My chat with Joseph Cohen, Founder & CEO of Univer.se

We spend so much time on the internet on our phones. For mobile consumption, we can now create and manage pictures, audio, and videos on our phones. The same cannot be said for websites. We create websites on desktops and then consume them on mobile. There is a reason “Is it mobile optimized?” has become a common refrain. I have always had conflicted feelings about creating websites on mobile for mobile. On the one hand, it makes no sense that mobile creation of websites is not mainstream. On the other hand, it’s easy to wonder why anyone would want to do that! The screen size is tiny! It doesn’t sound convenient! Desktop tools are so powerful!

To understand the mobile-based creation of websites, I talked to Joseph, founder of Univer.se, a mobile native website creator. This chat is packed with creative ideas and thought-provoking insights. Joseph is an incredible storyteller. He is an effective spokesperson for the idea of thinking of our phones as creation devices. We talked about his personal story, how Univer.se has evolved, and his reflections on company building. He also shared crisp frameworks on how to think about software as appliances versus language, the nature of constraints in designing software, and the relationship between ease of use and power of tools, among many other things.

Sar: I've always been intrigued by the philosophy behind what you do. It's both counterintuitive and logical at the same time, depending on how one looks at it. I want to understand your instincts and motivations that got you here. How did you start Univer.se? Help me see the world the way you see it. 

Joseph: I grew up in New York, but I like to say that I grew up on the internet. I got a bondi blue iMac when I was 10, which changed my life. It opened me up to the world, and I was like, “Hey, I want to spend my life here.” From then on, I started building websites, launching businesses, and designing products. But I'd always felt that building things on the internet didn’t work as my brain does. I can write code, but it is not how I want to spend my time. 

Historically, to create things online, you could either write code or use an off-the-shelf tool, like Squarespace or Wix. When writing code, you have complete power. But you have to bend your brain to work the way the machine works. On the other hand, when you use a legacy website builder, what you create is determined by the template you start with—they’re not open-ended. 

So you either have to learn the esoteric language, but you have no constraints, or you can take something off the shelf, but your tool is prescriptive. 

Sar: Right, the choice was between no and too many constraints. 

Joseph: It's not just that there are too many constraints. It's the nature of the constraints. Whenever someone writes a new application with code, it's entirely new or a unique combination of what's been written before.

When you use a product like Squarespace, it's not new 95% of the time, right? Because the system itself is designed so you cannot break its boundaries. It's not a generative, open-ended system. It's more like an appliance than a language, which you can combine in unlimited ways. So those have been the options. This all became clear while working on my first startup a decade ago. I went to UPenn for undergrad, and while there, we used this less-than-desirable learning management system called Blackboard.

Sar: So did we! 

Joseph: I designed a better version of it and started to get traction. It was a classic student entrepreneur project. I started to get investor interest, and after two years of school, I left and started a company around what I was building. We eventually sold it to another larger edtech company. And I was off to the races as an entrepreneur. 

It frustrated me endlessly that if I wanted to make our homepage or make an edit to our homepage, I needed to bug an engineer to do that. I knew exactly what I wanted to do.  Around the same time, I realized that mobile was not just about phones. It was about universal computing. Mobile devices were the first truly personal computers - computers everyone loved using all the time. All utopian ideas about the personal computer revolution from the ‘60s and ‘70s were going to come true because everyone had a computer in their pocket.

I thought to myself, if the primary method of experiencing the internet is mobile and will continue to be increasingly mobile, you need to be able to make the internet from a phone. What if there were a way for someone who wasn't a designer or developer to create on the internet from wherever they are on their phone? That would be incredible. That would change the world. This was 2013. The idea of how you do that was unclear. 

Trying to do it for a phone was incredibly difficult, but it was the kind of problem I was incredibly motivated by because, as a designer, you want the challenge of an insane constraint. It's entirely solvable with design intrinsically. There's no reason why you couldn't solve it. The question at Univer.se is, how do you empower everyone to build the internet? The company's raison d'etre is to solve that question.

The solutions will change as we learn new things, build new products, engage with customers and learn more about the world. You empower everyone to build the internet by starting on a phone because that's the universal computer. It should be visual, touch-based, and abstract complexity. It should meet you where you are. You should be able to grab a domain easily. You should get analytics. It should include a commerce engine. It all follows from that opening question. 

Sar: You didn't start by asking why we do not have a mobile native website builder. That wasn't the question at all. It was more abstract and powerful than that. Maybe one of the answers you eventually landed on and took off was the current manifestation of Univer.se. 

Joseph: Yeah. I'd say that our mission will also remain broader. Univer.se right now helps you build a kick-ass website from any device.

But we're in the business of internet building. And as the internet continues to evolve, we will be at the forefront of what it can do. For example, as the internet becomes more three-dimensional, how do you create that? It's not limited to websites. One of the things that people like to do online is sending emails. They'll use something like MailChimp. If you ever use one of these tools, they're not great. They’re creatively inflexible and not designed for mobile devices. 

We have built this editor based on a grid, our core offering for creating a website. There's no reason we couldn't use that for building an email. So it's the same core, just a different expression. Another one is how we think about native applications. I don't believe that most people or businesses should have their native applications because of the nature of the use case. So we have started with the most pragmatic and clear need: websites. But that's not the end in itself. And it is a deeper idea. We weren't in the website building business when I started the company.

Sar: What did you get started with?

Joseph: It took me about three years to figure out that this needs to be a website builder to go to market. The first version of the product was much more like a social network. And you were creating open-ended canvases called verses.  You could effectively make a webpage with it, but they were only visible inside our network. 

You couldn't access it from the web, which was the wrong idea because distribution is king. The killer feature of the Web is not HTML or CSS. It's the fact that it's available everywhere. In a world dominated by Facebook, Apple, and Google, the open space is the web. 

Sar: You talked about the nature of constraints and picking the right constraints earlier. You are now playing the hardest mode when it comes to website-making. Tools like Squarespace are built on the Web, which is far more open than the app ecosystems like Apple. You are building an app that produces websites that can be viewed from mobile and the web. Within that context, your design space is further constrained by what's doable with taps, drags, and clicks on a phone screen. Our users are not building on the Web but are trying to increase the output on the Web. It all feels so counterintuitive and technically challenging.

Joseph: Constraints are inevitable, and constraints are very generative. The question is, which constraints do you want? There are good constraints and bad constraints in my mind. In other words, our product has constraints. We designed a grid system that allows you to put blocks in certain areas and not other areas on the screen in our app. We have that system not because it’s technically easier to have those constraints but because people need them to make decisions.

If you have an infinite choice, it's impossible to start. So some constraints are valuable. The way I think about constraints is almost like architecture. In an elegantly architected system, there are principles and a foundation. Take a system like Notion. In Notion, you can nest components, build things, and reference them. So it's infinitely complex but built on strong principles. 

On the other hand, Microsoft Word has many features, but the product's architecture has no logic. They bolt on features, and you end up with this sprawling Frankenstein. So I would argue that a system like Squarespace is much more like Microsoft in its architecture design than Notion. For example, if someone wants to start a blog, Squarespace will just build a blog as a separate system and throw it on top of what they have. Squarespace recently came out with a product called the Fluid engine, which Univer.se heavily inspires, but they bolted it on. It's not intrinsic to the nature of the product, and you could feel that. 

You asked why I chose the hard mode in this sense. Apple App Store has made software mainstream. And today, it is still the best way to mainstream a behavior that has been the provenance of nerds in the past. And the whole thesis of the company is to take nerdy behavior and make it mainstream. It is the best front door. Apple has an incredible distribution channel. The tools available in iOS allow us to make a user experience that's leaps and bounds better than anything else. You can drag something around, and it follows your finger. And there's a sense of intuitiveness that comes because of the way that the platforms are built. When we got started, an iOS app was a great insertion point. The first version of Univer.se was very simple.

It was a toy app that let you build a single-page site. The pages didn't scroll. We didn't even have drag and drop, but you could use the grid system. You could put something online. And that was just a very good fit for the Apple ecosystem. There were many other website builders, but no one had a mobile native website builder. The bet paid off! Apple has given us tons of distribution. They have featured us many times, and we have a very close relationship with them. 

Sar: And are you guys still just on Apple? What do you think about other platforms and devices?  I would imagine the physical constraints of different device sizes can also create computing constraints. 

Joseph: That's a great question. Most of our usage is on iPhones, iPad, and Mac. 95% of the usage is on iPhones and iPad. There's feature parity across all devices, so that will continue to be true. Univer.se is not device-specific, even though most people know us as the mobile builder. It's not about mobile. It's about anywhere where you are. You should be able to build the web anywhere where you're experiencing the web.  The reality is that many people, if not most, don't use Apple devices all the time. So how do we empower them to build the internet? We are actively working on building a web-based version of Univer.se

Sar: Do you see tension with the core philosophy of staying native on mobile? Why not go after an Android app? 

Joseph: To be clear, the primary focus will be mobile web, which includes Android devices and other devices that don’t run iOS apps. We've just decided to focus on the browser as a platform, but we'll also build a native Android app in the future. We're focused on mobile web browsers, but the websites will work on a desktop browser. 

A lot of our traffic comes from social networks. Many users put their Univer.se website links in their profiles. When someone clicks on it and visits it, they might be intrigued enough to want a website like that. They would need to go to the App Store and download the app. That's too many steps. If we had a web-based experience without leaving Instagram, you could be making your website. Every place where a website is viewable is an opportunity to build your own. That's why we're excited about the browser as a medium; the browser's gotten so good on mobile that there's no reason we can't do this technically.  Ultimately, we want to expand access. You should be able to build a website regardless of where you are, what device you use, or whether you have our app on your phone. 

Sar: And would it get synced across platforms?

Joseph: Yeah, exactly. And that's how it all works now. Whether I'm on Safari or using the app, it's the same experience. I was working on my site over the weekend and started using my iPad; then I had a new idea for it, and I went on my phone, and it's all perfectly synced. That's how it should be. 

Sar: Most people face this universal problem of making their websites mobile optimized because they are building on desktop Web-based tools. Do you guys have the opposite problem? Technically you have to make sure that it does not show up funny when I pull it up on my laptop.

Joseph: You're right. We are biased towards the mobile experience, but that's because 95% of the traffic to our sites comes from a phone. So the experience for 95% of people is what you're designing it as, right?  It's much easier to scale something from small to big than from big to small. Because when you're going from small to big, it might look a little funny, but at the end of the day, it's not going to be broken.

Sar: Yeah, you are losing constraints instead of adding more constraints. 

Joseph: Yeah, exactly. This is another reason why it started on the phone. If you make it work on a phone, it's very easy to scale to other devices. 

Sar: Makes sense. Talk about the headspace or your customers' mindset when they build their website. The mindset with which people sit at their desks with their laptops is very different from when you are on your phone. What do you think of designing a computing experience for that on-the-go mindset? I imagine that mindset attracts a certain persona. 

Joseph: Our customers are what I call creative entrepreneurs and entrepreneurial creators. A creative entrepreneur is someone who's starting a new fashion brand. They created a brand from scratch, and they're making their website for the brand. An entrepreneurial creator is a musician who wants to make it big. They're building their brand. The fact that we have all the tools in one place makes it work for them, too. 

You use your phone everywhere. It's not like you're using it only when you are on a commute. It's with you when you sit on the couch or at a bar while working out in the field.  If someone is using Squarespace and you ask them what they are doing, they would say I'm doing work. No one thinks it's fun. It's a chore. However, the people who use Univer.se are having fun and will lose hours. Not because they're fussing with it in a frustrating experience, but because it's just fun. 

My core insight as a designer was if you make this stuff that people normally think about as work feel fun, you'll be unlocking a new demographic of builders. You can almost trick people into becoming developers. What's unique about using Univer.se is you are designing and developing a website simultaneously. If you like traditional web development, you might design a site in Figma and then code it in another system or pay someone to code it. There's a clear break between the design and development processes. Still, the truth is the best kind of creation has both things happening simultaneously because as you build, you're generating new ideas. You cut off the feedback loop between the execution and the ideation process by keeping them separate. So when people are in Univer.se, they don't know exactly what they want. They're feeling it out as they're using the tool. And that's just a very different process. 

Sar: The tighter feedback loops are often associated with simple, no-code tools. A common critique of those tools by people accustomed to alternative systems, the open-ended systems as you called them, is they are not functional enough.  Do you think feedback loops and functionality are at odds with each other?  The feedback loop is the longest when you design something in Figma and later code it. But that's often the path with the most customization and functionality. 

Joseph: People often think the ease of use is inversely related to the power of a tool, and that's not true. The question is how you package the power and complexity. You can have a tight feedback loop and an infinitely powerful tool. In my earlier point about Notion versus Word, both have complexity. However, the way that Notion does it is much more elegant and much easier to use. It allows powerful functionality if you want it, but it doesn't put it in your face upfront. And that's a similar philosophy to Univer.se, which is to say that it's super easy to get started, but we also have complexity.  You reveal the complexity on demand. You don't start with that complexity. It's about how you expose the complexity and how you manage it. That is the art of good design. You walk the user through it one step at a time. Once they understand it, they're moving quickly because they've internalized how that system works.

Sar: That makes sense. It's the idea of progressive discovery. How do you talk about what you do to your users?

Joseph: I've learned that the key is to package abstract, powerful ideas in a very familiar, almost prosaic way. Yes, our product is a website builder, but what does that mean? That's not a very descriptive technical description. To your point, we make an iOS app with this drag-and-drop grid-based, Lego-type system on mobile that produces an output that can be rendered everywhere. That's different in its implementation than any other product in our category.

A website builder is not an accurate description of the thing intrinsically. Still, I think good go-to-market and product positioning fit novel solutions to existing frameworks for understanding the world. So we use the most straightforward language to describe what we do. You're building a website and an online store. You're sending emails. You're adding a product to your site. We're not being creative in the framing of it. We're being creative in the implementation. Being creative in marketing can confuse people who could be viable customers. 

Sar: You have mentioned your grid editor a few times. It is the main innovation in Univer.se’s core user paradigm. How did you end up with that? I imagine you rejected many other things and landed on it. 

Joseph: After lots of trial and error, two previous user interfaces, and years of contributing inspiration. At first, I started with a blank canvas interface, like Figma, that was open-ended, but it didn’t feel right on the phone. It was an intimidating paralysis of choice. I then worked on a structured template-based tool. That was easy to grok, but it was closed-ended: there wasn’t room for people to make their own creations.

I’ve long been interested in grid systems and modular systems. Grid systems divide space and make it easier to wrap your mind around. Grids are used for all types of design: graphic design, architecture, apps, and cities. Modular systems have simple components that ladder up to complex possibilities. Our GRID system combines both to make building the internet feel like playing with Legos. The GRID is less intimidating than a blank canvas and way more open-ended than a form-based template editor. It’s a modular system composed of blocks that can be used for anything you want to add to a website: an image, text, or a buy button. 

Drag and drop blocks to move them, pinching to resize…the system is incredibly fun and powerfully open-ended. Our creators build entirely new, custom designs every day. That isn’t possible in a template-bound system like Squarespace. We see GRID as the most intuitive and powerful way to create things online. And we see no limit to its capabilities. There will be thousands of blocks made by creators around the world.

Sar: Talk about your business model. What has evolution been like? You have been leaning into commerce over the past couple of years. 

Joseph: We have a freemium product. You can use Univer.se for free. You get access to a lot of our tools. If you upgrade to a PRO subscription, you can get a custom domain and access to our commerce tools, analytics, and other things. If you're using Univer.se for a commercial purpose, you should upgrade. But many people are using it for birthday parties, and it doesn't make sense for them to upgrade. We built our commerce tools because it was the most requested feature.

From day one, people wanted us to allow them to build stores. I was pretty reluctant to do it because Shopify exists. It's a whole other business, a whole can of worms. And I was intimidated by that. We later realized Shopify's not the right product for our customers. It's too much. They don't necessarily know what an SKU is. They're earlier in their entrepreneurial journey than your typical Shopify customer. And they want to be able to design their website and build their business in the same place.

When they're starting a new business, the first thing they do is build a website. It's their front door into building a business. And so we were very naturally inserted into commerce.  We started small by allowing users to take payments. Over the past two years, we built out a full commerce engine. You could build an inventory and ship products and charge for shipping. A quarter of our users sign up for our commerce features. It's a major part of our business. We're not making much money on the take rate that comes with commerce. We know our commerce users more often than not convert to PRO.

Sar: What about the paid non-commerce users? What are they like? What do they do? How do you support them?

Joseph: It's a designer or an artist showing off a portfolio, or an individual planning a wedding, it's a rock collector, it's the full diversity of the internet. And many PRO users don't use our commerce tools because our commerce tools are optimized for selling physical goods. If you are a barber, our commerce tools don't yet support you. Over time, we'll continue to build out the suite of commerce tools.

Sar: When you look at all these desktop web-based website builders, they all have third-party ecosystems. What do you think about what integrations to build to support website infrastructure?

Joseph: We build out the most important integrations. That's as simple as a YouTube block for when you're adding a YouTube video to your site. 

We have built out custom solutions for specific services like Google Analytics. In the long run, we want to be the enablers of the platform, and we are not building everything ourselves. That is similar in spirit to how the tool itself is designed.

In other words, we will allow our users to become developers of the platform, and they will be able to create components in our system that they can exchange with other users. We want to enable a much wider demographic of users to become developers. We are inspired by Roblox, which has turned gamers into game developers. That's the idea. 

We recently launched a feature that will be a critical component of this puzzle. So right now, our system works by having your site on the grid and then adding blocks to it. But when you start making more complex blocks, you're making a block out of other smaller ones. A product block might have an image with a buy button and a description, which now becomes one new block. We call that a group block, which allows you to take any set of blocks and package them up. You can copy and paste it for your site. We're not yet enabling people to publish it for others to see and use. 

Sar: That's the next logical step because templates for you guys are specific block configurations. You want to help users save their templates and reuse them in the future. And later, users can showcase them. It's the playbook that Airtable, Coda, and Figma have used. But, none of them let users charge for them. Maybe that's something you guys should do. I'm sure I'm oversimplifying. 

Joseph: No, you are right! 

Sar: Can you think of an expensive mistake that may have cost you a lot of time and opportunity? 

Joseph: One of the things that I don't think I appreciated as much was how the nature of our product is inherently a lot. When someone's building a website, they need it to do many things, and they can't use it if it doesn't do one of those things. For example, if you were trying to get someone to switch from Sketch to Figma five years ago, it needed to have parity with Sketch. If it didn’t do gradient borders, you couldn’t use it if you're a certain designer. Our product's not quite like that because we're not replacing a professional tool for a professional type of user. However, it needs to do a lot of stuff. It needs to enable you to design any site in a particular way. It needs to support commerce tools. It needs to have your analytics and domain support. 

I think I took a little bit too much of the iterative approach. I was constantly trying to decide whether I build this or that. The truth is that for us, we need to have parallel streams that are building against our core jobs. Once we started doing that, we saw a lot more traction. I didn't think I needed more resources three years ago. We're now 60 people, and we raised a good amount of capital, which is the scale we should be at. Two years ago, we were 12 people and weren't giving the product what it needed.

Sar: You are in a well-defined market with many mature players. So you need to have all the basics first to get people to try and appreciate what's novel about the new product. If you don't have the basic things people care about, you don't even get a shot at delighting them.

Joseph: Yes, exactly! We still have our challenges in that way.

Sar: Do you recall any enlightening customer conversation in the early days that led to meaningful learning or a shift in your thinking?

When I first launched Univer.se, I wasn't sure if this was a little widget company. It's a mobile app. How complex are people going to go on their phones? I remember chatting with the guy who wrote a book on his phone and was building a website for the book on his phone. I asked, "Why do you write on your phone?" He said, "Well, that's where I put down my ideas. I'm in the park with my daughter, and I have an idea and write it." I realized there was an unlimited appetite for people to do things on their phones. For our customers, their phone is their computer. 

Sar: What is a long-held view about building teams that you are now reconsidering or changing your perspective on?

I flipped pretty hard on remote company-building. Univer.se has been distributed since before the pandemic, but I was deeply skeptical of building a company this way before I met Ryan, our CTO. He was the first person to join me and lived in South Carolina. He convinced me, not through his words but our working dynamic, that this was the future of work. 

We like to call Univer.se a “true internet company” in that we’re not only building things for the internet but also building on the internet. The internet is our office. As an online company, our interactions are mediated through software, which increases the power of computers to augment our work. Everything is recorded, searchable, and editable. We can work with the best talent, no matter where they live. Positive distributed teams align culture around things that matter—mission, values, motivations—and not an arbitrary distinction like geography. 

As we’ve grown, even as a distributed team, we’ve felt the crush of meetings. Up until recently, we were mostly North America-based. As we grew, I realized we could kill two birds with one stone by hiring folks internationally. We could reach great talent globally AND create a hard time-zone constraint that would force us to work more asynchronously and be less reliant on meetings. We’re in the middle of this transition. We’ve added a bunch of European team members, and just a couple of weeks ago, we added our first India-based employee! Over the past six months, we’ve made a ton of progress in working async, but there are still things to work out as we push to work at a high velocity that isn’t dependent on a bunch of people being on a video call at the same time. 

Sar: What do you think we are not paying enough attention to? It does not have to do anything with what we discussed,

Joseph: The internet is still in its infancy. Per the Wikipedia page for “domain name,” there are about 330 million registered domain names - on the internet of 5 billion people! If the 20th century was the car’s age, the 21st is the age of the internet. The exchange of ideas at the speed of light is an enabling technology that will continue to open up new doors in every domain: energy production, healthcare, manufacturing, love, and governance. Everything.  

I like to say that internet software combines the power of machines with the economics of media. I can send you an app at no marginal cost that does the work of what, in the past, would have required a room full of expensive machines. But good software, like good writing or art, is not inevitable. Technology does not arc towards a positive future on its own. The decisions we make in designing software make the difference between it preying on our weakest vulnerabilities and helping us become the best versions of ourselves. A glorious internet is possible.

Last chat :