 Hi everyone. Thanks for attending my webinar on using data storytelling to build products. I wanna thank you for attending this webinar. I wanna thank Product School for having me. A little bit about myself is that I've been in the tech industry for a while. I work at Facebook Spotify Pinterest. Fun fact is that at the end of last year I found out that I listened way too much Justin Bieber. I'm in the top half percent of listeners and I'm sharing this because my friends found it funny but also we're gonna be touching on Spotify later in this presentation. So what are we going over today? I'll be talking about what is data storytelling. I'll go over why it's important. I'll give you an example of how to create a data-powered story. Once that story is created we wanna know how to make it effective at the company settings so that you can drive impact with it. I'll talk about useful resources to develop data skills so you can create data-powered stories. And I also wanna share a presentation checklist that's very useful. And then I wanna summarize at the end and thank you again for attending this presentation. Cool, so what is data storytelling? I got this definition off of the Harvard Business School. I really recommend reading this article that covers what is data, what is a narrative but it also covers what is storytelling. And storytelling is really important because it's one of the most useful ways to communicate a point to an audience. So why are we here? I wanna cover a practical way to tell a data story and really in the tech industry if you can get insights from data, a lot of people view that as a superpower. So why do PMs need to know the superpower? And I wanna just point out that PMs are all about understanding a people problem, getting a group together and then getting that group to build a solution to solve that people problem. So to really emphasize this, building product is about solving problems for people. With data storytelling skills you have the power to understand a problem and get a group of people together, help solve it together. So I think of it like in the tech industry or in your tech company, everyone's super talented. There's just such an enormous amount of impressive people. But when you assemble together, like the vendors, you're able to take on much bigger high impact problems. So here's my framework for how to create a data powered story. You really wanna just answer three questions. What people problem are we trying to solve? How do we know if it's a real problem? How would the world be different if we solve this problem? The first question is really about setting the problem up. And then to make it more of a narrative, more of a story that's relatable, I like to share my personal experience, my friends experience, someone that I know. And or you can just, if it's a live session, you can pull in the audience and ask them if they've had this problem before or have them share a relevant situation. And then that goes into, how do you know if this problem is real? And that really comes down to data. So you can use qualitative data or quantitative data. The format I like to use once you have the data is that you show the data, you reason about it, you wanna explain it to the audience, make some assumptions potentially, then you make a recommendation on how to improve it. You wanna reiterate your points and then you make a full recommendation. In that full recommendation, you wanna make sure you're answering question three. How would the world be different if we solve this problem? And that really comes down to defining what the success criteria is. Because what happens in the tech industry is a lot of times people build things. After the fact, they look at the data and they look at what moved up and then they say, oh, we actually had impact because this thing moved. But in reality, if you really understand the user problem, you can actually define success well before building anything. And that comes down to, if someone gives you the result, you should be able to know if, oh, I actually helped solve this problem or it's a learning opportunity because you did not solve that problem. So let's go move on to the example. So first thing is you wanna answer what people problem are we trying to solve? And the problem I chose is, people like to discover and share music. However, it's both hard to listen to music that was shared and to share music to people. And my story is, so my name's John, I really like listening to music, sharing music and discovering music. I often use Spotify to share to Instagram, Messenger and Facebook. But I also discover music through Instagram, Messenger and Facebook. And overall, it's not that great at certain times because there's lots of steps to share music. When a friend tells me that they heard a good song or they found a really good playlist, I ask them to share it to me, but sometimes they don't even know how to share it. So I have to give them instructions, go to the option menu, look for the share button. And that's not really great. It also tells me that not too many people know about the share feature. And then on top of that, when I discover music, a lot of times when I tap on content and that deep links me over to Spotify, I get an error message. And it feels like half the time, but I think the real number is probably a quarter of a time or something around there. And it's just not a great experience, right? Because I don't get a list of my music, I end up giving up and I'm just frustrated. So here's the problem, right? People really love to discover and share music, including me. However, it's both hard to listen to music that was shared, but it's also hard to share music. And then you go into, well, how do you know if this problem is just not huge on? Like, is it actually impacting other people? Well, you can take a look at the data. And when you take a look at the data and you're presenting this to the audience, you really want to guide them through this story. So first of all, like step one, I want to tell people that it's important. So you want to answer a question, is sharing a highly used feature? And I want to point out that Spotify has 400 million MAU, 25 million users share music every month, 100 million users come to Spotify every month from shared links. And then I want to talk about, well, let's reverse it to make sure people understand it's important, really reinforce that. What if we remove the share feature? What happens to these key company metrics? And I can point out that our estimates show that registrations will go down. Retention would go down. Active users would go down. Consumption would go down. And then, well, now that you showed them some numbers on metrics going down, you want to explain to the audience, well, why would they go down? So let's take the case of registrations. I share music to my friends. Some of my friends potentially don't have Spotify, but they want to listen to that song, right? Because I shared the link. And then they might click on that link, download Spotify, register, and then consume this music or playlist. And that creates a new registration. So how about the case of retention, MAU and consumption? Well, a lot of times I share out this music or someone shares it to me. I press on it, then I go over to Spotify and I consume the music. So overall, I have more consumption. But how about my friend who potentially for that day wasn't going to use Spotify? But they ended up clicking on the link I shared to them. They go over Spotify, listen to something, and then automatically they become an active user. And overall, a lot of these people who weren't going to go on Spotify for a day, that really involves higher active user numbers, higher retention, if you have a sharing funnel, right? Because people are like, oh, I want to listen to this. My friend shared it to me. My favorite artist just shared something on Twitter or Facebook. And then here's an example of a slide I would create. Overall, I highly recommend making very simple slides, if you can. And you want to just put in a point where someone can just have a very simple takeaway. So on this slide, I put in, well, this happens 100 million times every month, but 22% of the time, when people come over Spotify from a shared link, nothing happens, right? You don't get to listen to music. It's pretty poor experience and these are numbers for it. And that's a lot of upset people. So we move on to making a recommendation, right? And I point out, well, this is happening 22% of the time. I think we should fix this because it'll overall improve our metrics, but also it'll improve the user experience. And then we can move on to the other side of the sharing experience. So it takes four to six steps to share something to your friends or share something to your fans or to your artist. And then you can look at the data. What are some proxies for understanding if this is a problem for friction because it's high friction or education problem. So maybe you can take a look at comparing users with big phones and small phones. People users who have bigger phones potentially have the, well, they do have the share button on the option menu visible. People with the smaller phones, they don't have the share button visible in the option menu. And you compare the two groups and you see, well, it looks like people who have a visible share much more than people who don't have a visible. So potentially it's that people just don't see it when they hit the option menus. So they just don't share it because they don't know that it exists. So what can you do to fix this or get learning from it? You can make it more readily available on a popular surface, such as the mouth playing view and AB test it. Or you could even just make sure that on the option menu, the share feature option is actually visible across users. And then you want to close it out. You want to summarize your takeaways. You want to summarize your recommendations. And you want to make sure you're defining your success criteria. And that really comes down to question three. So how would the world be different if we solve this problem? So I think that people will be consuming more music because people want to actually, people are pressing these links to go listen to music on Spotify, but 22% of the time they're not getting anything, right? So team can make a goal such as this, the goal could be this happens less than 5% of time for users. Then you can also track metrics such as consumption, active users. And then how about the other side of that? We're about sharing music. So we want people to share more music because we're assuming that people like to share more music. So you can track, you can make a goal on number of users repeatingly, repetitively sharing music over a certain time span. You can track that there's more users sharing. And you can also take a look at higher restoration and consumptions. Let's go to the next slide. Cool. And I just want to make sure I reiterate the framework here. So to create a data-powered story, you want to answer three questions. You want to say, you want to answer what people problem are we trying to solve? How do we know if it's a real problem? And you wanted to find a success criteria, which is, how would the world look different if we solve these problems? Well, this problem. So, and then once you have the data story done, what are the things you want to do to drive impact with it? Well, first up, you want to make sure that your slides are very readable without you presenting them. Because what happens is in a lot of meetings where I present something, people find it really interesting so they share it across the company. And you want to make sure that, well, if you're not there and narrate it, they're taking the same points off of those slides and you're getting, they're recognizing the problem, they're getting the same recommendations, how to fix it. And then once your slides are really good, you want to start sharing it to a smaller group of trusted people to get feedback. So I like to make sure I get like questions from them, stress test the idea, really clean up the slide so that it's prepared for a wider audience. And to see if they're really aligned on the problem when excited about it, when they go through it with me. So once it's ready, you want to share it to a wider audience. And I usually like to set up meetings with stakeholders or decision makers that I think are relevant. But I also like to share it on Slack, share it to teams through email and really peripherate the idea that this is a problem that we should fix. And once there's alignment across these, the relevant parties or stakeholders, then you're ready to really build a solution to solve that problem and start doing real product work. Cool. And then I just want to share how to really develop these data skills so that you can create data power of stories. You can learn SQL. I highly recommend learning SQL. So you can go on Udemy Coursera SQL School for it. SQL Zoo. Learn data visualizations because visuals are very good communication meetings for people. I tend to use bar charts and line charts mostly but Coursera has a lot of good courses for it. And third, I really recommend reading a lot of good articles. So these are three articles that I really like making good decisions as a product manager by Brendan too. Julie, who's a former VP at meta slash Facebook writes a lot of good product content. And these are my two favorites, metrics versus experiences and agony and ecstasy of building with data. And they're talk about how to define metrics, talks about how to use data to build good products. And here's the presentation checklist. Everything goes over your head. Go through this checklist. So the deck should tell a story. Each slide should be very simple, one point if possible. And then when you create your charts, you shouldn't assume that people actually know how to read charts. So you want to call it out for people. So in summary, if you create a really good data story or you can assemble a huge group of people or even a small group of people to tackle a really high impact problem. My example is a lot of times I create these presentations I share across the company and over several months people recognize that as a problem. And sometimes I could have up to five engineering teams working on a really high impact, very challenging problem for the company. So to reiterate what I said earlier, building products is about solving problems for people. With data storytelling skills, you have the power to understand a problem and get a group of people to help solve it. And that really comes down to answering three questions. What problem are we trying to solve for people? How do we know if it's a real problem? How would the world be different if we solve this problem? And then if you have any questions, feel free to reach out to me on LinkedIn. And I also want to just give credit to a few people real quick. So the presentation checklist is from Alex who's currently a data scientist in Meta. These questions are actually from Julie who's the former VP at Meta and she uses them in this talk called How a Facebook Designer Thinks. I want to call up Saurabh who's a product manager that I worked with and he's currently a product lead at YouTube. He's one of the best product managers in the world is my belief and he's really great. So I just want to thank these people for helping me with this presentation or inspire me in what content put into this presentation. And then I'll end with thank you for attending this webinar again. I'm hoping that you got a lot out of it and I want to thank Prague School for having me. Cool, good luck.