My chat with Caroline Clark, Cofounder & CEO of Arcade
Making demos interactive and fun
I had started following Caroline on Twitter when she was still in her previous job. I later found out she had left to start a company. I saw a few videos of her product over the past few months. I checked out the website and reached out to learn more. When we connected, Arcade was still in private beta. I had no idea when she was going to launch. Turns out today is launch day for Arcade! We decided to time publishing this piece with her launch. You can check out the company announcement here.
Sar: I must say that the website's fun and playful. The company feels appropriately named. What’s the backstory? You have a strong point of view on how the website should look and feel.
Caroline: When we started Arcade, my co-founder and I asked ourselves, what do we want people to think about Arcade, and how do we want them to feel after they use our product? And fun and playful were two words we decided on. The fact that you can pick up on that from the website is exciting. We want to be something people feel excited to use, and it's accessible to anyone, no matter who you are.
Sar: When I started looking into what you were doing, I wasn’t sure how I felt about what you did. You have a demo of your product on the homepage. It looks like you used Arcade to demo Arcade, which makes sense. Watching that demo made me realize what you did was cool and different, which is great evidence of your product working. You have taken product demos, which are quite passive and boring as a format, and turned them into an active and engaging format. It’s still a video, but it is an interactive viewing experience.
Typically, when watching product demos, it’s easy to gloss over details and get distracted. Based on that demo, you are making snap judgments about how you feel about a new thing. The stakes are high for the people who have made that demo. When I was watching your demo video, there were pointers and descriptions throughout the video, making me want to lean in and try harder to understand what was going on. I was distracted and had my phone with me while watching it on my laptop, but I didn’t miss the details because the video wouldn’t proceed without me clicking on things! And having to take action makes a lot of sense for videos showing people how a new thing works and what it does. Another thing I noticed was all the Arcade demos I have come across are quite short and to the point. Making the demo both actionable and short felt like a great combination. It was one of those experiences where little design details make a lot of difference in experiencing things that seem pretty mundane! Am I correct on how I’m thinking about this at a high level? I have a few other observations, but I'll first let you react.
Caroline: It's precisely how we think about it. I always coach our customers to create crisp and short demos. Don't try to make a documentary. There's something called the rule of seven, which is a psychological study. It’s been proven that our short-term memory can't hold more than seven steps. So it's hard for us to retain things anytime there are more than seven concepts, actions, or directions. We just cannot recall it quickly. So that's what I encourage our customers to do. That's why the phone numbers in the US are seven digits.
Sar: I had no idea about that connection!
Caroline: You are right to point out we're much more engaging because we force viewers to pay attention to what you have to do next. Another aspect of what we do is we make the demos editable. Once you record the demo, you can edit it through our flow builder. We have a Typeform-style paradigm where you can see different steps in moments in time and edit them. You can add, remove or swap them; you can change to order. You can add branches with different directions to take the viewer through. People can see what the difference is when they make changes.
Sar: Right, there's post-production work that comes with Arcade. It isn’t about recording a video 15 times until you get it right. It is about casually recording an experience, and then you guys break it down piece by piece in the flow editor. As the creator, I can engage with every piece and move things around. It's designed to get started with a few clicks and then customize it later. There's less pressure when it comes to recording it. I have used Loom for work. And I have always felt this pressure to get it right in one go, especially with external communications. The image that comes to mind is that of a Hollywood set where the actors are making multiple attempts to get the scene exactly right!
Caroline: You nailed it. We try to lower the pressure as much as possible. One of the key ways that we measure success is to look at how many people publish demos, host demos, and then how many views those demos receive. That's how we build our flywheel. We are a product-led growth company that primarily targets other product-led growth companies. That's why we care about publishing so much. When we think about the steps you need to take to publish a demo, then think about many small details in our product to lower that pressure. We are encouraging people to publish, and we want to make the entire experience feel delightful. This is why we reward users with confetti when they publish a demo. Then we suggest how the users could improve the demo through analytics – both through proactive notifications and our out-of-the-box analytics product – which encourages them to adjust and ultimately re-publish.
Sar: Where do these videos go after you publish them? What are people using Arcade for?
Caroline: Once you publish them, you can send the link directly to a customer with a private URL. If you launched a new feature, you could post videos to explain the features in the blog post. You can also put them on a website. That can be product pages or even the homepage. You can share them in tweets for marketing. People use Arcade to discover what users are engaging with. They make videos for different use cases, and they look at the analytics to see what has the highest engagement. You can track conversion from viewing to actions like new leads or signups.
Sar: Talk more about the analytics you offer.
Caroline: Analytics are designed to inform Arcade creators about which parts of the product experience resonate most with their audience. They can dive into traffic and engagement patterns to A/B test various flows, or CTAs, to help understand user interest. Marketers can use these insights to promote functionality earlier in the buyer’s journey, and product managers can use these insights to help them shape and influence their roadmap.
Sar: You didn’t launch with collaboration built into the product. Can you talk about how you thought about the MVP and your thinking on team collaboration within the product?
Caroline: Our primary goal was to make an intuitive product that allowed users to showcase their products to prospects and customers at every stage of the user journey. While we knew from experience that building product demos internally is usually a team sport - you need help from engineering, product, design, etc. - we wanted to empower the user to do it themselves. We wanted them to be able to create their demos at their own pace. But some early feedback from users is that they wanted an easy way to collaborate on Arcades within their companies, and the reality is that teams are behind every great story. Because of that, adding ‘multiplayer,’ as we call it, was a no-brainer. Now designers can add the latest mocks to an Arcade, product managers can highlight new features, etc. - and all of those updates happen automatically.
Sar: Let’s discuss your journey before founding the company. You worked at Atlassian in product marketing for three years and spent a year in venture capital at Sequoia. And then you left and started the company, right? Did your time in Atlassian inspire this idea?
Caroline: I joined Atlassian in 2015, which was at the company's growth stage. I was one of the first product marketers on a new product called Jira Service Desk. Atlassian, as a company, is extremely conscious of building authentic customer relationships. You have to showcase the product and build trust step by step. And that bleeds into how we build our product.
I will never forget the first day of my job. My manager said, "Oh, you need to go to a conference and sell the product to new customers.” I was completely unqualified. I had no idea what this product was. I'm trying to make a great impression. I tried hard and got 250 signups at the conference. I was excited to tell my manager about the results of the conference. He looked at my sheet and was horrified. He’s like, this is way too many names. And I'm like, what? That’s the opposite reaction of what you would expect. He said there was no way that many people were interested in the product. He was right. We did the campaign to follow up with those people. It was a poor-performing campaign. The lesson that has always stayed with me is that you must earn the customer’s trust. You have to earn the right for them to want to be able to understand what a product does to sign up for it. This culture at Atlassian – of earning customers’ trust every step of the way – is what inspired me to start Arcade. We started with interactive demos, but long term, it’s about supporting the customer journey.
Sar: What tools were you using for these product demos at the time?
Caroline: Screenshots and gifs! One time, I did an experiment where I had an engineer build an interactive visual for a blog post. It took about two months. It was shipped, and we had great engagement, but it was such a painful process that we were like, okay, let's not do this again.
Sar: You spent a year in Sydney working at Atlassian. What do you miss about living and working there? As a New Yorker, what surprised you the most about the culture? Any observations about the startup folks in that ecosystem?
Caroline: I loved Sydney. It’s truly a great place to live. It’s beautiful. Australia reminds me a lot of London culturally. When you walk around and meet people, it’s clear it’s a Commonwealth country.
My very subjective observation about Sydney over a place like New York City is that it felt less classist and had a larger middle class. But in places like New York City (and SF), there seems to be a more extreme disparity between the poor and the upper class. Many things drive this, but one of them is Australia’s education system. Instead of flying across the country to go to college, it was expected to go to the local university – subsidised. The quality of education across Australia is way less stark. That means that if you look closely at people’s friend groups, there was more variance in their backgrounds. It was normal for a wealthy Australian to be close friends with someone who grew up with fewer means because they met at University. It’s certainly possible for this to happen in the US, but it’s more of a novelty. What does that mean culturally? There’s more of a meritocratic approach to software. I would even argue that Atlassian’s product-led approach is almost a function of their Australian identity.
Sar: That’s fascinating. After your time at Atlassian, how did you end up at Sequoia? Why take that detour before starting Arcade?
Caroline: I went to business school at Stanford. Alfred Lin reached out to me with a cold email. I didn’t know much about venture capital. It seemed like a really interesting opportunity, so I talked to Alfred, and the rest is history. I joined the firm as Chief of Staff for Bryan Schreier. Sequoia’s rigor, discipline, and focus on quality were unlike anywhere else I've experienced. There are a lot of cultural norms at Sequoia that I've adopted at Arcade. For example, I try to instill a writing culture. When COVID hit, I decided I wanted to build something, and they were supportive.
Sar: You mentioned how painful it was to create engaging demos at Atlassian earlier. We discussed how Arcade helps reduce friction and the pressure in making demo videos. Do you hope companies and teams that previously didn’t rely much on demo videos would now consider doing it because Arcade makes it easier? What kinds of companies are using Arcade?
Caroline: We’re seeing companies use Arcades to support a variety of use cases - anything from early prospect education to training staff. For example, some teams at Carta use Arcade across their support documentation, social media campaigns, and websites. We have users at Sentry who incorporate Arcades into blog posts to communicate product updates. There’s a beekeeping app, Nektar, that uses Arcades in the field with commercial beekeepers to educate them on how to use their software. So yes, I think we’ll start seeing more companies rely on demos to educate users on the best way to find value from their products.
Sar: What low convictions bets or decisions paid off big time?
Caroline: We had a really interesting experience after we launched in January. We learned that there were a few big gaps in the product, one of which was that users were confused by the video component and the multi-select options within our chrome extension. They didn’t know whether to click on “Take screenshot,” “Record video” or “Record clicks.”
The journey with updating the chrome extension was a little winding. At first, we decided just to delete the “Record Video” button. We looked at the data – about 20% of users used it, which we deemed not high. We thought that would be the simple answer to our problems. But then users got very upset. It turns out that many of them were using video with us. We got a flood of support requests and emails.
So we decided to bring it back. We realised that by removing something, we learned how important it was. We also failed to realise that 20% were using it frequently.
We then built on top of the Mux API to generate video automatically with dragging and dropping, typing, and scrolling. It simplified our experience a ton and incorporated video within the core experience instead of making it a standalone part. Essentially, we went from pushing users down a multi-select journey to a single button that, behind the scenes, is a powerful recording of all user actions.
Sar: What product choices were you nervous about initially that have played out quite well?
Caroline: When we started, we were nervous about relying on screenshots and videos instead of something more “sophisticated,” like front-end components to represent an application. We looked into Storybook and many other applications but realized that every solution with super cool output was extremely challenging for an average user to create. It wasn’t accessible for a marketer, product manager, or customer success person.
Now that we’re 1k companies in, what has shocked us is we have heard from an absolute total of two people about needing something more interactive. Meanwhile, we have a 7-minute median time to publish across our user base. The big lesson here is that always, ALWAYS prioritize the product experience to accomplish a result and focus on your end user.
Sar: Talk to me about Arcade’s future.
Caroline: It’s getting easier and easier to build new software these days, as Vercel, Tailwind, and AWS all exist now. You have to have high-quality software to stand out and win. You have to experiment constantly. You need to validate it as soon as you have a design. Currently, customers use Arcade to demo a product they have already built. The bigger vision is to own the entire product life cycle. When you are just starting and haven’t coded, we want you to use us to validate your ideas. When you think of the word demo, you think of sales typically. But, in the earliest days, it is not just about sales; it is about validating whether you should even build a product. As you grow and the product matures, we want you to use us to decide what to build next.
We've been very focused on having a self-service product. It will be interesting to see where things evolve in the next few years, whether a volume-based product-led growth approach work or is it all about top-down enterprise sales, which many of our competitors are doing. Alfred Lin always says to be competitor-aware but not competitor obsessed.
Sar: So you want to enter prototyping territory? Right now, you have to have a product with a website and install the Arcade extension in your browser and take a video to demo your product. The stage you're talking about precedes that. This means you don't even have anything.
Caroline: Yeah, some of it is prototyping. But, it can go earlier than that. Think about a napkin drawing to show someone what you have in mind. So you can import any assets like screenshots and pictures and start creating workflows using our editor.
Sar: My last question is what’s been most challenging about building a company?
Caroline: Before I made a full-time jump to Arcade, I was concerned about health insurance. Honestly, that is a super easy problem (we went with Rippling).
The hardest thing about being a founder is the gray area. I'll give an example of a gray area, which I haven't specifically handled, but this is along the lines of what will come up a lot. Let's say you have an employee who is 11 months into the job, and you do not think they are a great fit for the next stage after they have been at the company for a while. Do you let them go even though they have contributed a lot for that year, or do you let them vest an additional month? Letting them vest means those shares are unavailable for a potential future employee. To be clear, I would let the person vest. It's perfectly acceptable and legal not to let the employee vest, and many founders opt not to do it. There are many of these emotional issues with no straightforward answers.
I want to have a reputation as someone fair and honest, with high integrity, and who does right by the people I work with. And that includes my customers, my investors, and my team members. Sometimes you can't please everyone. You need to have a point of view and principles to draw from. That's what keeps me up at night, and that's a complex part of building a company.
Chat from earlier this week :