing us back: what actually is private beta? It means you took a few users who were very good, or you brought them through Cloudinary's engine and opened a beta for them that suits them better? So that's nice. So in the alpha it's really freestyle, I just want to know that you saw something and gave me a thumbs up. In the private beta it's really a selected number of customers that I do the onboarding for; I don't expect them to understand on their own. I'll help them, and now we'll see what the issues are. And who was the persona on the other side who actually went through it? On the other side, in smaller organizations, it could be someone who's really a founder, a product manager, a marketing person, someone like that, depending on the use case, or a developer. In larger customers it's already more distinct. There's someone who has a role, has capabilities, knows what he needs to do. And is he today your buyer? That role is more complex than that, by the way, we're also still on the journey, yes. But today we're identifying that in organizations of this kind there are all kinds of buyers. Some simply get an assignment, go improve the... I see Gen AI with some CEO or someone with authority who says, what's going on, how are we improving? And then sometimes it'll simply reach the IT area and reach the innovation area, if there is one, because they're the ones who know how to respond or run some experiment. And at this stage it's convenient for me to work with those, but in the long term it's people for whom producing content costs money, which is more the creative people and marketing people. That's classic. But it depends on the use case; we, again, do several things. And some of them are pure developers who ultimately need to implement something on the website, and that's what they do, and they're the users and they're the buyers. Maybe they don't operate it. Again, there are all kinds of nuances here that we didn't get into, just because of the limits of the format, but it's absolutely something super important to know in order to build a product.

Eyal David: [42:07] And then this private beta happens after how long in the role? About half a year. Oh wow, that was quick. I thought it was... it's clear to me from the... Out of two years, that's nice. So after half a year you're already getting on calls with people, you take them by the hand, and you probably also manage to bring them to the moment where they see value.

Yair Miranda: [42:27] So first of all, getting on calls right away, yes. I told anyone who'd listen: listen, I have no problem getting on a call as a layman, as someone who doesn't understand the field. I also had experience, I came from all kinds of fields in my career, and many times it's fine to come and say, listen, I apologize in advance, I'll ask you stupid questions because I'm not from the field. And people are happy to explain. I mean at the level of interviewing ship captains when you're a landlubber, or interviewing doctors when you want to initiate something in physiotherapy. People love it. So within two weeks I was already getting on calls and saying, listen, my name is Yair, I'm from here and there, I want to talk to you about something at your company

Eyal David: [43:19] And forgive me in advance. Excellent, let me be precise, I'll bring us back there. So after this private beta, you were already getting on onboarding calls with potential customers, you already took them through a journey, they saw value at the end of this thing. You finished this stage, I don't know after how many months, I don't know how long it took, and then what insights did you come out of the private beta with? So we discovered that they don't understand

Customers didn't get it: rebuilding the user experience

Yair Miranda: [43:44] And that the problem is on our side. And then we paused the... we closed the valve on new customers for a moment. That was also more or less in the period we're in now, when customers are in Christmas and New Year's and we have a certain grace period. We simply rebuilt the funnel and the experience so they'd understand. So basically, okay, you improved the funnel, did you also touch the product itself, how was it on that side? No, the funnel, I mean, by funnel I don't mean the demand generation funnel but once you've landed in the product.

Eyal David: [44:18] What do you understand, what do you do? Because now, from all the insights you gathered from the onboarding you did on the calls, you took them and actually created them?

Yair Miranda: [44:25] Yes, and then we actually sat down, UX, development, the whole team. At some point I also brought in a growth expert to help me build a kind of PLG funnel inside the product and outside. And we started improving, meaning running several experiments and changes so that a user who arrives, who has intent and has what's needed, say a 3D model, will in the end come away with some aha moment or some wow effect. And after that we also discovered it's not enough; they also need to be able to share it with their colleagues. So there's a process we went through. What didn't they understand there? What didn't they understand, as you claimed? Several things. The product started, if I take you back earlier in the conversation, when you think at developer scale, you say, I have an API and I'll create some documentation here, and the developer will read it and understand, and here it already integrates with his PLM and the SKU taxonomy and all that. And then we discovered that when you take someone, even if he's a developer, and he reaches your user interface, and in the end all you gave him is something similar to an API but in a UI, then he doesn't understand you. And then we said, okay, let's strip it down, let's simplify it, let's make it really simple so he'll succeed, let's create an experience of success. So we made it much more graphic, much more alive, much more guided, some kind of wizard, and giving him a "here, if you do this, this is what you'll see." We put in all kinds of samples he'd succeed with, we built them, we did everything so he'd manage to get to the place where he suddenly sees a product page with

Eyal David: [46:13] His product. And now, at this moment, you got past this stage, he got the aha moment, you still didn't charge money, I understand, the metric was actually to see how many reach success.

Yair Miranda: [46:23] The metrics we set, good question, I haven't talked about metrics until now. At this stage it was how many users are serving from a website that's live. Serving is actually pulling an image? What does that mean? Serving is serving an image, serving a model, doing something from a website that's not localhost and not some... I want to see traffic and count how many of those there are, in order to understand that someone, without me talking to them and without doing too much, you know, white-glove for them, went live. That's your closing of the loop, basically: all the approvals happened, they're happy, they put it there. Yes, and I'll say that sometimes it's very surprising who goes live on your site, because we've really been investing in enterprise for a while now, but we left the self-serve PLG machine running. And we have hundreds of models of steaks and cuts of meat of different kinds on the site, for some website that simply demonstrates what different cuts of meat look like, and we keep coming in the morning and seeing another few dozen meat models that went up, and it's amazing, because it's probably a vertical I wouldn't have targeted. Crazy, crazy. I apologize for the trigger to anyone who's vegan. Actually it also makes a lot of sense, because then you really save on food. It's a thing with a very short lifespan, so it makes a lot of sense that you'd create some

Eyal David: [47:59] Model, because otherwise he needs to go photograph it, because it spoils very fast, and photograph it again. My wife photographs food, so I know that. So okay, we got past this stage.

The shift to enterprise and looking ahead

Yair Miranda: [48:08] What was the next stage, actually? You rebuilt everything, as you said. So at this stage we more or less came to the conclusion that it's worth going to bigger customers. Because it actually works, but we saw that whoever really converts is bigger. And maybe it's not worth staying in this area of SMBs and small customers and all kinds of others who, in the end, have no real business behind them. And then we shifted more in that direction, which actually means stopping all the demand generation for a moment, and the campaigns on Google and LinkedIn and wherever there were campaigns, and simply building a CRM page and starting to convince and get into the sales and renewal processes and so on of existing Cloudinary customers. Because that's ultimately my unfair advantage versus a founder, or versus if I were outside now, beyond the fence, trying to sell: I wouldn't have access to this. And then we simply shifted, and that's where we actually are today. We didn't stop anything we built in the past, but our focus right now is all in on the big customers.

Eyal David: [49:33] Which makes sense, I think we both agree. So tell me, do you have some gap in the sale, because again, as we said, the buyer isn't necessarily the same buyer, right? So how do you bridge that, because I'm not Microsoft coming and selling a suite now to the same, to the same person.

Yair Miranda: [49:51] What does it look like? So that's a good question, because as a founder I want to know that in a while I'll have cracked it. But there are so many organizations, they're not cookie-cutter in their organizational structure, and there are so many cases where, when I arrive, the person who's easy for me to reach is, say, from the development side, IT and so on, and still he's significant enough in the organization, or he has enough connections, or he's an entrepreneur at heart, in all kinds of, you know, scenarios where I can still make a sale. And it's happened to me quite a few times that the moment the penny suddenly dropped for someone who's the CTO, that's who I was with, then, let's wait a moment, let's do another meeting, and suddenly he brings his CMO and the marketing people or whoever needs to approve the budget. So okay, maybe this way we won't manage to reach all the targets in '27, but '25 is amazing, I don't need more than that for sure. So what actually is the metric that drives you today? Just customers? Today it's mainly customers and usage, meaning I want to see that I'm not only

Eyal David: [51:05] Collecting customers, but that they're really using it. And are you still measuring that with the metric of, the metric of how many actually publish in the end?

Yair Miranda: [51:16] Yes and no, because today it's much easier to say I have a deal, there's money, there's ARR, so it's much more immediate, and the usage will come. But first I really want to, show me the money, that's fine. That's for sure. I think also another thing: they don't necessarily use it only on websites today, right? There are many more uses; if I'm an enterprise, I can't measure them if there's an ad here on Facebook or whatever.

Eyal David: [51:42] Totally cool. Now actually the desire will probably also be to know, benchmarking within an industry, what works better, right? That's actually the next thing, no?

Yair Miranda: [51:53] Yes, but we haven't gotten there. Meaning, you know, there's an infinity of verticals and an infinity of data points that I'd be very happy to have, but in the end I can get help, and I do get help, from analysts and data and so on that I have at Cloudinary or that I can buy outside. But I keep asking myself: if I really were now a founder in a garage with the same number of people and the same amount of money and so on, would I spend that money, would I now go

Eyal David: [52:27] And spend my time on these benchmarks? And many times the answer is no, so that's why I don't. That's for sure. I mean, but like, the next stage those customers will want, beyond getting the models, will be to get some insights, to get some further solution from you. Like, what are you actually hearing?

Yair Miranda: [52:42] So social proof is a very important thing. We, right, enterprise, we have enterprises who can tell what it did for them. So that's super important. And there's this moment also with the users or the buyers or the influencer in the organization: if he sees the output and says, wow, wait, you can get to this? And remind me how long it takes or how much it costs, when he knows how much it costs him today and how exhausting it is for him. So at least trying, everyone wants. And now it's on us to get them to also put it in at scale in their organizations, and that's also the reason we're coming out of beta, so we can do that.

Eyal David: [53:28] Stunning. Listen, lots of luck, Yair, it was really mind-blowing. I really enjoyed recording this episode. I want to ask you one more question. So where do you actually see this thing in 10 more years, where do you imagine yourself within this thing? We kept being careful about the 5 years, so I jumped to 10.

Yair Miranda: [53:46] Yes. So 10 is really a long time, and I think that in 10 years, like when GPT entered our lives, and today we don't know. Three years, right? Yes, nothing, really, right. So people worked on it long before, but nobody knew it, and today everyone knows, that's usually what people take as a reference. So I think that in 10 years there won't be a question at all of how you produce content, and this whole thing of this disruption will happen big time, and that's how organizations will work. So if you ask me, in 10 years I think it'll be mainstream, and Cloudinary will be on it. Excellent, and I also wish you another unpaid leave in China. Totally, we haven't scheduled it yet, but it's definitely already in the plan. Amen, Yair, how fun. So maybe come back in another year and tell us where you are, physically, or let's see where we are. We'll see. Okay, so thank you very much. My pleasure. If people want to contact you, questions and so on? LinkedIn, yes, LinkedIn. And we're also hiring in general, and probably by the time the episode airs we'll already be talking about more pods, or more pods opening, so this thing of being a startup inside an established company, I recommend it to everyone. So follow me on LinkedIn and it'll be published there. We'll put a link in the notes. Yalla, thank you very much, Yair, really. My pleasure.

Eyal David: [55:30] Hi friends, thanks for listening. If you found this podcast valuable, you can of course subscribe and follow us for more episodes on Spotify, Apple Podcasts or any other app. And of course if you couldn't find us on some app, I'd be happy if you write to us. We'd love five stars on every platform, and for you to follow us so more listeners can discover us and find the podcast. You can also find the previous episodes in any app or on the YouTube channel, we have links in the description. So until next time, yalla, be efficient, and bye bye.