Kate Barrett (00:06):

This is Founder Forward, the podcast from NEA where we explore the company-building journey with candid commentary from founders and investors, some legendary, some just getting started, all moving forward. I'm Kate Barrett.

(00:28):

On this episode of Founder Forward, I spoke to Sven Al Hamad, Founder and CEO of Webiny, an open-source, serverless framework enabling enterprises to build, manage, and deliver content at scale.

Sven Al Hamad (00:42):

I'm Sven Al Hamad. I'm the Co-Founder of Webiny. And Webiny is an open-source enterprise CMS that's built on top of the AWS serverless infrastructure.

Kate Barrett (00:54):

We talked about how Sven's experience building websites compelled him to create something better, Webiny's approach to gaining traction in a massive market, and how challenges at the macro level shape aspects of the company-building journey.

(01:08):

And on every episode of Founder Forward, we like to fast-forward to get some insights from a founder and investor who've walked this road together before. Today, we'll hear from Ali Ghodsi, Co-Founder and CEO of Databricks, the open-source data analytics platform built on Apache Spark, and Pete Sonsini, General Partner at NEA.

Pete Sonsini (01:29):

Hello, my name is Pete Sonsini. I'm a General Partner with NEA, and I run our Enterprise Software Group and joined by Ali Ghodsi, Co-Founder and CEO, of course, at Databricks, with whom I've worked for many years now. And we're here to talk about our journey together and where we're headed. Thank you for joining us.

Ali Ghodsi (01:47):

Thank you so much. Super excited to be here, Pete.

Kate Barrett (01:50):

Ali and Pete discussed what it was like to turn a successful open-source project into a really big company, the role customers can play in creating new products, and the importance of relationships at every level, from developers to partners to investors.

(02:06):

The goal of every founder is to solve a problem for the audience they serve. But as Sven, Ali, and Pete shared, founders who really listen and look for the white space can find it shapes their journey in unexpected and important ways. Let's dive in.

(02:30):

Sven, you co-founded Webiny in its current form in 2019, is that right?

Sven Al Hamad (02:35):

That's correct, yes.

Kate Barrett (02:37):

But it was a company before that which you led as CEO too. Just tell me about that and your path to founding this company.

Sven Al Hamad (02:44):

Well, it's all related, right? That's why I wanted to kind of keep the same name. So early on, Webiny started as your typical digital agency, so building websites and applications for different customers around the world. I started that pretty fresh out of college, and my first employee was Pavel, who's my co-founder today, whom I met on college. So we go more than 15 years back. And so that's where our journey started.

(03:13):

And the thing that sparked the interest of moving from digital agency into product was the lack of good products to do what we were doing in the agency. But it took a few attempts and it took some, well, trials and errors and finding the right market into which to go to because the competition was pretty stark at that moment.

Kate Barrett (03:37):

Sven, I think that's something a lot of founders contend with in the early days, right? We certainly heard echoes of that in Ali and Pete's conversation as they talked about the company's early efforts to gain traction.

Pete Sonsini (03:53):

Maybe let's start with just the beginning of Databricks and the invention of Apache Spark and why you decided to form a company behind it, the thinking, and go from there.

Ali Ghodsi (04:03):

Yeah, I mean, there was this 2009, '10, '11, '12 period where we had created this technology called Apache Spark. Back then, as you know, all the rage was about this thing called Hadoop. I mean, now we think it's dead and it's sort of gone, but back then it was the most hyped thing. And while Hadoop enabled something new, we felt that we could do so much better. It felt like the technology we had even before we had sort of completed it, we felt that it was several orders of magnitude better in multiple dimensions. We felt that we're 100 times faster, 100% simpler to write code, just two lines in Spark would be maybe 1,000 lines in Hadoop. And we felt that we could do many, many more use cases.

(04:45):

But it was just so hard to get the market to recognize that. It was just money flowing to these different Hadoop companies and Hadoop this and that. And nobody wanted to listen to us. And it was hard to do that from UC Berkeley sitting there in your ivory tower publishing papers. So it felt like unless we actually start a company, we're not going to really be able to break through the mold. So that was kind of the thinking. And it felt like we should be able to do it because it just felt like it's so obvious, if we just can get people to see this thing, they're going to see it themselves. We don't need to convince them.

Pete Sonsini (05:21):

And you actually had a lot of traction with Spark, though, before you started the company. So there were some signals that this was something that people found interesting before you actually raised capital though.

Ali Ghodsi (05:33):

Yeah, I mean, we had these things called Amp Camp, which was these kind of events we would say, "Come to UC Berkeley and we'll teach you how to use Spark." And we would get hundreds of people coming into those rooms.

(05:43):

So it definitely had way more traction than the usual academic stuff that we were doing, which sometimes garnered no interest from the outside. Despite getting rewards and awards and whatnot for the papers, we would get no interest. For Spark, we actually were getting interest from outside, but it was still nothing compared to what was going on with Hadoop and there were no companies commercially using it and talking about it. So it was still just basically zero despite all the attention that we were getting internally at UC Berkeley.

(06:16):

I mean, our strategy was can we just give it to these companies and they just run with it and they realize that it's awesome? And we tried, actually. We tried all kinds of strategies. We tried to send interns to these startups and tried to get the interns to be moles pushing the Apache Spark project, but it wasn't working. So 2013 felt like we just have to do it ourselves.

Kate Barrett (06:41):

Does that resonate with you, Sven, this challenge of getting the market to see the value of what you're building? Or since Webiny really evolved from its target market versus in academia, has your experience been different?

Sven Al Hamad (06:53):

In our digital agency days, we were building different tools and solutions for customers and we were always frustrated that there wasn't something better. So we've learned what better should look like, and then it was just the question of, well, first building it, and then showing it to everybody else.

Kate Barrett (07:13):

And in this case there, I think a big difference is that you were using this technology day in and day out not because it was something you'd sort of developed in the academic ivory tower, but it was your business and it was your customers and you were kind of observing that, wow, there are some issues. What were the specific things that you really felt like you could do better?

Sven Al Hamad (07:33):

Yeah. So the thing we knew right away, there was a lot of solutions to do that on the market already. That meant that the barrier to entry the market is quite high and then you need to go above that barrier to be better. The thing is that through those experience of building hundreds of different projects and websites, talking to hundreds of different customers, seeing where they struggle as users, and where we struggle as developers in using such products, we were just keeping constantly a mental map of those issues, those products.

(08:07):

We understood that the CMS market is not just about CMSs, it's not just about, "Hey, I want to build a website," but that website has content. That content needs to be delivered through omnichannel. There might be mobile apps involved in that. There might be different platforms involved in that.

(08:23):

So there was so much more that we felt a CMS should do, but there was nothing on the market that was attainable to us and that we felt would solve our needs. But then there was a challenge of how do you build such an amazing product, which is another story.

Kate Barrett (08:41):

And that takes us then probably into the serverless discussion and what the opportunity you saw for that was.

Sven Al Hamad (08:47):

Yes. So exactly, because looking at the market where we saw just hundreds of different CMSs. So how do we find our place? We had to be different. Building just a better product is very challenging unless you have one uniqueness about your product that nobody else has. And through that uniqueness you get the first customers and the first funding and then you build on top and you build on top.

(09:13):

And we were kind of sitting on the sidelines of that market, observing the market, learning the market, just waiting for the right thing which would make us unique. And that's when we discovered serverless and the whole thing clicked.

(09:27):

Serverless we feel is the future of infrastructure. It costs 80% less in terms of cloud infrastructure. It costs 60% less to maintain. It is even better for the environment. It's a funny thing, but AWS gives you a sense of measuring the energy use to run your website. Serverless helps drastically with that. It's fault-tolerant. It's scalable. And suddenly, if you build a CMS on top of it, nobody else has this combination of these unique features that it can perform at scale and be super secure and way, way cheaper than any other solution of the market.

(10:07):

The other problem was that when we started, serverless was in very early days. There was no tutorials on how do you build full Stack applications on top of it. So we couldn't just go and learn serverless, we had to pioneer a lot of things ourselves and that meant that we were a bit slower in the beginning with the product because there was just a lot of figuring stuff after along the way. But serverless was that unique element to our solution and still is today.

Kate Barrett (10:35):

Yeah, I mean, just as the technology and tools shape aspect of your product and help you differentiate, so do customers, right? That was a theme that Pete and Ali explored and it was really surprising to hear how big of a role just listening to customers played in Databricks' product development.

Pete Sonsini (10:57):

So you had a super successful project and then you were pulled, your customers basically guided you towards some of the other products that you could build on top of it. So you just stayed customer-focused and the opportunity kind of opened up in front of your eyes as opposed to you guys sitting back and saying, "I think the market needs this."

Ali Ghodsi (11:18):

Yeah, we were going to them and saying, "Hey, I think you need this." And they say, "That's fine, but I can't use it because of this other thing." Do you have enterprise security? Do you even have a way for me to store the data? How do I make sure that I can store it correctly? This Spark thing you developed has too many knobs. How do I configure all the knobs?

(11:37):

How do I do more of the operations of machine learning because I have many different people in my organization. Some of them are great, they know how to use Spark, but there's five of those guys. I have another hundred over there that'd like to get the benefits of it, but how do we share the results with them? How do we get them to help me do that? And then I need a way to make sure that all my data is high quality. I don't think Spark helps me with that. Can you help me with that? That's a big problem. If we can't fix the data quality problem we have over there, then I don't think we can really use Spark even to begin with.

(12:09):

So then it started becoming more and more clear that there's four different types of individuals involved and we're only addressing one of them. And if we can't address all four, there's no way we can make these enterprises successful. So there was the data engineers, which is where we started, the people that moved data and know how to use Spark. But then there was data scientists who wanted to build machine learning models. We were not really addressing them. We didn't really have a great product for them. And then there was the business analysts, the people that aren't sophisticated programmers who just want to get the insights. And then you have other stakeholders in the company, like administrators and other people that are not even in the data space, they also need to approve budgets. They also need to see the value of your platform, otherwise they're going to feel like it's something that they don't understand.

(12:51):

So we started building out the platform to address all of their different needs. And we started beefing up the product management function significantly. And we adopted what we called jobs-to-be-done methodology, which we were big fans of in product management. And we just sent them out and had a lot of these deep, in-depth contextual interviews with the customers trying to understand what is it you exactly need to get this going? And we started building towards those things. And that was 2015-16.

Pete Sonsini (13:19):

Yeah. And it's the winning strategy. I mean, it's so clear right now that our strategy with Delta and letting the data stay in the Delta Lake and not needing to move it all to some proprietary format and it's just one platform for data science, machine learning, analytics without a lot of heavy lift with moving data and stuff like that. It's a proprietary format. It just makes perfect sense.

(13:46):

And how you pull that off is, in hindsight it's brilliant because it's definitely the right answer as opposed to building some new mousetrap, which move all your data into this mousetrap and it'll just work. You made it work where it was, which is just incredible. And I think it's because you stay true to the open source benefits and to the open formatting and there's some things that you stay true to that were really in your roots as an open source company I think that allowed it to happen, with MLflow and Delta and stuff like that.

Ali Ghodsi (14:15):

Absolutely. I think we were humbled in the early 2000 even before the company that there are these big market forces. It's hard to fight them. We saw it with Hadoop, then later we saw it with the cloud taking over. In some cases, we were on the right side. And the other sides, it was difficult. It took a long time for the trend to happen.

(14:33):

So we wanted to really sort of not go against the big market trends. One was people had lots of data in these data lakes. They didn't want to move it into a proprietary place. We were going to have to spend a lot of energy convincing them to move it into a proprietary system. And anyway, the world is going to be more and more open and more and more standardized. And that data has gravity, so why fight that? And there's no benefits either. The cloud had many benefits, so we were very pro to cloud. But moving it into these proprietary stores didn't have any benefit. So we stayed true to let's just keep it where it is and let's make it valuable where that data is and build that platform.

(15:12):

So then we started one brick at a time building Databricks, and I think we're still on that journey. I don't think we're even close to being finished. It's like we're still looking at what are the problems. And I think we're in the beginning of this data analytics AI revolution that's going to sweep over the whole planet and all organizations. So there's a lot of work to be done, but we just need to continue listening to the customers. And that's when kind of the business model solidified.

Kate Barrett (15:40):

Sven, I'm curious, how do you think about building Webiny's platform as it expands to different types of users, different functions, and how do you prioritize those needs?

Sven Al Hamad (15:51):

Yeah. So we've got two main user types, let's say it like that. It's the developer usually picking out a solution, evaluating it from the technical perspective, installing it, maintaining it, being responsible for it. But it's CMS and CMSs are used by content editors, marketers, non-technical people. So we understand that those are the two personas we need to satisfy.

(16:18):

We started focusing first on the developers, getting that first-touch experience in a better state. Still a lot more work for us to do there. And we slowly then transitioned into, "Hey, what does this content editor want to do?" "Oh, they want to do peer review processes before they publish the content. They want to involve other departments in a peer review process. This marketing person wants to build not just static landing pages but wants to pull dynamic data from the headless CMS." So just by talking slowly to those users.

(16:48):

And in order to talk to users, you need to get your product out no matter how good or bad it is, but it is a conversation starter. And then they will say, "I've did these three steps, and then I got stuck here on the fourth and in the fourth step I wanted to do this." Okay, write it down on the roadmap.

(17:05):

And they mentioned a really good question. It's how? How will you do this? How will you achieve what you want to do? What is your use case? Just understanding the steps they are planning to take in solving their pain with your product, you will see where your product does well and where it falls short. And when it does fall short, they will tell you.

(17:27):

Really good user is the user that suffers through your project. You got the users that hit a problem and they just go away. It's hard to get them back. Then you have the users that say, "Oh, everything's great." Cool, but you're not getting super valuable feedback to improve your product. But the ones that suffer through by using your product, the ones that are trying to build something, they're kind of succeeding, but it's a lot of effort, it's hard. If you get that type of a user on a call, they have a lot of feedback, a lot of things to say about your product. We don't want to be perfect at this stage.

Kate Barrett (18:05):

Was that hard to get in that mindset at any point? I mean, I imagine coming from an agency environment to one where you're just putting stuff out there and it's kind of half-baked and seeing how people react to it, was that a hard transition for you?

Sven Al Hamad (18:18):

If you have a user that tells you you did this five things great, but I can't use your product because you're not doing this sixth thing, I want to get that sixth thing out of the door as quickly as possible just so he can kind of get over that hump. So it's not hard. It's actually, we're incentivized by that type of feedback. Just get something out, "Hey, here's something, it kind of gets you what you want. We'll make it better in the future." The user is super happy. I'm able to now go to the next step.

(18:49):

So coming from the agency days same as me, when you're working with clients, yes, everything has to be perfect, right? The client has to be happy. Here, it's different story. You got a much bigger product to build usually across a few years. You got a bigger team that's dedicated to that. But also you don't have one customer, you have thousands that are telling you this is burning, this is burning, this is burning. So it's a different environment and the environment in a way forces you to get into that mindset, I would say.

Kate Barrett (19:21):

The velocity that has to confer as an organization and just being able to move quickly probably makes up for all the rest.

Sven Al Hamad (19:29):

Yes. But the bigger challenge is actually achieving that velocity from the product/engineering standpoint, that you have the right mindset within the engineering team that we don't need the code to be perfect, right? We just need to do this thing. Let's cut some corners even if needed in this first release. Let's get the feedback, then we'll improve it. It's the mindset across the organization has to follow that approach. Otherwise, it doesn't work. You will be too slow in your release cycles. You will be too slow in satisfying the needs of your users.

Kate Barrett (20:07):

As those product development cycles ramp up, it's natural to start thinking how do you get your product in more user's hands without having a giant sales team? For Databricks, that meant forging some really powerful partnerships.

Pete Sonsini (20:26):

But the Microsoft partnership that we struck really early on, you and I spent a lot of time on that really early. Just to rewind to that for a second, because it was a hugely successful partnership and you brilliantly navigated structuring it and pulling it off. Really a win-win for Microsoft and for Databricks. Talk about how you pulled off that deal and, I mean, one of the most successful partnership deals really you've seen and maybe since the VMware IBM days perhaps.

Ali Ghodsi (20:55):

Yeah, look, Microsoft has 77% of Fortune 2000 on enterprise agreements. So that means 77%, so close to 80%, right? So four out of five companies on Fortune 2000 already are Microsoft customers. That's been the case for 40 years, before cloud, before all of this kind of stuff, which means at Microsoft, the name of the game is when they release a new product, they don't need to go find customers for it. They put it in the enterprise agreement. So then it's just a matter of can you get people to click on that product and use it?

(21:29):

The whole idea was if Databricks can become part of that enterprise agreement, it can become a skew in that basically four out of five companies on the planet are our customers immediately overnight. Then all that we need to do is get them to click on the product and use it, which is still you have to educate them. You have to make them successful. But it means you no longer have to go negotiate with them. You don't need to send your salespeople in there to get them to talk to procurement and discuss how do we buy this from you and talk to security and figure out is this secure and can you trust us and what's the indemnity and what's the limitation of liability and all those data privacy and all those kind of things.

(22:07):

We are Microsoft, right? The idea was to have a partnership that could do that. And it was hard to pull off because Microsoft was this giant company. At the time, we were far, far below 100 million ARR when we did the partnership. This was the year we started talking to them about partnership.

Pete Sonsini (22:23):

It was like 20 million or something.

Ali Ghodsi (22:25):

Yeah, exactly. That's the year we did roughly 20 million revenue. So working with the big machine of Microsoft, and this was sort of a special deal that the department was putting together, working with all their lawyers and all the different stakeholders, it was massive.

(22:40):

It was 150, 160-page custom agreement and we didn't have a big legal team. Our legal team was maybe two people at the time. So I combed through that agreement line by line. I knew every sort of paragraph. Of course, we used outside counsel and so on. It was super important because you're signing this big deal that could have massive implications on the future of your company. And of course, we ran it and the board was involved in every step of the way because I personally felt like I don't want to sign this thing, put my name on it.

Pete Sonsini (23:07):

I remember that. But how did you get their interest though? I mean, you were personally compelling with management. They saw the growth of Spark and they knew that you were important to machine learning, which was a hot thing, but there were a lot of companies that probably could check that box. I mean, how did you just pull it off? Was it really about a relationship that you struck with one individual, which it usually is?

Ali Ghodsi (23:32):

That is definitely the case, but there were several of those individuals at Microsoft. I think Satya Nadella was very important. From day one, he wanted a new Microsoft which is partner-friendly, and he told me first meeting I had with him three times I think in that meeting, "I want you to make a lot of money." So it was a partner-friendly sort of organization suddenly after the CEO change had happened. So that was important. So that tone was coming from the top that we need to make this happen.

(24:00):

The second thing is convince the product and engineering teams who always want to build stuff themselves not to invent it here, I'm going to build this myself, convincing them that these Databrick's guys have something that's really, really good. It's a slick product. And I remember they put together one meeting where they brought in people from Microsoft Research and they brought in all their heavy hitters, professors that were working at Microsoft and so on, I mean famous names.

(24:23):

And they had two hours where they just bombarded me with technical questions. How would you solve this problem? How would you solve that problem? And I think after that meeting, they kind of felt like these guys know what they're doing. They're pretty damn good. They're no idiots. So I think there was a thumbs up after that meeting. That was important.

(24:39):

But what was fascinating was that once this thing hit the ground, we started building it. It grew zero to 100 million ARR in record time, in far less than a year. I think in eight or nine months, we did 100 million ARR overnight. And it's back to the everybody's already a customer, you just have to get their mouse pointer to go click on the product and use it and then magic happens.

Sven Al Hamad (25:04):

So from the perspective of partnerships, partnerships are for getting to the enterprise deals. I think they are crucial. We are getting invited to join different product demos and things like that by different partners, be that system integrators, be that big cloud vendors. They got the deal flow, so you need to play with them.

(25:26):

There's a challenge, however, as a small startup, some of these partnerships require formal process. They require that you have a minimal commitment or a spend on their platforms in order to get deals from their platforms. There's a lot of challenges for a very small company to take on a burden even of partnering with such a large vendor.

(25:48):

We've tried some unsuccessfully purely because we're not yet at that right scale and the effort required is too high. Some partners told us directly, "Hey, call us when you have minimum five salespeople because we are going to send you a lot of deals and unless you are able to scale to that level, you won't be able to handle them right." But at the same time, for the stage we are at, some of those things are not a priority to get to that scale right away.

(26:20):

We do have partners which are smaller system integrators or smaller digital agencies, and they are bringing the deal flow to us because partnerships have a really great value because they have established relationships with the customer. And then you just piggyback on that and that means a lot because you don't need to build the trust from ground up. And that helps with closing the deals. And especially in the enterprise world, six to 12 months is the time you need to be ready to invest into a single enterprise deal. And for a small organization with just one and a half salesperson, that is a lot. You can't do many of these deals in parallel.

Kate Barrett (27:03):

Well, and you would need a legal team and you would need, I'm sure there's a whole bunch of...

Sven Al Hamad (27:06):

Oh, I'm the legal team. I'm the...

Kate Barrett (27:07):

You're the legal team?

Sven Al Hamad (27:09):

My other salesperson is a legal advisor.

Kate Barrett (27:12):

Oh, gosh.

Sven Al Hamad (27:13):

We're doing everything we can ourselves at certain stages. If it gets tricky, then we pull in the legal teams. But legal teams are also, I mean, in certain cases, they are not worth the investment if the deal is not six-or-seven- figure deal, which for a startup you don't get a lot of these in the beginning.

(27:31):

So I understand where Databricks came and the stage they were at when they joined in partnership with Microsoft, but they were still a later stage than we are today. But you need to know what is the right thing for you to do at a certain stage because going into some of these deals, it can take away so much focus and so much effort from 50 or 60% of your organization and then the product is going to suffer. The community is going to suffer. It's just going to potentially ruin your company if you're not ready for dealing at that scale.

(28:11):

But luckily, I mean, you've got partners that are smaller ones and that's where we are starting. We were learning how do you navigate a relationship with a partner and then what is the relationship and how do you navigate that once the partner introduces you to their customer. There's a lot of learnings to be done there.

(28:31):

And unlike sales, in terms of direct sales, which is documented, there's frameworks, there's a lot of learning materials, navigating partners is an area that has a lot of resources, at least what we found, that could apply to the company at our scale. It's more me asking fellow founders in terms of, "Hey, what did you feel like worked for you?" And for some it worked one thing, for others it worked something else.

Kate Barrett (29:00):

That is a really important point, I think. A founder needs to balance conviction and ambition with caution. I mean, Pete spoke similarly about Ali striking that balance in his approach.

Pete Sonsini (29:15):

One thing I've always really admired about you and your CEO style is that you're obviously super confident person and so, so talented, but you stay humble and that you really actually seek advice from everybody and do it in a way that disarms people and gets them giving you information, which I think is an important part of your management style, but you also seek it from everybody. And not everybody does that. And you seek it from the lowest level, just all the way up.

(29:44):

So the people that you've worked with and how you've made your decisions and gathered information, I think would be really interesting to talk about a little bit. And then from there, who you've decided to work with, both on the board level as well as your hires and some of your key hires that may have taken longer than some of us would've liked.

Ali Ghodsi (30:06):

Yeah. Look, for me, it's all about the trust that you build with people around you. A lot of things just come down to that. In my previous life as a researcher, you publish papers, talking to computers, they're very binary. It's super simple contract. When it comes to humans, it's all about trust that you're building with them, and that's a relationship you build over the years. The only way you can do that is by talking often and forming that relationship and being vulnerable both ways, right? "Hey, Pete, this thing is kind of bothering me. I'm thinking about it. How do you think about it? What do you think I should do? You were at VMware, you struck that mega, mega deal. You know how it works. Am I thinking through this the right way? What do you really think?"

(30:46):

So that way you can sort of by opening up and being vulnerable about the things that you're thinking about, people will then try to help you and maybe they're on your side now. Instead of I come in and say, "Pete, I've already figured everything out. I absolutely want to do it this way," and not listen to you. And then you try to tell me how you've done things in your previous career. And I said, "I don't care about that. That was a different era. This is what I want to do." That doesn't build trust.

(31:08):

So that way we can figure things out together. Ultimately, I have to make the decision anyway, so decision is mine, but that way I'm actually partnering with people around me and picking their brains and picking their experiences. I think the experiences that people have had are extremely important and extremely valuable.

(31:22):

And we always think that what's happening right now is so unique and amazing and we're doing something that's never done been before, but there were many great companies in the past and there were lots of great experiences. And those great companies today, maybe they're not anymore those sexy, hot companies that they used to be, but actually many of the elements of what was going on back then is similar to what's going on now.

(31:42):

So tap into that. Especially for me as a first-time CEO, first-time entrepreneur, first job actually outside of academia, I wanted to make sure that I'm learning from what happened in the past. So I spent a lot of time building those relationships and picking people's brains and then letting that inform the decision that ultimately is my decision, but it's informed. And people change my mind.

(32:02):

And those people also came to save us when the going got rough. For instance, 2015 was a rough patch for Databricks. It was a rough patch actually in the market in general, but funding kind of dried up. It was this era of everybody keeps doubling their valuation and doubling the number of employees they have and revenue doesn't really matter. So everybody was on that kind of journey. And then suddenly funding dries up for everybody and it's like, "No, why should I give you any more money? Why would you double your valuation and get more money unless you can prove that you've doubled your revenue?"

(32:35):

So we went through sort of a similar thing in 2015. It started really well in March 2015, but then by end of 2015, we really couldn't raise any funds. So basically what was going to happen is we would have to go back to the employees and say, "This rocket ship we're on, we're not going to get a big up round. In fact, we weren't even able to secure money, or if we were able to secure money, it's maybe even lower than the previous valuation. And we have some hard decisions to make." And you could see the movie, we might have to let people go and so on. And that kind of just changes the trajectory of the company.

(33:09):

And really appreciated the partnership with you, Pete, that year towards the end, throwing your lifeline for us when you didn't have to because we were picking your brain on how to do the fundraise. So you were kind of conflicted. So you knew what was going on. You knew that basically we don't have any options. And so you could have just said, "Hey, I'm going to be cutthroat here and not throw you the lifeline because nobody else is either."

Pete Sonsini (33:34):

Yeah.

Ali Ghodsi (33:35):

But you did throw it. And what I meant is we could go back to the employees and say, "Hey, told you this is an amazing company. We just raised money at a massive valuation. We're continuing to build the future." And we didn't need to do that kind of cultural shift towards letting people go and everything becomes gloomy and so on.

Pete Sonsini (33:52):

Yeah, I remember that. I remember that. Fortunately, I had enough experience as a venture guy at that point to know that it was an easy decision and I shouldn't try to squeeze you guys for a lower valuation. So I'm glad that I'd seen a few things at that point, that it was clearly the right decision to support you and keep the momentum going with the company. So it actually was not a hard decision at all.

Ali Ghodsi (34:16):

Well, thanks for all the support, Pete, all these years. We wouldn't be here without you.

Pete Sonsini (34:20):

Thanks so much. I don't know about that, but thank you.

Kate Barrett (34:25):

Sven, it's been so much fun talking to you today. Thank you so much for joining me, answering lots of my questions about all this, but really it's been awesome to get an idea of what you're building and where you're headed.

Sven Al Hamad (34:35):

Thank you, Kate. I felt challenged, but in a good way.

Kate Barrett (34:40):

Hopefully, not too challenged from me.

Sven Al Hamad (34:42):

Not too challenged, but in a good way. It made me think about certain things that I haven't thought for in a while and I just kind of had to reassess, "Hey, why do I value this?" It resurfaces some of the, let's say, fundamentals behind which we're building Webiny. So it was a really good refresher and had a ton of fun.

Kate Barrett (35:07):

It's so inspiring to hear a founder talk about building solutions that meet the needs of the communities they serve. It's inspiring and instructive to hear a founder describe how that journey can be more of a circle than a straight line, that listening to your customers and to the market can shape a company's trajectory and accelerate its growth. I think exploring these dynamics of company-building offers a lot that founders can take away and apply in their own journey no matter what type of business they're building.

(35:35):

I'd like to thank Sven, Ali, and Pete for sharing their inspiring journeys with us. It's been a pleasure to hear about their experiences, and I hope you learned a lot too. Thank you for joining us.

(35:49):

Founder Forward is a production from NEA, made in partnership with FRQNCY Media. From NEA, I'm your Host and Executive Producer, Kate Barrett, with support from Ashley Mitchell, Erica Sunkin, and Shanna Hendriks. From FRQNCY Media, Michelle Khoui is our Executive Producer. Enna Garkusha is our Supervising Producer. Jordan Rizzieri is our Producer, and Catherine Devine and Emily Krumberger are our Associate Producers. Our Mixer and Sound Designer is Claire Bidigare-Curtis with dialogue editing by Sydnee Evans.

(36:22):

For more on NEA, visit nea.com. You can subscribe to Founder Forward on Apple Podcasts, Spotify, and wherever you get your favorite podcasts.