20Product: How Canva Builds Products: Lessons Learned, What Works? What Flopped? The Top 5 Product Lessons in Scaling to 185M Monthly Active Users with Canva Co-Founder and Chief Product Officer, Cameron Adams

0:00Speed is definitely important. Like you can't take five years to launch a product, but it also needs to reach a certain level where people get excited about it. You don't just want to launch something that people feel like it just gets the job done, but they're not that crazy excited about it. Launching something at Canva, that people will spread, that they will tell others about, has been the biggest growth driver for us. You are listening to 20 product with me, Harry, The most steady now 20 product is the monthly show where we take you inside the world's best product teams and sit down with the best product leaders in the world.

0:36Today we're joined by none other than Cameron Adams, Chief Product Officer and co -founder of Canva, where he is responsible for heading up the design and product teams. Since launching Canva in 2013, Canva's global community has grown to over 185 million monthly users in 190 countries. In 2021, Canva was valued at $40 billion. Following a $200 million funding round, this saw Canva become one of the most valuable, private software companies in the world. It's an incredible story and this is a masterclass from Cameron Adams, co -founder at Canva. But before we dive in, ever wondered what your customers actually do?

1:16The team over at They Do have been living it for decades and they have recently built a product that is frankly enabling the world's biggest businesses to literally know what they do. Harnessing journey mining with AI, you can turn unstructured data into structured customer journeys in minutes, saving weeks of work and months of analysis. They do journey management platform is already helping over 100 leading businesses, including Ford, Johnson and Johnson, Royal London and Home Depot, to know what they do. Don't waste another step, check out theydo .com. But before we continue, let me tell you about one schema.

1:51One schema is the embeddable CSV importer that will save your team months of development time. While building a basic CSV importer may seem simple, product teams typically spend over three to six months building and rebuilding their CSV import experiences. Why? Customer files are full of unexpected formats for dates, numbers and addresses, and your first pass at building an importer never has all the edge cases handled. Failed imports lead to unhappy customers and endless emails and tickets for your support and engineering teams. Enter one schema with the largest library of pre -built validations and intelligent transforms.

2:27One schema helps you launch a guided CSV import experience in just 30 minutes. Get your customers to the value of your product in minutes and say goodbye to frustrating messages like import error on line 53 with over 4x faster performance compared with alternatives. It's clear why one skimmer is the choice for top start -ups like RAM, scale AI, and Pave to streamline their CSV import process. Importing clean customer data into your product is easier than ever. Learn more by visiting oneschema .com slash 20. And finally, we need to talk about Pendo. A really simple way to describe Pendo's value is to simply say, get your users to do what you want them to do.

3:05What is Pendo? The only all -in -one product experience platform for any type of application. What are the features that make Pendo so awesome? Pendo's differentiation is in its platform. Every capability from analytics to in -app guidance, to session replay, mobile feedback management and road mapping are all purpose -built to work together. What is the social validity? 10 ,000 companies use Pendo and we also manage mind the product, the world's largest community of product management professionals. Where do we want to drive people? Give it a try and visit pendo .io. slash 20 product hyphen podcast to learn how your team can use Pendo to start building better digital experiences There you can also check out Pendo's lineup of free certification courses 12 hours of in -depth training for your product management teams on topics from AI to product analytics to product led growth Cam I am so excited for this.

3:59I was just chatting to Cliff beforehand and I was like Why do I take this conversation? He's like, I'm not gonna do the accent because you just ridiculed me for it. I was actually gonna do it before, but he said you've got to ask him, how did he come to meet me and Mel? And he said there was a little bit of confusion on your end about why you were there as an advisor or actually a hire. Talk to me about that. Well firstly, don't ask Cliff about content. But secondly, I got intrigued to melding Cliff through my old boss at Google, Lars Rasmussen. So he was one of the founders of Google Maps. And I worked with him on a project called Google Wave and we did that for like four and a half years.

4:36I left Google to do this other startup with a couple of other engineers from Google and we were going very happily doing that. We were in the process of fundraising and Lars said, go talk to these people about some tech that they need. And I went and talked to them. They just moved to Sydney over from Perth at the time. They had a school yearbook business. Running this thing in a doby flash, it helped schools make their yearbooks and they kind of wanted to take it to the next level. But they weren't going to do that with Flash so I knew a bit about JavaScript and HTML so I went to chat to them about that.

5:09Walked into their office, Mel was sitting at their conference table, trying to look very professional because they had like five staff at the time. I walked in and she's like, oh you're here for the PHP developer role and I'm like, no I'm here to tell you like how to actually run the tech side of your business. So we sat down, we chatted a bit about the latest tech at the time, what browsers were capable of, and particularly in the scheme of creating design. So how can you do visual design inside a browser, which was quite tricky at the time. We explored it a bit, we went for coffee afterwards.

5:43Mel told me about a bigger vision beyond yearbooks and it was democratizing design. It was bringing design to people who hadn't been able to access it before, which was really intriguing to me because my background is a graphic designer and also computer scientists. So merging design with democratizing technology was perfect for me. But me and my co -founders in my other startup called Fluent were like, very happy. We were gonna take over the world and we were just about to land $5 million in funding. Little did we know that we weren't just about to land $5 million in funding, but that's beside the point.

6:16So put it in the back of my mind of what Mel and Cliff were doing and went off to do what I wanted to do. And a few months later, after our startup had kind of created a little bit. That idea of democratizing design really came back to me. So my wife, we were looking at possibly moving to San Francisco because there wasn't much happening in Australia at the time. My wife wasn't that keen to move to San Francisco, neither was I. So she said, why didn't you go hit up melancholy again and see how they're going with trying to democratize design? So I did and I hadn't got funding or anything so far.

6:50I hadn't found a tech co -founder that could help them build the product. So we decided to join forces. That was 2012. Weird one. How much of success do you think is luck versus skill? Obviously you're an incredible leader and incredibly skilled guy, but actually you'll start up not getting funded. It was probably the luckiest thing in many ways happening. Yeah, it's all these sliding doors moments. I wouldn't necessarily call it luck. I think it's about putting yourself in a position to capitalize. That is talking to people, networking, having the skills, like working with technology, trying different products, experimenting, Canva we started in 2012, but Micre is going for like 12 years before that.

7:33And there's a ton of ideas that I tried with a ton of people that didn't work out because it was the wrong time or there's actually the wrong people, we weren't a good team. Google Wave itself, like the project I work on at Google, was really precient in what it wanted to do in terms of real -time communications and collaborations, but it was early for a bunch of reasons. Like, browsers weren't quite ready, people weren't quite ready to work in that way. So, six years later, launch it if you become something like Slack or Microsoft Teams, it can take over the world, but in 2007, when we tried to do it at Google, it just wasn't the right time.

8:09So I spoke to Mike Haberon also before this, and he's both about creating the first Canva landing page, and the amount of time that you you spanned on it, making it and quoting him just right, and so much attention to detail. My question to you is, we're told in the generation of the lean startup, the importance of speed of iteration, speed, speed, speed, and it feels like it goes in conflict with the making at just right and the time to make it perfect. How do you feel about speed versus quality and that importance and balance between the two? The Lens startup book actually launched in the year that we launched is like 2012.

8:46Investors were quoting it at us and like telling us to launch as quickly as possible. We kind of ignored them because we knew what we wanted in the product and we knew we wanted something high quality. Speed is definitely important, like you can't take five years to launch a product, but it also needs to reach a certain level where people get excited about it. You don't just want to launch something that people feel like it just gets the job done But they're not that crazy excited about it launching something at Canva that people will spread They will tell others about that they will shout from the rooftops about has been one of the most important parts of our success Because those people telling other people one amazing product it is what a delightful experience they had what fun they had designing something has been the biggest growth driver for us.

9:36How do you create such fanatical users? I mean, for me, when I look at your growth, one of the most astonishing things is counter -talk. You know, the unbelievable UGC content that your community creates. How do you create such fanatical users? I'm sorry for such a basic question. Part of it is that detail, like creating that landing page that has a great animation on it that people spend five minutes just on the landing page playing around with that before they sign in. Like it's those things that people a feel the love like they can tell it's a well -crafted product that they want to use and Be bring something to their life that they didn't have before like it's a very small thing It's a moment of delight But it's not like using a spreadsheet or doing some boring job that they do every single day It's about giving them something new in their life that Interest them that unlock something those details in the camera product have driven so much use a love like we get tweets about the little things we put in, like the Easter eggs that we have.

10:34We have a duck that floats by every time you upload a hundred images. We thought about the onboarding process and making that fun so that when people first come into Canva, they discover that they can actually design things and this moment of delight when they discover that is what we want to trigger. That's like our aha moment. How important is the speed to delight moment? Definitely needs to be in the first user journey. I don't think it has to happen in the first 30 seconds, but you need to give them a feel for the product right from the very get go. If you have a crappy sign -up page that looks like it was made on my space in 2006, like that's setting the tone for your product.

11:12So they don't have to be immediately wowed, but you need to start giving them kind of the vibe of the product right from the very get go, which is where those details really fit in. And then that experience should build up to that wow moment. They've signed in, they're giving you some details. And for us, that wow moment was when they had created something that felt unique to them, that they had put enough of themselves into. Then they realized that through this process of onboarding, they had become a designer essentially. They had done something that they didn't think they could do before. That was the trigger moment where they're like, wow, this product is for me and I can use this in 101 different ways.

11:51In terms of that ability to have superhuman powers of design and to democratize design, it has to be simple. And I think it's really interesting when I spoke to Andrew before. He said that you're a master of making things just simple enough. And I really am always struck with this in product, which is like, is simple always good? How do you think about just simple enough when product designing? The way that we think about the product experience is that we want anyone in the world to be able to access design and to get the benefits of it. But also, as people spend more time with your product, it needs to unfill and unlock more stuff in it.

12:28They need to continue to become better designers as they use it, get better outputs. So there's this 80 -20 rule that we apply where, when you first get into Canva, you should be able to get 80 % of the value. But then as you spend more time with it, you unlock the remaining 20%, which makes you an expert, it makes you put more into the product and what you're doing and discover things and create those moments of delight constantly. Previously you didn't care about typography after using Canvas for six months. You discover all these things about typography and we give you access to those. So it's a real power unlock.

13:01So like simplicity is the top layer, but you also need depth under there for the people who spend the time to explore it. What's been the hardest thing about the move -in to enterprise? We've had a couple of cracks at Enterprise. So we actually launched, we had a journalist who did some good in investigation and said, you launched Camera Enterprise like three years ago. And what we launched three years ago wasn't really an Enterprise product. It was just a way for someone to get in contact with a salesperson. So we hadn't, I don't think at that stage really listened to the Enterprise audience.

13:32We're just like, you want to scale Canva to lots and lots of people. Here's a self -service thing that you can do to like get your credit card, I didn't get a thousand people using camera. It wasn't a true enterprise experience. Something that I worry about is the unbundling of this space. We obviously have content team as well. We just see so much specialization, whether it's short form video generators from YouTube links to podcast artwork AI tools. To what extent are you concerned about the unbundling and to highly specialized products that do one thing very, very well? I think we're seeing particularly in the AI space, we're seeing a lot of hype and churn and real microservices and you're also seeing a lot of those startups kind of fall by the wayside as well as the real value of them comes to light and it's not very huge.

14:23For us we see immense value in creating a true platform that enables proper workflows. We mentioned this AI tool, it was one of the questions that I worry about with those AI tools is to see infinite supply of content that comes out of the media, you give them a YouTube link and suddenly you get 30 clips. And they're not great quality necessarily, definitely not nearly as good as what we do say, for example, but just human is still so much better. But there's an infinite supply that does gobble up human attention. And to your point on like being able to create 30 brand assets in a while, we used to create one.

14:54It's kind of good, but I worry that the value of content decreases as the infinite supply becomes more prominent. I don't think the value decreases, it's definitely more people creating stuff but as you very well know great content rises to the top. I don't think it does. I think that's like saying build a great product and they will come. Well yeah but in a time when there's very few or when there's less great products I think now discovery is everything and actually discovering consumers finding you is the simple harvesting. There are thousands of podcasts that are amazing but podcasts is the shit that's to be in now because it's so hard and that's used to odd years ago.

15:33I mean, you can argue with me, Cam. I can argue with you. You see the same thing in the music industry. Like, what you hear on the radio isn't necessarily the best song in the world, but it's about discovery, it's about being a fully rounded artist. Someone who can have a personality who offers people something unique, gives them a different take on the world, as well as creating great music. So Taylor Swift is an amazing marketing machine. She thinks about the attention she can garner the way that she presents herself the whole strategy of her to her like all of that is part of Taylor Swift It's not just Taylor Swift sitting at a desk writing a song.

16:12What was the product change that caused the biggest spike in user adoption? And what did you learn from that? Oh, that's an interesting question. Text to image was quite pivotal for us. We launched text to image 18 months ago. We were first product to like truly scale generative images to over a hundred million people. And that was really pivotal for us because it was our first dip of our toe in the water of generative AI. And that has sparked an incredible amount of product innovation and strategy over the last 18 months. Pretty much every touch point in the product now involves some level of AI.

16:48What was the response? It was amazing. I think we wanted to get it out early. it was very only in the piece and actually only took us six weeks to launch from way to go. From someone saying, let's do crazy idea, let's do like type in a box and you get an image out of it. Six weeks later we deployed all the infrastructure, we did integrate it into the product and we launched it to 100 million folks. And we did that because it was an incredible piece of technology that we wanted to like bring to people, but also we wanted to open ALAs infrastructure. But that one that was based on stable diffusion and since like iterated upon that a whole ton.

17:25And it's time by us here where you move between different providers. Yeah, definitely. You know, there is a heap of competition now in terms of large language models, image providers, any sort of AI model you're thinking of, there's probably like 50 different versions of it. And I think that's great because not having a monopoly from open AI or Google or Anthropic or whatever model you want to look at is beneficial for developing a product. I'm naive, and the speed and ease to transition between the two is little. It's getting a lot faster. Early days, it was a bit harder because there were different infrastructural needs, but for them to be successful products themselves, offering tools to developers, they need to make it easy to just wrap your product around it.

18:10So the APIs are getting better. The way that you call different models is getting standardised, So it's becoming a lot easier just to switch over. Can I say when you rely on someone else's infrastructure, like many are now on OpenAI or on stable diffusion models or whatever that is, it kind of dictates product roadmap. And before you'd set the product roadmap, and really only you and your execution determine whether you're able to complete it. Yeah. When suddenly OpenAI can release a new model, and whoa, it changes the game. And we had the CPU from Loom on the show, you know, the video. Yeah, yeah.

18:41and she said, it's the hardest time ever as a product leader, because I'm suddenly, my whole product roadmap is turned upside down by an open AI improvement. How do you feel that the world of product creation has changed now when this is the case? I wouldn't say it's changed. I think that particular area is experienced rapid acceleration, but I think it's the same for any, the early stages of any technology. There's room for innovation, improvement, and discovery. In five years time, I would say that AI innovation will slow by then and a lot more will be standardised. Social patterns will be established of like how we interact with AI, the possibilities of what LLM can do, are very well understood and known and it's just become a part of everyday product development.

19:26How will we interact with AI? I think a lot more than just typing into a chat box and even a lot more than just talking to it. that interface of working with AI. It's a really great demo and it's been a great inroad into exploring it, but it's going to become a lot more nuanced. AI is getting a need to... Most people are great at coming up with the very first idea. It's one of the things we learned at Canva when you present people with a blank page, they just like, freeze up. They don't know what to do. And at the moment, AI is just like that. You have to like, go to it and say, I want this thing and it will give it back to you.

20:01It's not that great at helping you explore ideas and iterate on them, give you the very first prompt that you should be prompting it with. And that relationship between you and the AI needs to get a lot better as a collaborative partner. And that means you need to develop the interface a lot more. So our magic media tool, which is the evolution of text to image, is a great example of that. It started off as a text box, but most people can't describe an image. They don't know the exact style they wanted in. They can't tell you between the different types of artistic drawing that's possibly available for a duck that you want.

20:37So actually on that then, we had Gustav from Spotify on the show. And he said, probably one of the biggest questions you're asking interviews for product people is show me your prompts. And that will be a determining fact of the quality of a product person in the next few years. Do you think that's right? And how do you think about that? The quality of prompts being a determining factor in terms of quality of person? I think prompting is definitely a skill that we're all going to need, but in terms of it being the defining part of a job interview I would hope it doesn't go down that path if in five years time the way that we interact with AI is Purely just through prompts.

21:14I think it will have displayed a total lack of imagination from us as product builders How will we see the stabilization of LLM's like you said? Well, how will that end -state look in five years time, do you think? I think it all looked very similar to what cloud -based infrastructure looks like now in terms of AWS, Google Cloud, Azure, how I often do switch your cloud provider. Not a huge amount, and there is definitely a bit of luck in there, but it is possible, and the starting point is also very different for different people. Like when you're starting your project, if you're familiar with Azure, you can go on that.

21:50If you're familiar with AWS, you'll start with that. If you're familiar with Google Cloud, you'll start with that. But you do have choice there. It's not just one provider that you have to do and they become monopolistic and they're not rolling out anything new and they just take a bunch of money from you. There's still a huge amount of competition there. Exactly the same will happen with LLM. So they'll be innovating. Google will add a feature, open AOI, will add a feature, some other startup that we don't even know about yet will add a feature and everyone else will rush to catch up and then innovate on top of that.

22:21But you'll have choice. My question is, do you charge more for magic media and for AI capabilities? It's all wrapped up in our camera pro or our camera enterprise subscription. Is that your margin structure then? Not really. It is a tiny bit of like a hard user of the magic media. It's creating a lot of images with AI. The idea is that would actually then dampen your margins. There is a little bit of cost in there, but it's the approach we've taken for the last 12 years. Putting an immense amount of value into our subscription is why people love it. You need to constantly be adding to your product, getting people to pay the same subscription for the exact same product as it was seven years ago just isn't gonna fly.

23:02Plus, we've also seen the costs on AI just continue to shrink. But do you think AI expands the revenue lines of existing great companies like Canva? Would you just think it allows them to have much better products with happier customers? That's the question, like does AI actually allow big companies to make more money or just create better products? In the short term, it might allow them to make more money. So like, I think Microsoft and Google both have add -on packages that you pay for to get access to the AI stuff. I think that's a very short term thing where it'll just become commoditized, people will become used to it, and it'll just become part of the basic package.

23:39And I think they've done that to kind of shelter themselves to any possible price repercussions from people just smashing their AI servers. But yeah, we view it as just a way of building a better product and it's a continuation of democratizing design. 10 years ago, we used the cloud and mobile technology to do that. AI is just the next thing now that enables us to deliver an even better product experience. Do you think the structure of product teams changes with AI being such a core part of the development design distribution process? It does a bit, definitely the way that we developed product has changed over the last two years.

24:16How so? So we've become a bit more R &D focused in the sense that technology needs to be explored a lot more. For the last 10 years, we've pretty much known the technology stack and what the technology is capable of. And you can look at a mock -up of something and go, yeah, we know how to build that. We can get the Java servers to do this and JavaScript in the browser will be able to do that. So it's kind of a known quantity. Whereas now you can come up with an idea for a new feature that you want to roll out. So say you want to roll out. We just rolled out in hands voice so you can take a video, clean up the audio on what someone's saying and it does it with one click.

24:52But when you initially do that, you're not quite sure whether you can pull it off because the AI model needs to be developed to do that. There's a whole bunch of smoke and mirrors that you need to do to get it to work and there's a lot more uncertainty in there. So you need to be comfortable with that uncertainty and the product needs to evolve with where you're seeing the technology go. Six months ago, we might have set out to do enhanced voice. Some of the engineers will try it out and see how far they can get with the quality. Maybe there's a particular use case where it doesn't work. So say you're in a taxi, it doesn't work for that, right?