Hi everyone, my name is Patrick Akio and today I'm joined by Samuel Bake, Chief Product Officer over at V dot IO. If you've ever wanted to do something with AI and video editing, you have definitely heard about V because they are the biggest browser based AI powered video editing platform right now out in the market. They ship Gen. AI features. They deliver value to users like no other and I wondered exactly how they did that. Luckily, we touch on all of that and much, much more in this
conversation, so enjoy. I used to think that, you know, strategy was really, really important and I needed to think everything true and then everybody should work on the things that I think or things that feel fall within my my plans. Master planner. Yeah, yeah. And I, I think that I thought that because this is what people talk about a lot when they talk about successful business, successful products.
But then when I started to look at, you know, what is the reality, what gets successful within my organization, I realized that very often I'm not right. And that's a very humbling experience. And now I'm like, OK, actually, maybe I do not know. And maybe we should do things where people can come up with suggestions and actually ship those so that we can actually figure out what works and what doesn't. That being said, there are a few constraints that I said.
So for example, you know, we've had people make video games within Veed as a hackathon project. Now that is really fun. It's so cool. It's very cool. Yeah. It is not something that is going to fly with our audience. So we often tell people like, hey, this is our audience. And these are a couple of themes. So for example, the last one we did with the entire company, we said make a video better with AI.
That's that's the theme and and make it and, and it can't be like a proof of concept, it needs to actually work. Yeah, and that was a great theme. We got very different things and some of those things actually made it into production as well. That's super cool. Yeah. I mean, even the that's me, right? Because I also love video games. The concept where people just made a video game in, in Veed, I
would love that. And even that if, if I were responsible, I would still try and figure out maybe, maybe that could be like an April Fool's thing and then it goes viral. Like it's still great. It's a great idea. I, I think it's, it's great. A great. I mean, we talk about room to play. This is the this is room to play.
Yeah, yeah, for sure. Yeah. Talk to me about how you've been putting then Jai solutions in production because I feel like I've talked to many product people, even people in startups and there is a different challenge with doing that and making sure that you have kind of the same production level readiness with Gen. AI that you would have otherwise with software. Yeah, I think, I think the, the real, the short answer she won't
like. And I think actually you shouldn't try to get there because what I realized with Gen. AI is that people they have different expectations. They do not expect things to always be right. Like if you have ever used Cha Chi PT, you know that it's really cool and magical, but it's not always right. And you know that every couple of months they're going to send out a new model update. It's going to get better and it's it's going to be right a
few more times. And that is kind of the customer expectation around Chen AI. So when we build Gen. AI tools, we're not trying to obviously, well, well, we're actually trying to make it perfect, but we are also OK with it not being perfect. So that's one thing. And then the other thing is like, OK, but if it's not right, it should not break the entire experience. So there should be ways to iterate on it, tell our product that hey, this is not good, re prompt things, things like that.
So that's how we are trying to, you know, take care of that. And I think actually the the next generation workflow around video or AI generated content is going to be AI generates 1234 options for you. You tell it what you like, what you don't like, and it iterates really quickly and that's how you get to something you love. Options giving options is always
amazing. Options also come at a cost, but in the end having the option from a user consumer perspective, I love that you talked about consumer expectations. Is that something that you did market research on or how do you get to kind of that consumer expectation and that awareness of that? Yeah, Honestly, really hard to do market research on that. Very expensive, takes a long time. I think this is something where we just kind of felt the market.
We just noticed using other very successful products or things that are kind of adjacent to video creation. So think about like image creation and those types of things. We noticed that this was just kind of the the way people treated that. So that's that's how we got to that. And so far, we haven't been proven wrong. Gotcha. I, I was interviewing and I'll, I'll share this because I think it's valuable.
I was interviewing at YouTube, probably my dream job, right, Because I do podcasting, I interact with the platform. So product manager at YouTube, I biffed the interview. I'm still this salty on that, but we'll put that separately. And I asked the question, OK, how much access do you have for your end users, right? Because me as a creator, I would love to be involved in the creating process of analytics, for example, or making YouTube better.
And they were like, well, we have a, we have a team. And then everything goes to the team and it has to be approved. And sometimes you get like a survey about actually talking to people in an organization like YouTube and Google is a big, big bottleneck. Like you have to go through multiple hurdles. How's that for you? How easy do you have access to your users? So I think this is the reason why companies can beat Google or
YouTube or whatever at things. Sometimes people are like, you don't want to compete with these large companies, but I think you this is where you have a competitive edge. We have tons of access to users and I think it's really important. So when I started at Veet, we didn't have much data and Saba, the CEO, was like, well, if you don't have data, just talk to a lot of users and if you hear something very often, it's probably the thing we should be working on. I thought they was genius.
It's super simple. Yeah, it's super simple, but sometimes the smart things are very simple. So we started speaking to users a lot and now everybody in product, so every product manager, every designer needs to speak to at least one user a week. We have a user research team speaking even more, but everybody needs to speak quite regularly. And I think that just creates this muscle so that you can kind of channel having a user at the table when you're making
important decisions. So there's one thing we do. Another thing we do, which has been really fun is we have this V feedback program in which we have a couple of power users in our Slack. We can share ideas with them, we can share prototypes with them, give them access to things and get feedback really, really quickly. Great way to kind of get those creators to kind of bond with us. And lastly and this is I highly recommend every company trying this.
We stole this idea from Stripe I believe, but we invite customers to our all hands. So when we have the entire company in a meeting, we interview one or two customers and this, and everybody in a company can submit questions and this really makes those users come alive. We now have an idea of what our users look like, what they value, and yeah, that just makes for better product decisions. That is, I've never heard that that you invite your customers,
your users to your all hands. Like if if you talk about risk and you're talking about control that can go anywhere, I feel like. What are some of the insights you've gained through doing that? Yeah, I think, I think so far nothing has has gone wrong. But what was really fun is to just ask people like, OK, So what is the most time consuming making a video? What are some of the pain points?
And very often we hear about pain points that are not necessarily pain points in our software, but are kind of adjacent to our software. And we feel like, oh, we can actually maybe help users with this. So for example, people often complain about forgetting their lines. So we build a teleprompter. Then people were reading from the teleprompter and they were not looking into the camera anymore, and they were confused
about that. So then we build a feature that uses AI to make sure that your eyes always look at the camera. This kind of iteration you can only get from talking to users. There's no metric that will show you these problems. You really need to speak to users and they need to kind of talk about their they, they will not say like, hey, I'm looking in the camera all the time. Often they're, they're saying other things. They're like, Oh my, my recordings fail. And we're like, why?
Well, I'm not concentrated. I forget my lines. And you ask why, why, why, why, why, And at some point you get to this inside and build a feature for it. Yeah, I feel like the the skill of figuring out what the user problems are for me, that's incredibly valuable, right? And I love the example that you gave. If there would be a a prompter and I would have to look at it and then I see myself not looking at the camera, then it's
useless, right. But then is that I'm not going to be like, well, that's like you have to ask me why I'm not using that before I say, well, that's the issue. I would never tell you. Well, out of all the hundreds of features you have, I don't use this one because of XY and Z. Then I'm the expert. Exactly. Exactly. So yeah, user interviewing is a, is a skill and it's really hard, but if you do it well, you can get fewer goals out of it. Yeah.
I mean one of the most fun teams I've been in was my last one was at a bank where I feel like the engineering team really understood the domain. Me as a product person. I also took them to stakeholder calls, alignment things. I made sure that the domain awareness was there. It was in the sustainability field and people that get attracted to that also really love that domain.
So there's there this inherent curiosity, but I feel like the the kind of signals you put into place, right, having product people do use interviews once a week or talk to users once a week. Having used a research team that does more of that, inviting users to your all hands. And all of engineering is also there. And they see that people become more and more aware of the domain, and the better they understand it, the better they
can execute, I feel like. Yeah, 100% because like I believe good product is like a ton of little tiny decisions. Some decisions get made by a product manager, some decisions get made by a designer, some get made by someone in support and engineering. And all those decisions matter. And I think if you get create that strong awareness about who your customer is, what they care about, more people will make the
right decisions. Yeah. Talk to me about trust, because from a product sense, even me moving from software engineering to product, I was like, I can see those micro decisions. And then because I was good at that, I would sometimes still engage and I'd have to be like, actually my team told me that. It's like, you're not really software engineering anymore. It's like, yeah, like I had to be like, yeah, this is your guy's decision. It was hard for me to let go. And now I'm aware of that.
So I kind of fell and I, I got a mirror in front of me and now I learned. But from your perspective, like you've gone, I think high over more and more. How hard is it letting the decisions just flow and putting trust in the team? Well, I got to be honest with you, it's very hard, like, because one thing I do not want to do is tell people exactly what to do or what I would do. No, that's just not a nice way to work.
There are details that I think are so important that I might intervene, but those are I, I tend to keep at the minimum. So I, I, I was struggling with this and then I realized, you know, what I should work on is create the context, the frameworks, the things so that everybody would make the right decision according to my beliefs. So that's what I'm working on right now.
Like I'm, I'm making sure that people understand what we value as a company, what metrics we want to drive, what our users care about, all those types of things so that when I'm not in the room, they will make the right decision. And to be honest, whenever I feel like a team went completely off, off guard of the rails, I, I, I kind of asked myself what could have, I could have done differently in order to, to have
them make the right decision. So sometimes there's these teams that, you know, they, they prioritize the feature. And I'm like, why would we work on this? And then I asked the team, why would we work on this? And I say, well, because this is important, this is important, this is important. And I'm like, oh, wait a minute, I have not told you that this is not important. Or, you know, I've, I've not repeated this idea. I have often enough. So those things really help.
Really focusing on the context, that's one thing. And then in order to make this a good working system, one of the things that we've introduced at Veet, which in the beginning was not received very well, was let's review. I, I sit with every product team every week. And in the beginning people were like, oh, that's micromanaging. You're 2 on top of things. That's not nice. But now those meetings are 10 minutes. We just check in like, how's it going?
Do you have any questions? Like what are decisions that you're making that you want to have input on? It gets, it goes really quick. Teams spend very little time on reporting. And I have a sense of where things are going. I feel chill and I can, if there is a problem, I can at least voice my opinion in time. Yeah. And I think checking in with people is like super, super important, super important. How do you do demos with regards to new features in the
organization or even outside? I mean, I see, I see what you do outside. It's on LinkedIn. I love it when you just say, OK, this is something cool and I see you as a user using that. That's amazing. But how do you go about demos? Yeah, demos are really part of it. Like I think if you look at building and one and when, when, what are the most rewarding moments in building? It's the moment you can show others what you're working on. So we love the demos and we, I,
I think we have a demo culture. Basically, when something is shipped, whoever shipped it needs to make a demo in Veet and I need to make a video because we're a video company, right? Great way to also enforce people to work with, with our tool. And we made, made this a competition. So basically people need to make a fun video about the thing that they've shipped, like a minute long. And this means we have this select channel that is filled with demos. And then every month we pick a
demo of the month. And then that person the next month picks that demo of the month. And it's fun because people get creative. Everybody does it a little bit differently and it's just a really fun way to experience our product and celebrate what we've built. And then on top of that, I love kind of yapping about our product on LinkedIn and stuff like that. I also think it's really important for me as a, as a leader of a company that helps people create videos, to create videos myself.
So I anything that gets shipped, I make a video about and that we ship a lot. So I make like one or two videos a week now about the things that we're building. Yeah, I mean, I I love it. For me, like as a person that creates videos, like podcast is easy because I have someone in the back, Jamie that does the, it does the camera switching and everything. But I used to edit and for me, editing and listening to myself speak on camera, that's how you
got to be a better communicator. Also feedback from other people, they would be like, hey, you always do this thing where you say XY and Z and I'd be like, I don't like that. No, not anymore. So I would immediately consciously fix that. And also with editing, I would hit listen to myself hearsay, umm or not using a pause and I would try and fix that actively. I became a better communicator by editing my own content basically.
And I feel like I became a power user of whatever video editing software I was trying to use because I edit video needs to go out weekly. I try and optimize that process out of the get go. I'm also software engineer. So give me all the hotkeys, give me all the advantages from a tool perspective. And if you have your whole organization doing that, I'm, I'm assuming engineering does it as well. They become better communicators, they become better users.
They know user pain points because they are the user has so many advantages it's ridiculous. Yeah, it's, it's, it's really a no brainer and it's sad that we only kind of started doing this two years ago. I feel like as a as a company we would have been even further had we had we introduced that earlier. Yeah, for sure. I mean, you've been in different organizations and I feel like Veed is B to C, right? Your consumer is the user and that can go super scale basically.
I don't know how many users you have now, but how is it building something that can reach this mega scale? Yeah, I think to be honest, I we're B to C ish. But I think our consumer are not people that are using it for fun. Like the people we have a lot of people using it for fun, but they're not paying. The people that are paying are using V to do something for their companies or, or, but those can be really small businesses that can be creators, things like that.
So building for that scale is, is fun. It's really fun because you need to the, the, there is quite a widespread of things that you can do, which is hard with prioritization, but it also is kind of energizing and, and you let you see your product being used in very different ways. So for example, whenever I introduce people to the product, I always show them a couple of pictures of users that I've
interviewed. And I've designed these as like Pokémon cards because I'm a little bit of a nerd. And on those Pokémon cards, like one of them is a college teacher, the other is a marketeer, the other one works in learning and development, and the other one is a startup founder. And it just shows you how wide the variety of users are using our product. What unites them is some of the jobs that they want to get done
with our product. But it's been really fun and interesting to see the kind of videos people are producing. Yeah. I can imagine. I mean, for me it'd be really fun because video and concert creation has become this passion. I don't know if it was a passion of yours kind of getting into this, but where your personal passion aligns with what you do on a day today, that's magic. Like that's what I would wish on everyone. That's the best feeling.
And I feel like with people that are actually, it's expected of them to create content and to use videos, they will either realize that this is not really a passion of them. So then that expectation doesn't really fit with them, doesn't really glue well, or it's what they love. So it's like amazing. Yeah, Yeah. For me, it's definitely become a passion. And I also feel like when I speak to users like you have this, you can say like, I feel you.
I feel you when you need to take that take for the 10 time. Like you, you've recorded everything and you realized your camera was out of focus. I feel you. And that's like such a nice form of user empathy that is just only there when you really feel, feel it. Yeah. Yeah. I mean, I, I could put myself into the kind of ears of the listener and think, OK, it's really lucky to have this platform and access to users and, and kind of this tangible software, right? Video editing.
It's genius how you've done it in companies. But when I think of my own software or when I think of my own product, I don't have that. Like what can I still implement? If you're in an internal organization, you have a product in internally, they're the customer. It's not customer facing or it's not B to B. What are still things that you can take away that you can implement in your own product? What advice would you give?
Yeah, I think, I mean, you can still talk to users or try to figure out what they're, what they, what they care about. Like I think in a way, sometimes it's nicer. And I'm kind of contradicting what I said before, but it's nice if you're not the user because if you are the user, your opinion becomes a little bit more valuable than some of the things that you're hearing when you are not a user.
You need to do more kind of academic style research and really get into the depth of what the people want. And sometimes there are, you need to get creative here. Like some people do not have access to users of their product. So then you need to be like, OK, but where do those users of my product, where do they spend their time? What do they listen to? What do they watch? Are they on Reddit?
Are they listening to podcasts? And you can still kind of channel what what motivates them through those ways. So I think there's many different ways to kind of learn from your users, even if you do not have direct access. 1 cheeky 1 is look at your competitors. So I think competitors kind of share a lot of data publicly about their products.
Yeah, unintentionally. So if you want to know what features make money for any product, you go to their pricing page and you look at the top three bullet points. Those are the most driving for for revenue, right? So like, there's a lot of ways that you can get creative to still get that access to people even if you can't speak to them directly. Yeah, yeah, I love that. I mean, I've never thought of it, but indeed not being a user, it gives you a different viewpoint.
How do you think of how product management has evolved? You me touched on kind of technical product managers versus domain experts from more traditional business analysts, people that don't really know about the software. And for me, I only have the background I have, right. I could really leverage software engineering, take stakeholders with me and talk about, OK, this is what good engineering looks like and we're not there yet, so we need to focus on XY and ZI
really relied upon that. And I could not see me being effective as product managers, as product manager without that knowledge yet. There's a lot of people that are effective without that technical knowledge. So there is a way. What's your view on kind of product managers and the technical expertise? Yeah, I, I, for, I've, I've had many different views and, and I think what I, what I'm afraid of is like dogmas.
So like I know at Google they only hire technical product managers and I think that's great for Google maybe. But I also see Google really struggling, for example, to market their products. And I think that is a result of this technical focus. So I believe that any product manager being into your team, you need to think about, OK, do they have an edge? That is different than like any product manager. So that can be a technical
background. But for example, at Vet, one of our most successful product managers, she has a background in creativity. She used to be a photographer, then worked on a few creative apps and then became PM at VET. And I think that's another example of somebody that has an edge that is very different from a technical background, but can really help kind of change the conversation and and bring in unique insights that we don't have in in the team.
So I think when thinking about bringing on product managers, I think that's one of the things that that that we index on. Yeah, got you. How has for you product management changed with regards to AI? We touched on AI, putting it and embedding it in your software, the thing that is your bread and butter that you deliver to customers. But there's another angle with regards to productivity. I've talked to many engineers, there's many opinions on using AI code tools.
People are definitely saying it's here to stay. So engineering is evolving. How's product evolving with regards to personal? Productivity yeah so one thing I noticed when AI kind of chechi PT boom started was that I was getting sent a lot more documents I don't we have a saying in V too long didn't read TLDR or TLDR culture. I don't care about long documents, stop sending them. So I thought that was a really bad thing. People were like, oh, it's easier to write, so I'm going to
write more. That's like the opposite of what you should do. Now what I see is I see AI helping us in two ways. One is as kind of like a sparring buddy for PMSPM can be quite a lonely job. And everybody that you often spar with, they they want to get something out of it. So if you speak to engineers, they maybe want to get more engineering time on the project that they really value. And if you speak to design, they care more about little details that they want to get into the
product. So engineering can be a bit lonely. And I think actually ChatGPT or Clod or whatever LLM you use can really help you kind of get your ideas across and make develop your ideas even more. So I see some of my best PMS really using LLMS for that. And I think that is great. And, and, and the quality of ideas kind of increases. So that's that's one thing I
really like. And then the other thing I see is that one thing anyone in a company should think about is like, how can I make my idea more tangible? How can I make it more visual or more, yeah, more interesting? How can I give the idea legs? And I think now with all the tooling, PMS can really improve that. So they can make prototypes, they can make mock ups way easier than before. And that's something I'm seeing
a lot. So for example, we build an AI feature and then some of the people on the team, the PM and the designer build a tool to test out different prompts for that AI feature. And that is amazing, right? We're saving engineering time. We're, we're, we're having the PM and, and the signer think about prompt engineering.
Super, super cool. So I think that is giving me a kind of a, a tiny view into what the future is like, because I do believe that in the future, anyone in the team is going to contribute to the code in a different way. And I think I'm very curious to see where that is going. Yeah, yeah. I'm very curious how organizations are going to be set up like that to be able to do so, right. Right now it's it's easy to create and from a product end, sometimes you need user
research. You have many hats on, to be honest, depending on the accessibility of people and if people have like a discipline, you can do anything basically. And if you don't have those people, it's up to you to kind of fill that gap because it's still valuable. Having a sparring partner or something, what you can also call kind of an expert to bounce ideas off of is incredibly valuable. But I've also noticed I become lazy because critical thinking kind of goes down the drain.
I accept what I see there. So there needs to be a level of discipline, I feel like, and it's still a responsibility. Yeah, I mean, I think I, I read this study came out a while back, I forgot what the name was, but basically people that use LLMS to write essays, they become worse at writing essays even if they stop using the LLMS. So definitely there are some concerning things around this. I think we're going to sort that out. But yeah, I, I can definitely see people get more lazy.
And I think then for me as a leader, I just should be razor sharp there and be like, OK, if not done your homework, do it again. And we'll we'll see. We'll see you next week. Yeah. I mean, that's the that's the responsibility, right? You have the nice part and the the thing we have to actually put your foot down. Yeah, I get that.
We talked about putting Gen. AI solutions into production and I really wonder how you decide in the end what is going to be valuable or what is more of hype and what is not necessarily relevant for the product, but still looks interesting. Keep an eye on how do you categorize kind of what you see from features that are all of a sudden available. Yeah, this is really hard. Prioritization in this age of AI
is super, super hard. And one of the things that is very interesting about AI is that we're moving from toy to tool. Like there's a lot of fun. Like, oh, you can make this weird meme now and like the anime memes a while back, but those really drive usage. And then slowly but surely people are like, OK, I can use it for this, but then I can also use it for this. That is very interesting for my business.
So that means that as a company we need to be focused a little bit more on fun and toys and making those things. We realized that one thing that would is really important with AI is speed of shipping. So people want to buy tools from products that are moving ahead. And it's almost like making hit music. You just got to make a lot of hits. So we're really prioritizing things that are easy to build, like things that we can build in a week or two weeks. That's one way we prioritize.
And then the other way is can we be the first to market? Can we build something that is unique, whether that's bringing a famous AI model like Google VO to market for the first time or bringing something unique that is very that we only can do on our platform. And, and those two things are, are kind of how we prioritize right now. And to be honest, we find it very hard to say. This is just like with music, I guess this is going to be a hit. This is not going to be a hit. We rarely know.
But then, you know, if you ship 10 things, if you ship fast, one or two of them are going to be hit. Yeah, I really like that approach of shipping fast and then delivering. And proof is in the pudding, right? You will see once you deliver. Otherwise, we're going to have endless discussions on yes or no, and if it's a yes by default and then we touch and we figure it out on the go or we steer or we actually put in the trash because it turned out to be trash.
Well, whatever, we move on to the next one. The ones where you mentioned, OK, if we can do something that has never done before, I feel like that has kind of a higher risk, but also higher return where it's a bigger bet, right? You don't know if it's going to be valuable. It's still still in the same mindset, but there's also probably more effort into that because it's not going to take one or two weeks likely. Yeah, yeah.
That's correct. And there you have to have a bit more insights and a bit more more belief and confidence. So we're trying to to find that confidence somewhere and then just build it. But yeah, I think what we're seeing is that novelty, newness is so, so important right now. So we're trying to figure out, OK, right now you can make a video that looks very realistic, but the person cannot say
anything. Can we build something where they can say something stuff like this or we've seen this trend in like most videos are basically moving images, right? So we can look into the future a little bit by looking at what text to image is doing in certain startups and then being OK, we make a video version of that. Yeah, I like that a lot. I mean, the the pitfall that I see and some organizations do this well and some organizations
just struggle with this. If speed is like the mantra then and someone said this and I love it. It's like even marathon runners, they don't go Sprint after Sprint. Yet in this framework that we have, we only Sprint and then especially if speed is also the name of the game with the rest of delivery to production. And if it fails, you just get up and you move on to the next. People really might get kind of burnt out in this fast-paced
environment. I also heard you talk about play and there's also play in the organization, video creations, hackathons. How do you balance this very much so delivery oriented speed mindset versus also the resiliency and still having fun part? Yeah, I think, well, first of all, I think people underestimate that if you build something and people use it, that's the best feeling in the world.
So I don't know if there's a running analogy for this, but like I, I, I rarely see people being super excited to work 8 months on an internal migration. There are people like that. And I love that, that we have those. They're very rare. Most people, they build something, they see people. I mean, that's just a magic, especially with a creative tool like feed. So I think just in general, it's kind of fulfilling to work on
these things. But what we hear a lot in our organization is like, hey, we've built like 10 fast things, but we need to do some maintenance work. We need to make the experience more smooth and things like that. So we try to find a balance there. We have some new things that are really exciting. That's where a huge portion of our team is, what a huge portion of our team is working on.
And then we also have initiatives that are a bit more longer term, things that we know are also going to be important 268 months down the line. Like there's a few things that people use V for that just need to be rock solid. And there are things where we know, okay, if we built this, a lot of our users are going to be happy. So for those bigger initiatives we use, we leverage research a lot and we do a lot of user research and we do bigger
projects, bigger bets. And then on the other hand, we're like, let's just see what sticks. It's not that simple, but a little bit like that. Let's just see what sticks. And then when we know that something sticks, we'll double down on it. So we saw that, for example, two years ago, we built a ChatGPT plug in. We build in a day. We thought it was going to be nothing. It was a huge hit. People loved it, VGPT. And then we were like, oh, OK, well then we're going to build
this out. And then we build it out. And then people using that plug in came into the product and they were like, oh, but we want the product to be this and not that. And that helped us inspire like a lot of long term work as well. Yeah. I love that we touched on feedback and more so the qualitative feedback. You also mentioned product data,
product analytics. Has product analytics really evolved with regards to Gen. AI and kind of the metrics that you look at or is it kind of still the same and it holds true with regards to new functionality, AI or not? I, I think what is harder with AI is that there's trends go faster. So we used to look a lot at search traffic, for example, when it comes to video, and we would see like, oh, podcast video, that's a new trend in search and that those trends would like slowly build up.
And we would be like, oh, let's build a feature for it. And then we instantly knew we would have distribution. That is really hard right now because new technologies enable new trends. So like the anime video meme stuff, that was like somebody figured out how to do it. It became an instant trend. Everybody was Googling it. But you know, by that moment we were too late. So I think that is not possible anymore. You have to take more bets there.
And then another thing I'm seeing changing in the data landscape is you can now do a quantitative analysis on qualitative data using AI. So we transcribe all the user interviews and we look at different teams and it can allow, it allows us to, yeah, really quickly understand better something about a certain problem.
So for example, if we are talking about recording, we used to be able to look at recording data, just the numbers, but now we can match that with like, oh, what have users said about this part of the recording flow in the thousands of user interviews that we've done? So that's been a huge, huge improvement for us. I feel like that's gold. That's something that wouldn't be possible otherwise.
And especially if you have this culture where you build up this knowledge base of user interviews, you can also think of, OK, what has already been said, what are the trends across interviews? Because what's the last interview is the freshest in my mind. Like that's the easiest. And then leveraging AI to go through that catalog. I think that's genius. Yeah, it's incredibly valuable. Yeah, it's worked really well for us. And we also record all those
interviews. So then you can make and this, if you ever want to convince somebody, make a compilation video of users talking about a problem. So there was one designer that was really passionate about redoing our dashboard in the in the product. And I was a little hesitant. And then he made this video. Oh, yeah. It's like a compilation video of all this. You're saying I don't know where, where, where's the button? How do I create a project, a conflict?
And I was like, OK, OK, OK, let's, let's get it, let's work on this. So, so I think that's like, you know, you use the, the product data, you use the, the qualitative data with AI and then you turn that into something that is easy to digest, then you're absolutely killing it. It's really funny that you've armed everyone to kind of this use this tool set. And even when they disagree with you, they can still use it to persuade whomever. And that in the end is also
product, right? Doesn't matter if you're CPO or if you're product manager or whatever. However, in the end it's about taking people along with you and kind of rationale and reasoning about what you think is valuable. If you can't convince otherwise, then if you can't convince people, then it's not going to fly. Basically. That's the hard part. That's also part of the job. Yeah, yeah, that's what I meant by like building a big context window for everybody in the company.
And, you know, I'd love to be wrong. Like, I just want to make sure that we're building the right thing, so if people can figure out better ideas, then please. Yeah, I guess. Yeah, I love that. You touched on one thing, which is kind of my last thought. I think trend based products, because our product features, I should say, everything that goes viral, there might be a place in your product. But you already said for this thing we were too late, right, Which also means there's a start
and an end. Were you there at the start, then everyone would have been like, OK there and we can do that there, but there would have also been an end. So then you build features which kind of have this shelf life that's new. It's like spoiled milk, basically. I feel like we haven't seen that before. And how do you accommodate for that? Do you jump on when you think it's valuable, but you don't know what the payoff is going to be? In any case, it's still speed.
Everything's about speed. Yeah, Yeah, I think, I think this is very new to product. I think the ephemerality of products has has increased by a very large magnitude because AI is developing so fast. Now there's ways to predict trends. One of them is every new model that comes out enables new use cases. So if you know what models are going to come out, and if you're like Veet and you have early access to a lot of those models, you can kind of brainstorm about, oh, what could we do with
this that we couldn't do before? So that helps you predict those trends. But yeah, you're right. I think that not everything that we're building right now is going to be super relevant 612 months from now. I think the things that people use ChatGPT for 12 months ago are not the same things that they're using it for today. And that's okay. But what you need to do then is to build an interface that can allow for, can allow those use
cases to kind of transform. So we have now this thing which we call the AI Playground, where people can use our AI models to build different things and we can build easily build new use cases on top of it. We can add new AI things to it. And in that way we can really quickly jump onto trends, but also create a common interface to kind of build different types of videos. And that's been that's been
really successful. And we're thinking a lot now as product team, how can we build interfaces that are kind of trend agnostic, but can also adopt to whatever is popping right now? That's it's really interesting like for me, anticipating trends is one of the hardest things right Everyone tries it with like viral videos Hooking into communities is extremely interesting, like anime. I'm an anime fan of gapping. I've been watching anime for like the last 20 years of my life.
I can speak a little bit of Japanese and it's all because of anime. So when that trend happened, so happy. Like of course, of course I used it. Of course I was going to use it in whatever product I would find. It makes sense to hook into communities, but it's incredibly hard to anticipate and to figure it out even with kind of early access. I feel like keeping your options open and making sure you're fast when it happens I think is the best thing you can do. Yeah, exactly.
And and we, you know at we, we should never underestimate our distribution. Like we have 10s of millions of people coming to the site every month. So if we jump on a trend fast enough, they'll probably see it here 1st and not somewhere else. So even if we're not the first, we might be the first time people people hear about this. This is something I also learned when we were doing subtitles. So we do these automated subtitles, which are very
popular. I thought that thing was kind of over a couple years ago, but still to this day I, I speak to customers that are saying you can do AI generated subtitles. It's mind blowing. And I'm like, oh wow, we've had this for six years. Like what's going on? I mean, you are on top of the game, right? And you're an expert in your field.
You look at what the competition does, so then you're probably miles ahead of the average user who sees something, probably gets into video and content creation, and it's like, OK, what is out there? What can I do and just get stuff done and whatever you can have to get get make that better. It's going to blow their mind probably. I was definitely like that. Like I I used the tool and then I knew there would be at some point auto transcription.
I would use others tools that their sole purpose was to do auto transcription and then I would use it begrudgingly be like, if these tools have that, you're out of business. Basically, I'm just reducing this temporarily and exactly when it hits, I'm all in. Basically I want to use one tool and that's it. I want to have one platform and I want to get really good at that platform. And ideally I want it to be free, but I understand I also have to pay if I want to have good stuff.
So that's me as an as an editor basically. You're very much like many of the other editors. People want all in one solutions. They don't love paying for it, but if we make a good enough, they're willing to compare some change. That's it. Yeah. Man, I've really loved this conversation, I must say. Fascinating what you've achieved and also how you're thinking
about AI from a product sense. Are we thinking about AI from a productivity sense in your organization, but also your advice for other organizations in this conversation? Is there anything missing that you would still like to share before we round off? No, I think, I think this is this is really fun. Yeah, thanks.
Thanks for for having me. And I I think maybe the one thing I would say is try out our platform and and have a play with it. And I'm very curious to to hear what people people do with it. Awesome thank you so much for listening. Check out Sam. All the socials will be in the description below. If you want to support the show, leave a like that's the only thing you have to do. It's free, you don't have to pay for it, especially on any of the platforms. Apple podcast, Spotify or YouTube.
If you're your fan, do all three and otherwise we'll see you on the next one.