E1 - Podcast transcription

RVB: 00:00:00.684 Hello, everyone. My name is Rik, Rik Van Bruggen from Neo4j. And I am super excited to be on this podcast recording together with my dear friend, Stefan Wendin. Hi, Stefan.

SW: 00:00:15.276 Hello, Rik. Nice to see you again and nice meeting up in person the other week. Super cool.

RVB: 00:00:21.355 Yeah. That was great. That was fantastic. We actually got a chance to see each other live. Right? But today we're not going to be face to face. We're actually remote on two sides of a Zoom call. And we are embarking on a quest. Right? So that's a little bit of a special thing that me and Stefan have been creating all summer. This is the first episode of a multi-episode series. And what are we going to do? This is actually a little bit of an experiment, and I'm hoping it will turn out well. We prepared for it. But it's going to be all about graph value. Right? What we want to do in the next couple of series is to explore together with each other and with you-- to explore and to document and to structure our thinking about, how do we get graphs, these wonderful structures that we all know and love-- how do we get them to deliver value for an organisation? Right? And that's a big ask. That's a big assignment. We're going to be exploring this for a number of different angles over a number of different weeks over a number of different episodes. We're going to be writing about this. We're going to be publishing articles about this. And hopefully, also, we're going to be able to present this on a number of different fora because we think it's important. Graphs have a lot of value, but it's very poorly understood in many cases. So we're embarking on this journey. Bear with us. It's going to take a couple of weeks, and today is the first episode. Right? So it's the first episode, where we're going to be publishing a little bit of our thinking around graph value. So the first question, Stefan?

SW: 00:02:14.746 Yeah. The first question, why the heck do this quest for graph value? I guess that's my first question. I think that's enough to double down on.

RVB: 00:02:25.066 There's a lot there. Right? I mean, why is this important? Why do we need to have a quest for graph value? Why are we so eager to document it? Well, multiple things here, and I'll start with my first big thought about this. I mean, we can talk a little bit about it. Right, Stefan?

SW: 00:02:45.037 Yeah.

RVB: 00:02:45.087 But the first big thought here is that we've been publishing everywhere that graphs, property graphs specifically, are actually super multifunctional, multifaceted, that you can use them for almost everything. Graphs are everywhere. We have these beautiful little stickers that we put on our laptops. It's actually really fascinating. It's a multipurpose, multifunctional data structure that has so many different use cases. But the question obviously is, is it valuable? Right? Is it worth anything? Right? And I think we're trying to answer that question, whether or not it's useful for someone to use a data structure like graphs for a particular use case. Right? I mean, you've seen this in multiple cases. Right, Stefan? There's use cases where it fits well, and there's use cases where it fits a little bit less well.

SW: 00:03:45.590 Yeah, totally. And basically, I guess, trying to go from a Swiss Army knife, because you can apply graphs to everywhere, but kind of figure out where to use it and kind of go from this idea of the jack of all trades to the jack of all values maybe could be a way of thinking here. So super interesting looking into that. And it's totally kind of-- as you say, it's almost like it's this kind of curse and-- I don't know. How much time do you spend on this? I'm super curious since you are also part of the sales and operations.

RVB: 00:04:28.549 Yeah. I mean, I think it's the heart of our work as-- well, as salespeople but also-- I mean, I want to call myself more than a salesperson. I want to call myself a problem-solver for clients [crosstalk]--

SW: 00:04:42.122 Good. It was a trap.

RVB: 00:04:45.313 [laughter] But I think in the core of what I need to do as a salesperson for our clients is really to help them understand that this thing is valuable. Right? And I don't know if you've actually ever seen Emil, our CEO, and for example, Philip Rathle, our head of product management-- he talks about this all the time. We call it the graph problem problem.

SW: 00:05:14.643 Oh, classic.

RVB: 00:05:15.819 Right. It's the problem of identifying graph problems. Right? So that's the graph problem problem. And I think it's a valuable thing to think about because, for us, in order for Neo4j to flourish and to be used in so many different environments as possible, we need to find more graph problems, right, because graph problems are inherently going to be solved with graph technology. But I find that the key to solving the graph problem problem is that we need to understand what value is associated with our solution. So I guess the quest for graph value, the first and primary reason why I think we need to be able to do this, is because otherwise, if we don't answer this with a value case, graphs are never going to go anywhere significantly. They're not going to become a major force in data or data modelling in our industry. We need to be able to demonstrate value. If we don't demonstrate value, it's not going to go anywhere. Simple as that. Right?

SW: 00:06:31.664 Yeah. No, totally. Yeah, totally, I see myself in that as well, and I'm going to second what you say. I usually say I'm the worst salesperson ever. However, I'm extremely good at showing value, and then people happen to buy it. Because, I think, for me, that's the only way to do it, double down on the value case and try to figure out and spend time thinking about it because we have this idea of-- especially within data, but I would also argue in human life, this idea or mental model of tables, boxes, and it's so inherent in the way we think. So we already kind of gave up the idea of a [networking?] graph. So we immediately go to the, "Oh, I'm going to put this in this box." It doesn't matter if it's a fraudulent use case behaviour prediction or whatever it is. It's back to the box idea. Right? But there's also this other thing that I know that you spend a lot of time on thinking of this. You know where I'm going, so--

RVB: 00:07:34.959 Maybe, yeah. [laughter]

SW: 00:07:35.714 Maybe. You can just spit it out then.

RVB: 00:07:38.159 Well, I guess, the second reason why I think a quest for graph value is really useful is-- maybe I'll just explain this in my personal terminology here. I'm a salesperson, and my favourite sales book in the world is a book by Marc Miller, which is called Selling is Dead. Right?

SW: 00:08:03.950 Ooh. Selling is Dead, a very good title.

RVB: 00:08:05.891 It's a fairly old book - it's like 20 years old or whatever - but I still believe in it wholeheartedly. And what Marc Miller is basically articulating in that book is the fact that, as a technology salesperson, as a technology problem-solver, you need to understand this fundamental idea that was first articulated by Kahneman and Tversky - they got a Nobel Prize for this - which is prospect theory. And this has nothing to do with prospecting. Right? It has everything to do with how people evaluate their prospects, right, so their future and their decisions that they need to make about their future, in times of uncertainty. Right? It's a fascinating domain. I love this so much because I think it's so applicable and it helps you understand how people think about investing into graph technology from every angle. Right?

RVB: 00:09:12.222 So just briefly trying to articulate this, prospect theory is this idea that when you're faced with uncertainty, when you're trying to make a decision about something that you don't have all the information about - you don't know all the data, right? - what humans tend to do-- and this is not up for debate. Right? This has been proven time and time again. It's just a scientific fact. Humans tend to undervalue the returns, the benefits of a particular decision. Right? So they think actually that the returns of a particular decision are going to be a little bit lower than what they are really going to be, and they overvalue the costs of that particular decision. They think that it's going to be a little bit more expensive than it's actually going to be. So we're actually predictably irrational here. Right? We do this all the time. When we face a decision that is about our future and is based on incomplete, uncertain information, that's what we're going to do. And this is a little bit the fight-or-flee, primal brain type of thing. In order to protect ourselves, in order to be successful as a species, we're going to err on the safe side. Right? We're going to make a decision that is a little bit thinking about benefits a little bit lower and thinking about costs a bit higher. Right?

RVB: 00:10:47.656 But what that means for a technology decision, right-- "Am I going to solve this technology problem with graphs, yes or no?" That's my decision that I need to make, and there's plenty of uncertainty in that decision. Right? There's a lot of things that I don't know about that decision. I'm going to underevaluate the benefits and overevaluate the costs again. Right? So prospect theory is what's going to drive this primal reptile-brain kind of decision that we are going to make. And we're going to obviously make some mistakes. Right? So when I talk about a quest for graph value, one of the main reasons why I want to do that is because I want people to understand that we need to do this differently. We need to think about graph value in a different way. And it's not up for debate. It's just the way humans work. And successful project managers, salespeople, technologists that want to work with graphs better realise this.

SW: 00:12:02.621 Yeah. What comes in mind is that we almost have the graph problem problem, but we have the prospect problem theory theory or something. Right? Because as we talked--

RVB: 00:12:16.112 [laughter] Yeah, exactly.

SW: 00:12:16.393 --about this, I also see that a lot of people having a challenge [adopting?] this mindset because there is a risk. There's an uncertainty in here as well. So as you were speaking, it become very clear to me. The way I usually explain it is with allowance. I have two kids, imaginary kids. One [I was?] going to call Mohammed, and one [I was?] going to call Sarah. In this experiment, I'm going to go to Mohammed, and I'm going to give Mohammed €50. He will be super happy because that's five times more than he usually gets. As you know, I don't have kids, so I don't know how much allowance they have.

RVB: 00:12:55.218 Hypothetical, yeah.

SW: 00:12:55.817 Hypothetical, yeah. So he gets 50. He's super happy about it and go like, "Wow. I'm going to buy whatever candy for it." And then I go to Sarah, my other kid, and to her, I give €100. I give two €50 bills. And she goes like, "Hoo hoo hoo! Mohammed, look, I'm doubling down on you. I got so much more. This is going to be success." But then, since I am such a bad father, I'm just going to go say, "Sorry. I'm just messing with you. Give me one back." And now they both have 50. Mohammed in this case is happy. Sarah is grumpy--

RVB: 00:13:34.976 Frustrated, yeah.

SW: 00:13:35.146 --because she had 100, but now she only have 50. It doesn't matter that it's completely fair, which it is by numbers, but it's just this idea of things. So I think it's a super interesting theory to also apply on yourself, apply on conversations, because in the world, I mean, pretty much everything is uncertain when we think about it. And we try to make sense of it, and this is why we have this reptile or reflex or--

RVB: 00:14:04.206 Yeah. And I think it's fascinating because ever since I read that book, that Marc Miller book, Selling is Dead, it's like I see it everywhere. It's like every single prospect that I talk to, every single community member that I see struggling. I mean, I see it everywhere and not just with graphs. I see it with people buying their first car, or I see it with people having trouble deciding on their next bike or all of that stuff. Every time you need to make a decision where the criteria or the decision-making information is uncertain, that's what people do, and they do it all the time. It's so obvious. And so as we want to get better at articulating value, right, I think we need to think about it. We just need to think about it and be conscious about it. It's an obvious thing to do, and it changes the world when you actually see it and when you actually see it play out. I think it's extremely fascinating, but--

SW: 00:15:16.642 Cool. On the topic of uncertainty that we covered, any kind of tips on how to kind of deal with such a thing?

RVB: 00:15:24.183 That's in the next episode. [laughter]

SW: 00:15:26.899 Oof. Bummer. I was so curious for it.

RVB: 00:15:28.567 Boom. Exactly. [laughter]

SW: 00:15:30.640 A cliffhanger.

RVB: 00:15:31.842 We're going to come back to that, for sure. I mean, the good news is that there are really good ways of dealing with that, and we'll talk a lot more about that. And again, I think, part of the reason for doing this quest, for doing these series is that we want to be specific and give tangible tips really to help people solve that. But I did have another reason why I wanted to go on a quest, if you don't mind me--

SW: 00:16:00.556 Oh, yeah. I want to know. Let me know.

RVB: 00:16:03.538 I mean, yeah, I have a third reason why--

SW: 00:16:06.877 You're not buying a new bike, right? I have to call your wife if that's the case. Okay. Good. Shh, shh, shhh.

RVB: 00:16:13.145 No. [laughter] No. I mean, the third reason why I wanted to do these series and talk about graph value is because-- you know me a little bit, and I think the audience knows me a little bit. I love technology. I'm a little bit of a geek. I got lost in sales, and I love the technical parts of it. I write about it. I do all kinds of stuff with-- I'm tinkering with Cypher when I have a free moment. It's really something that I enjoy doing. And I see around me that there's a lot of technical people that love graphs. Right? And it's actually intentional on our behalf, on Neo4j's behalf. We actually believe that graphs are super, super interesting from a technical point of view and that the technical value will translate into business value in reality. Right? But I think therein lies a little bit of a challenge. The fact that we have technical people, whether you're a programmer or a mathematician or some kind of an analytical biologist or whatever it is that your background is, but you're kind of technical, right, and you're flocking to graphs. You love graphs. Right? I see this all the time. There's these Neo4j fans in the community, and they're usually quite technical. Right? And I love technical people. I mean, I'm probably one of them myself. I got lost on--

SW: 00:18:01.747 Yeah, totally. You're hopeless. [laughter] That's why [I am up here?] as well, so.

RVB: 00:18:07.190 Yeah, exactly. That's true. But that's usually where the conversation about graphs starts. Right? And there's a little bit of a challenge in there because if I know technical people in any way, I also find that they're usually not the best communicators. Right? And I don't mean this in a nasty way. I think it's almost like a compliment, right, because they have a very particular mindset. Just like some commercial vacuum cleaner salespeople will have a particular mindset, we're all a little bit different. But the fundamental reason why I think this quest for graph value needs to address this point is that if it's only the technical people that are going to look at this and find value in graphs - and they are not able to articulate it in a way that resonates with less technical people - then, again, we have a little bit of an issue. Right? We're going to get limited. We're just going to get limited in what we can bring to the table, what we can deliver to the world. So I feel that with this series I want to address that third point as well to try and help technical practitioners, data scientists, developers, salespeople that got lost in technology, whoever they are, but I want to try and help them find a structured way to articulate graph value. Right? And that's a really important piece. And I don't want that to be a nasty conversation, but I think it's something that we should just recognise. I mean, if we keep assuming that practitioners and technical people are going to be fantastic at articulating graph value, I think we're pipe-dreaming. It doesn't work that way.

SW: 00:20:10.344 Yeah. No, totally. And looking back to my career, what most people don't know is I'm a super introvert. Back in the days, I hated talking to people. I just want to be left alone in my little lab where I can build my things. It doesn't matter if it's a music instrument or building a film camera or whatever it is. I just want to fiddle around with my stuff and build cool things and don't be bothered. Right? I don't want to do presentations. Most of the time, I don't want to attend even to the meeting, if I look back to my younger self. But during that journey, I kind of always worked with someone that was very good to pitch it. When you have that person - I mean, it can be Steve Jobs and Woz or whatever kind of combination you want to go with - when you are aligned and synched, magical things happen. Right? But also, in that, I learned that I cannot be dependent on other people because if I want to build things that are cool, I need to start to articulate that.

SW: 00:21:15.717 So from my point, I'm just going to share all the stuff that I learned and what helped me because, at the end of the day, the way I kind of look at my job nowadays it's very much like curling. So the practitioner is the one shooting away the-- I don't know what it's called in English, but shooting away the object. And I'm the one with the broom there, paving the way and sharing the knowledge that we have. And I think your background is super useful for this. So for me, and not even in the ballpark of becoming [inaudible] two guys sharing the love for graphs, all the stuff that helped us on this journey. So basically just bridging the [gaps?], so.

RVB: 00:22:04.032 Yep. And I think we've come up with a little bit of a game plan to do this. Right? So there's at least four, maybe five more episodes that we're going to be publishing. So what we're going to do next episode is try to sketch out, how can you find graph value? How can you really map it out and identify it? Then we're going to spend some time on building the case for graph value, just how do you articulate it? Right? So it's all fair and well that you have found the-- or that you've solved the graph problem problem and you've found the graph value, but how do you build the case for it? Right? Then, of course, we need to articulate the case. We need to present it in a meaningful way. And then, I guess, the last-- or we're going to try and bring it back together by articulating how we build it. Right? How do we achieve it, the graph value? That's the plan.

SW: 00:23:07.247 Sounds like a plan. Also, do we have any conclusions for this?

RVB: 00:23:14.340 [laughter] I guess it's a little bit of an experiment, but I'm super excited about it. We're building a lot of content that I'm hoping will be really useful for not just commercial organisations but also community people that are doing graphs for good or wherever they are and doing their own projects. But I'm hoping it will be an interesting journey. I'm looking forward to it.

SW: 00:23:41.678 Yeah, me too. I think we could say that we have a hypothesis that we need to kind of be better and create the more structured approach to graph value. And now it's time for us to justify or falsify that. So let's see how that goes. It has been a wonderful time chatting to you, Rik, as always. I think we, for once, kind of kept it calm and structured. So we're going to tick--

RVB: 00:24:08.618 What's up with that? [laughter]

SW: 00:24:10.119 We're going to tick one of those boxes. People might not know who we are now, but-- super good. Thanks a lot.

RVB: 00:24:18.421 Thank you, and I'll speak to you on the next episode. Thanks, everyone. Have a nice day.