The following interview is a conversation we had with Bob van Luijt, CEO and Co-Founder of Weaviate, on our podcast Category Visionaries. You can view the full episode here: Over $67 Million Raised to Power the Future of Vector
Bob Van Luijt
Thanks for having me, Brett. Great to be here.
Brett
Not a problem. And I’m super excited about our conversation. Let’s go ahead and just kick off with a little bit more about who you are and your background.
Bob Van Luijt
Yeah, sure. So I’m Bob, as you said. So Bob is the easy part of my name. The rest is Dutch. It has the famous u I j t in there. So in Dutch, you would say. And that gives away where I’m from. I’m Dutch, and I’ve been working in software for a long time, and it has to do with the fact that I’m born in 85. So when I was coming off h, I guess so when I was like around 15, younger even, we got the Internet at home, and I very quickly figured out, hey, wait a second, I’m building some skills on the web, and people are interested in that. And so I created my first business around it before I was 20, actually. And it was like, it was nothing fancy.
Bob Van Luijt
It was just, I just sold websites and those kind of things. And then when it was time for me to go to college, I studied music, actually. And in the Netherlands and in Boston all the time, I was just writing software, basically to sell websites, e commerce things, and those kind of things. And then after I was done studying, I decided that I really wanted to go into software. So I started the consulting business. And when I was in the consulting business, I got in touch in. It was around 2014, 2015 with machine learning, and I decided that’s what I want to do. And that’s what ended up being Weaviate today. So that’s in a nutshell. My background.
Brett
Take us back to 2019. When you first were having the idea to start Weaviate, what were those early conversations like? And what was it about Weaviate that made you say, yep, this is it? I’m going to go build this into a business.
Bob Van Luijt
Yeah. So it was slightly before that because it’s important to know that we create is open source. So the open source project existed before we started the business. The original story actually goes back to 2015. So I was working with language models today. When you read the news, you read about large language models back then, they were like very tiny and they were all often focusing on single words. But one of the things that these models did was that they outputted something called vector embeddings. It’s a technical term, but what it basically means they too, back then, they took individual words and they gave coordinates to these words in space. And that sounds very technical, but it’s actually pretty simple. If I give you an example.
Bob Van Luijt
So, for example, if you take the two words Eiffel Tower and you combine them together, then the distance between the word Paris was smaller than, for example, a London. And that had to do with the fact that if you read a book or if you read something online, you can kind of see that certain words, they sit together. Chances that Paris is mentioned in the context of the Eiffel Tower is, of course, way bigger than, I don’t know, New York is mentioned in the context of the Eiffel Tower. And that’s kind of where everything started. So what we see now with these large language models, those kind of things, it all goes back to that moment. It was something called glove and fastex.
Bob Van Luijt
And one of the things that I figured out was like, hey, wait a second, we can do better search and recommendation with this. And infrastructure that dealt with that did not exist. So that’s how we started today. We call it a vector database, but back then we didn’t even call it a vector database. It was a category that did not exist. And then I was at a Google I O conference in 2016 where the CEO of Google, Sondra Pichai, said, like, we’re going to move from mobile first to AI first. And I was like, I know exactly what to do. They’re taking these vector embeddings that I’ve seen in these machine learning models and they use them for search for recommendation page indexing.
Bob Van Luijt
Of course, I didn’t know exactly what they were doing, but I got it immediately and I just was like, I’m going to start a project that is going to be a search engine that focuses on doing that not at the skill of Google, but doing that for any developer, at any company, at any skill that wants to build with machine learning, too. And important to bear in mind, Brett, that it took me years to talk about this concept because AI now is very hot. But I can tell you it was not in 2016 and it started to come up in 2019. And in 2019 we founded the business and we raised money and those kind of things.
Bob Van Luijt
But there was really some time that I was like traveling through the US, traveling through Europe, even went to Japan where I was on stage showing people what you could do with machine learning and search. But it took a long time.
Brett
What prompted you to decide to start the company then in 2019? And in 2016, when you were first starting the project, did you know for sure you were going to build a company eventually?
Bob Van Luijt
That’s a great question. So as a quick preamble to answer the question, I never worked for a boss, so I always had my own businesses, but those were more in software consulting. And someone is like, I want to have a product. I want to build a product business rather than a pure consulting business. And when the project started to grow, I was like, this is it. This is the opportunity, this is a product. So I just jumped right in and it was not something that I planned. It just went organic and everything came together. I wanted to have a product business, I wanted to do something in machine learning. And then there was Weaviate. Then it just all came together, met amazing co founders. So it was like everything came together and I was like, this is the moment.
Bob Van Luijt
We’re going to turn this into a product business rather than a consulting business.
Brett
And as you started the company, what was your mindset in terms of marketing?
Bob Van Luijt
So I think that marketing is the second most important thing in the business. So I think that the order of importance in building product businesses is product marketing sales. And I know that some people have different opinions on it, especially in b, two c. It sometimes even starts with marketing. But if you run a b, two b business and you sell to developers, marketing is extremely important because you can create the most awesome developer tool around. But if nobody knows that it exists, then nobody’s going to use it. And the nice thing of creating marketing strategies towards developers is that developers are very hard to market. But that’s actually, that makes it fun. So there’s all these strategies that you can apply that we’re applying at. We’ve yet to get some mind share of the developers.
Bob Van Luijt
And what’s important to know for your audience is, I don’t know if people know this, but the trick in a software business nowadays, as it hasn’t always been the case, but is that you try to create a bottom up approach. And bottom up refers to bottom up through the organization. So rather than selling to CTOs and CIO’s, you sell bottom up. So you go to the developers, you make the developers adopt your technology and they move it up in the organization. And all the marketing strategies and those kind of things are built around that go to market motion.
Brett
Tactically speaking, what were some of those early tactics and strategies that you deployed to get the first paying customers?
Bob Van Luijt
Yeah, so there’s a famous strategy which is called the three h strategy or h three. I always mix it up. I believe it’s called three h strategy. And it’s a marketing strategy that actually comes from youtubers. So what you do is that you basically try to create a funnel. And this funnel is created out for influencers is called hero hub hygiene content. And what you basically do is that hub content is something that’s very wide. For example, let’s say that you want to market a eyeliner or something, or like a beauty product. Then you might hire like a movie star or something like that. And you make sure that people see these people associated with your product, even if the product is just simply not mentioned at all. You just want to get a lot of eyeballs. You want to make something very fancy.
Bob Van Luijt
And then a step lower you have something called hub content. And in hub content you might zoom in on what’s in the hero content and was like, hey, actually the eyeliner that this person is wearing or whatever beauty product you have is our product. And hygiene refers to the fact that you explain how to apply the eyeliner and those kind of things. And the idea is that top down, that means that with hero content, you just try to get to a lot of eyeballs. With hub content, if people are interested, you show them how they can do something with your product. And the hygiene part is basically people who are already sold to your product, show them how they can use it. And it turns out that this youtuber strategy works very well for software.
Bob Van Luijt
So what we have, for example, we create is, I’ll give you two examples where we have a podcast. So with the podcast, we just talk to people, invite people who. Anybody has something important to say in the space of AI. They’re like guests on our podcast. So there can be people that create the models, that can be people who create tooling, that can be researchers, that can be business people. Anybody has something to say. Maybe the name we’ve yet is only mentioned, maybe in the intro, or you can only see it in logo. What happens there is that you fetch a lot of eyeballs, people are very interested in it, but it’s just associating yourself with that space, in this case of AI, and we do that in content on social media too.
Bob Van Luijt
So you will see a lot of we create content that’s on the wider space of AI and we create is somewhere in there, but it might be just radically mentioned. Then hub content goes a step deeper, where we tell people like, hey, if you want to build something like this or you want to do something like we’ve just shown you in these, we talked about in the podcast of what you’ve seen on the social media posts. This is how you build it yourself. And then of course, that’s why we show how you do that with VP. And then the hygiene part is something we have in we create that. We call, for example, the academy.
Bob Van Luijt
And in the academy, people really learn to work with the database in detail, but those are people who are already sold on using the database or maybe still in POC phase, but very close to being sold on using the database. And that marketing strategy, that works very well for a software.
Brett
How do you think about measuring the Roi of the podcast? Obviously, I’m a big fan of podcasting and a lot of the companies that I work with and the conversations I have, there’s always concerns about Roi, Roi. So how do you think about ROi when it comes to your podcast efforts?
Bob Van Luijt
Yeah, so what I just shared is we could see that as a big funnel, as I always like say, I just see it as an extension of the funnel. And how we measure success with that. Is the success really at the top of the funnel? So that can be through the podcast that can be to content that we’re creating is mostly measured through shares, likes and those kind of things, because it’s very hard to measure it in a form. Like if a dollar comes in and that somebody was, oh, yeah, four months ago, I saw this social media post, or I saw this, I listened to this podcast and you came on my radar, and now, therefore, I’m buying right now. That’s not how it works.
Bob Van Luijt
I often compare it more with these soap commercials we saw back in the days or these toilet paper commercials. There were so many of them. But the thing was, the moment that you had to buy soap or toilet paper in the supermarket, you associated the brand with something that you knew. So I have a lot of people that come to me like, oh, we love the weavy podcast, or oh, we love the posts that you guys put out, but they haven’t bought anything yet. Not even a competitor, just nothing. But now there’s this brand awareness. So at this top of the funnel, I am perfectly fine with the fact that it’s also an investment in brand awareness. But what we measure is things like retweets, likes, engagement signups to our slack channel signups to free sandboxes and those kind of things.
Bob Van Luijt
It is more if you go a little bit further down in that h three funnel. So if you go to the hub and especially to the hygiene content, then it’s way easier to measure. So everything that’s really top of funnel, of that marketing funnel, I’m okay with just experimenting and just not being sure how certain things will work out. But the further you go down in the funnel, the more clear it becomes what the ROI is.
Brett
What about building your marketing team? What does the marketing team look like today? And then where did it begin? Who were some of those first marketing hires that you made?
Bob Van Luijt
Yeah. So at Weaviate I’m taking the liberty to look at marketing in the broad sense of the word. We have three teams that are doing something related to marketing. So one team is what we call the developer growth team, another team is what we call the developer relations team and the other team we call just the marketing team. But that’s more product marketing. Now what these three teams do is that the developer growth team focuses on winning new users. So the idea is like getting people in, making people aware that our developers in this kit, making developers, where do we exist? Telling developers, hey, this is what you can build, hey, you want to do this cool AI stuff, this is how you can do it. And so on and so forth.
Bob Van Luijt
They build example applications, they make videos, they make the podcast, they make content, all these kind of things. The focus is on quantity just as much as we can, capturing as much mind share as we can. Then the second team that we have is the developer relations team. The developer relation team focuses on developers that have already decided that they probably want to work with weve. So here it’s more of focus on quality, if you will, in the sense that the content that we’re creating for them, that comes in the form of the software documentation, that comes in the form of we have the, we’ve academy those kind of things, the certain blog posts that people really can go into depth if they’re building AI apps. And then the marketing itself is focusing more on product marketing.
Bob Van Luijt
So for example, last year in November I believe, or early December, there was this really big conference in Las Vegas from Amazon or from Amazon Web Services, I should say Amazon Reinvent. And that is one thing that the marketing team focuses on and they all work in harmony together. And that’s how I’ve structured marketing as a whole. Inside brief yet.
Brett
What about thought leadership? How do you think about thought leadership in terms of your marketing efforts and go to marketing efforts?
Bob Van Luijt
It’s very similar to that h three model or three h model. I’m very much the guy who likes to think in the form of models. So everything that I can’t model out, I like. What do I mean with that? So let’s take myself right, what I’m doing right now on a meta level, this podcast right now, that is an example of hero content, right? It’s me as the founder of Eva talking about developer marketing. And your audience will listen to this and it’s helping us with brand awareness. Like, hey, this is what’s happening in the eye space. This is what these folks are working on. And it’s gonna be harder to measure how that might play out. But it’s a way to reach an audience, right? It’s building brand.
Bob Van Luijt
In two weeks from now, I’m going to be at a big conference, speaking in new York. It’s called the data driven New York. I believe something like that might be wrong, but something like that there. I’m more going to talk about our technology. I’m going to give demos and those kind of things. But it’s a pretty big audience there. That’s hub content, because that audience is really there to learn about AI. They might want to engage with a product, they might need it for their day to day job. And hygiene might be something. And I have to admit that I don’t do that much anymore, but in that ph kind of context, that would be just me sitting with a customer and said like, hey, show me your problem. Let me show you how we can help and then we really dive in.
Bob Van Luijt
Or you could do that in a meetup or in a panel or in a video and those kind of things. Back in the day when the company was smaller, I did all these things. Now I focus more on hub content, but that’s how I do it. And I also write articles about business building on Forbes and those kinds of things. That is a way, just a form of thought leadership in the wider sense of the word. So not only about what we do as a business, not only about AI, but also software, business building and those kind of things. And one of the things that I always, what I do too is that sometimes, for example, people invite me at universities or something to talk to the students.
Bob Van Luijt
And then I always ask for it to be recorded because if the talk is recorded, we can publish it on, for example, YouTube. And then it has a marketing function again. So it’s constantly looking at like, okay, what I’m doing. Where does that sit in that create strategy and how can it help the business?
Brett
What about your market category? Is the actual. This show is brought to you by Front Lines Media podcast production studio that helps B2B founders launch, manage, and grow their own podcast. Now, if you’re a founder, you may be thinking, I don’t have time to host a podcast. I’ve got a company to build. Well, that’s exactly what we built our service to do. You show up and host, and we handle literally everything else. To set up a call to discuss launching your own podcast, visit frontlines.io podcast. Now back today’s episode. What about your market category? Is the actual category vector database, or what’s the category?
Bob Van Luijt
Yes, it is. And this is something that has been one of the most fun things in my career right now. So if you. Vector database is a subcategory of the database market, and the database market itself is a submarket of software infrastructure. And that’s exciting because it’s kind of an invisible category, because it’s not like, I don’t know, Coca Cola that you see it everywhere, right? But the thing is, the software and technology is so omnipresent that these markets are enormous. They’re gigantic. And what is so much fun to do is that the moment that machine learning AI took the stage, if you will, there was this opportunity for all these new infrastructure companies, including in the database space. So we’re not only creating a new product that we want to position, but together with our competitors, we’re also creating a new category.
Bob Van Luijt
And that’s maybe one of the most fun things I’ve done when it comes to marketing in my career until now.
Brett
Our audience loves to learn about growth and any numbers and metrics that you can share. So is there anything that you can share that just highlights the growth and traction that you’re seeing today?
Bob Van Luijt
Sure. So let me share how we do that. So I don’t have the exact numbers, but now we got like a couple of millions of downloads because we’re open source. So that’s where it starts, right? Those are not customers, but those are users. And so the difference between a user and customer is that, as I always like to say, a user uses your product, but the customer has a dollar going from their bank account to your bank account. So a customer is always a user, but a user does not have to be a customer. But that’s now in the. There’s a couple of million, if I had to say from the top, I had four or five. One of the things that we’ve did that we launched last April, so this, less than a year ago, was our serverless offering.
Bob Van Luijt
And we have a enterprise offering, and internally we call our surplus offering. We call that our retail customers. That’s because these developers, they behave like retailers. They spin stuff up, they shut it down, and those kind of things. They want easy access at the swipe of credit card. And then we have our enterprise customers, and those are more sophisticated, bigger deals where people want to have certain deployments of the database. And that really took off. We launched the product just after Chedgpt was launched, because that was in November 2022, I believe, if I say this correctly. And we launched our paid product in April 2023. And so were right in the slipstream of that. And that’s all going very well. So you’re talking to a very happy least founder.
Brett
Good. I like talking to happy founders. Now, you reflect on this growth and all the success that you and the company have had. What do you think you’ve really gotten right? If you had to choose, like, one big thing people, how did you get the right people?
Bob Van Luijt
Ooh, that’s an excellent question. And that’s a good question to think about the answer. So let me try to get to an end. Let me try to think out loud and get to answer, because it’s an excellent question. But it’s also a very complex question. I very early on figured out, like, oh, boy, if I want to do this right, I need amazing people, because I just can’t do all this stuff alone. I mean, I have strong opinions, but they are very loosely held. And sometimes I meet people in my space that are so smart or so creative. So I was like, I need to somehow work with these people. And I was a little bit helped by the market because AI is now, of course, very hot. So that’s, for certain people, an attractive market to work in.
Bob Van Luijt
But the question for me became like, how do I do that, right? How do I find the right people and build the right team? Because, again, these kind of things, you can’t do them alone, right? And one thing that I got convinced about is that it’s important to me that people enjoy working at the company. So a word like fun or like having fun in building the company is extremely important. And so I just met these great people, and I just, you know, try to give them a very warm welcome in the company. And that worked very well for them. And later on, I started to really double down on that. So, for example, from the company values that we have, the first company value that we have is kindness.
Bob Van Luijt
Because one of the things that I learned, especially when competition became more fierce, and, of course, we had certain targets and certain goals and those kind of things, one of the things that I learned was sometimes that was also tough, right? So it’s like, it’s not always easy. So if I would tell people to have fun or to be happy or whatever, right? That was kind of shallow because I was like, I can say that, but of course, I can’t look into people’s minds. And then all of a sudden, it struck me that I was like, well, there’s this one thing that sits in the vicinity of all these words, and that is, like, being kind or kindness. And I was like, because kindness is something you don’t receive. It’s something you give. And I really started to double down on that.
Bob Van Luijt
So it’s like, we have to be kind to each other, to our customers, to our community, but also to our competitors and the people that we engage with. And so what started to happen is that we started to build this amazing group of people that really live by that. So we got emails from customers recently, I got a text message from a customer that I actually, with permission, tweeted out where somebody, literally a customer asked me, like, why is everybody in your company so kind? And that is how I’ve been doing it. And now to your question, so how did I find these people? I think that people see that. So the language we use, the way that we market our product, people see that’s who we are, how we behave, and that attracts people.
Bob Van Luijt
We have received hundreds, if not thousands of resumes of people who want to work at the company. So that’s how I’ve been doing it. Not sure if that answers your question, but that’s how I’ve been doing it.
Brett
It’s a complex question, and I think there’s not easy answers. So I think that was a very useful answer given the complexity of what I asked. Now, let’s imagine, Bob, that I come to you and I say, okay, Bob, I’m going to start a startup. I’m going to target developers. What would be the number one piece of marketing advice that you’d give me before I begin on that journey?
Bob Van Luijt
Help people be successful in what they want to build. Don’t push your technology. Help them be successful using your technology. There’s so many examples that I can give, and that includes ourselves, where we’ve been doing a wrong. So that you want to push your technology, but you should do it the other way around. What is the problem that somebody is trying to solve? How can your technology help achieve that? And you want to become very sophisticated, not only in conversations or when you give demos, but in the content that you’re writing, in the videos that you’re making. The keyword here is help. Help people be successful with whatever they’re trying to solve. And don’t try to please everybody. Right? It’s fine if people. If you say, okay, you know, I have a problem, but it turns out I don’t need your product, that’s fine.
Bob Van Luijt
And then you just introduce them to another product or another solution. But the thing is, focus on helping.
Brett
As I mentioned there in the intro, you’ve raised over 67 million to date. What have you learned about fundraising throughout this journey?
Bob Van Luijt
What have I learned about fundraising? One of the things that I’ve been. So I’ve only had one bitch deck in my life, and that was for our seed round. And all the other rounds came very organically, that is not to be confused, that people just randomly wire money. That’s not how it works. But I started to build these relationships very early on. Sometimes we talk, like over a year. So the company or the firm who let our last investment round, let’s say that’s a firm called index. And the investor, Erin, she wrote a beautiful blog post about how we’ve met in San Francisco, how we talked about open source, about business building, and it was like over a year before the investment.
Bob Van Luijt
So what starts to happen is you have these conversations, you stay in touch, you communicate with each other, you meet each other to different companies, and then you get to this state where it becomes very logical to work together. It’s like, all right, now it makes a lot of sense to work together. And so to your question, what I’ve learned is that if you need to force it too much, there’s something wrong in that. It can be your product, it can be your go to market, that can be your market. It can be anything. But then there’s something off. It should not take too much friction. If there’s too much friction, there’s something wrong. And that can also, by the way, that could also be the founders. Sometimes people go like, hey, I have this great product, but I’m not raising any money.
Bob Van Luijt
And, well, you know, it’s you. It’s not products. My point is, sometimes you also want to look in the mirror.
Brett
I love it. All right, Bob, final question. Since we’re almost up on time here, let’s zoom out three to five years into the future. What’s this big picture vision that you’re building?
Bob Van Luijt
So, as I mentioned before, we have database technology. That’s a form of infrastructure. And I am convinced that AI, in the broad sense of the word, will not only be something that’s like, sprinkled over products that you and I use today. No, no, it will be at the heart of these products. So almost nobody is building new products without AI today. We need to have what I like to call AI native infrastructure for that. So we’ve had, as an AI native database, that means we help you be successful building AI native technologies. Yes, today we focus a lot on the database because that’s how we started, that’s where all our customers are coming from. But we want to grow into this AI native platform where we enable developers. Doesn’t matter what size company you work for, it doesn’t matter where you’re coming from.
Bob Van Luijt
If you’re a developer that wants to build AI native applications, we want to be here to help from the infrastructure level. And that’s, what was it, three to five years from now? The future of amazing.
Brett
Love the vision, and I really love this conversation. We are up on time, so we’ll need to wrap here before we do. If there’s any founders that are listening in that want to follow along with your journey, where should they go?
Bob Van Luijt
So I’m very active on LinkedIn, Twitter, X, but people can also reach out to me over email. It’s just pop at Weaviate.io does wonders. I love to hear from founders, and if I can help. Sometimes it takes a lot of time, but I try to always respond.
Brett
Amazing. Bob, thanks so much for taking the time. It’s been a lot of fun.
Bob Van Luijt
Thanks for having me, Brett. Appreciate it.
Brett
No problem. Keep in touch. This episode of Category Visionaries is brought to you by Front Lines Media, Silicon Valley’s leading podcast production studio. If you’re a B2B founder looking for help launching and growing your own podcast, visit Frontlines.io podcast. And for the latest episode, search for category visionaries on your podcast platform of choice. Thanks for listening, and we’ll catch you on the next episode.
Twitter
facebook
linkedin