Cofounder, Dialpad
Alex Cornell - How to present your designs and get buy-in for your ideas (Dive Club Ep. 58)
When you think about storytelling in design Alex Cornell is often the person that comes to mind. And that's a big reason why he's one of the most requested guests on the show. So this episode is a deep dive into the finer details of communication. We talk about:
The story behind his startup Cocoon
How Alex leverages his background in video
Alex’s precision and obsession with language
Why Alex left linear to work on generative AI at Meta
Why getting buy-in for your ideas is kind of like a math proof
Behind-the-scenes of the wild videos Alex made at Facebook
Lessons learned designing the Substa...
About Alex Cornell
In a June 2024 appearance on the Dive Club podcast, Alex Cornell discussed his approach to design communication and his career moves. Cornell described his view that successful work involves spending significant time on how an idea is delivered, so that audiences recognize the idea's power without noticing the delivery mechanics. He compared gaining buy-in for design ideas to a math proof, recommending that presenters start with broad, uncontroversial premises and layer new information one step at a time. He also advised against showing many design options without context, instead suggesting that designers establish the framework behind their decisions and present a scoped set of prototypes with clear trade-offs. Cornell discussed his startup Cocoon, which initially launched as a single-group, no-notifications app—a constraint he called a valuable learning experience before the app later added multiple groups. He noted that TikTok helped him communicate Cocoon's value in a way he had previously struggled to articulate. Cornell also addressed his return to Meta to work on generative AI, saying he reflected on his favorite professional environment and realized that working with the people and project focus at Facebook/Meta was where he learned and enjoyed himself most. He expressed a preference for hybrid over fully remote work, stating that if he were to start a company it would be hybrid because of the value of in-person collaboration.
Transcript (56 segments)
N
Narrator0:11
Websites in Framer today, all you need is the Figma to Framer plugin. Just select the layers you want to copy, run the plugin, paste into Framer, and then you are ready to publish it directly to the internet. It feels a little bit like magic, and you can experience it today. All you have to do is head to dive.club/slf/framer to get started. Okay, now on to the episode.
A
Alex Cornell0:33
It's all communication, but there's so many hidden knobs that you can tweak, and most people either don't know they're there, or they don't really care, or they don't think they're that useful. I think that's sort of how I view work as successful these days: I'll spend all my time thinking about the idea, and then eventually a lot more time on how it gets to you.
H
Host1:10
Welcome to Dive Club. My name is Rid, and this is where designers never stop learning. When you think about storytelling and design, Alex Cornell is often the first name that comes to mind, and that's a really big reason why he was one of the most requested guests on this show. So this episode is a masterclass in the finer details of communication. We get into the backstory of his startup Cocoon, and he even shares some lessons learned while designing the Substack and Linear mobile apps. But first, let's hear more about Alex's background in video, because it plays a major role in who he is as a designer today.
A
Alex Cornell1:48
You get your first video camera and take it on vacation and see what you can make with it. It's pretty exciting. So my first job, you could tell other people how to use this thing, and since nobody knew how to use any of the technology in this room, nobody would ever visit the media lab. So it would just be me there all day with these very powerful computers that were there theoretically for students to come in and use, but it was just basically me in there the whole time, which was amazing. I think that laid the technical groundwork for understanding how to use video at all, which translated to YouTube when it became popular, I think it was like 2006 or 7. I was really into music as well, so combining those two things—making videos and making music—I spent a lot of time doing that until maybe 2010 or so, and again just honed the technical side of it. I think that eventually translated into a more...
H
Host3:11
You have this really fun chunk of time on your LinkedIn where you're making videos for startups specifically. Can you talk about that a bit?
A
Alex Cornell3:20
My first company was called UberConference, which is now called Dialpad. When we started, we were kind of trying a few different apps, and as I was making videos for those to promote them—because at that time, it's hard to remember, but back then videos were sort of the way that you could get your startup noticed. It wasn't just a nice to have; it was like, gosh, if you had a video and it was kind of good, it could really blow out your initial audience in ways that aren't possible anymore. So it was worth investing some time in.
I was friends with Adam of Sandwich Video, and he had a business—probably the ultimate business for this. Still to this day they do incredible work. I kind of thought if I could just do a fifteenth of what he's doing, that would be really fun and really lucrative, because people were paying hundreds of thousands of dollars for these things because nobody knew how to do it. So it's like, if we're going to get one, we have to pay somebody to do it. I think all of that led into me eventually deciding to leave the design side of the startup world and attempt to create a commercial business, making commercials for other companies, which was very fun.
There are a few examples of these places where you have a hornet's nest worth of people that somehow all collaborate in a way that is really efficient, which is pretty unusual. So I love that about it. But the thing that I guess I should have known, because I've done client work before, is that video work is sort of the ultimate client work. You go on a day shoot, you come back, you send the client what you have, and they're like, 'Oh, we like it, but maybe not that location.' You think, okay, well we can change that, but it's just going to require the machine to start up again completely, going all the way back to the beginning. That part of it started to become pretty tiresome.
I think there was maybe two years specifically honed there, but there was one missing piece to that part of what I was doing, which was the writing side of it. When you're writing scripts for companies and trying to take on their vocabulary and their tone, it forces you to learn to write in a very different way than you maybe would have before. So while I was enjoying the video side, I think I learned to express myself verbally in that period of time way more than I had before. That skill is sort of the ultimate foundation to pretty much everything I have done after that. A lot of times people point to video because it's the shiny object that you see, but when I see the work and when I see it resonate with...
H
Host7:11
I hear you talk about storytelling, right? They've seen your videos, and this is something that you've almost become synonymous with in some ways. So maybe we could use this as an opportunity to go a little bit deeper into this connection between your movie background and what you're doing more day-to-day as an interface designer. You talked a little bit about the writing, but what are the ways that this background in video and script writing has impacted the way that you approach storytelling as a designer?
A
Alex Cornell7:41
The fundamental that's the same is you're trying to get an idea from your head to somebody else's. That's it. If you're designing a logo or a menu or you're making a movie, at the core of it...
That's why I've become so obsessive over it. I think something that people are usually somewhat astounded by, and I think a little bit weirded out by, is that I often will meet with people to discuss this and I'll tell them how I work. It's a level of precision and obsession of language that I think would seem totally over the top, but I think that's why it can be so effective. As an example, if I were to sit down to write something, let's say I was going to make an argument for why using a guitar pick is better than fingerpicking, and I'll put that in a video. The very first word that I say in that video, I will spend a crazy amount of time thinking about.
Ideally, it would start with 'guitar' or some other word that's much more immediately interesting. Then once I have that sentence, I'll spend probably an hour figuring out all the different ways I could read it. So let's say the sentence is 'Learning to play the guitar is the most effective.' You could say it like that, or you could say it like this, or you could say it in a way that feels like it's going to lead to another sentence. There are a thousand different ways you can say each word and intonate your voice all over the place. When I'm saying the sentence, I can feel all the little knobs that I'm turning to make it ride one way or another.
What will happen is you'll hit the verse, the first couple phrases will sound great, and then you'll go out of key. But you're like, 'Man, I really love the way I came in on that.' So you'll freeze that and then you'll punch in, as they say, to get the next part. It's not that uncommon for a vocal take to be a composite of all these different takes. It's very common in music, it's also common in voiceover, but it's not super common in, 'I'm just going to make a video for my team today and send it to them this afternoon.' If people knew the crazy things that I was doing before I sent that video, they would think that I was either totally insane or wasting time. But I will composite voiceover takes to send a video.
All of this is just communication, but there's so many hidden knobs that you can tweak, and most people either don't know they're there, or they don't really care, or they don't think they're that useful. At the end of it, I look at these things and love them for their efficiency and their clarity, and then also enjoy that most people don't even care about that stuff. To them, ideally, all they care about is the quality of the idea that is conveyed. I think that's how I view work as successful these days: I'll spend all my time thinking about the idea, and then eventually a lot more time on how it gets to you. Ideally, people will only recognize the power of the idea and they won't notice the stuff.
H
Host12:10
You need to set the context effectively, outline the problem, make them feel the pain that the user's feeling. Voice inflection, I love it.
A
Alex Cornell12:23
There is an underlying assumption that the idea is of good quality and that you picked a framing that will be effective as well. I definitely consider these things as somewhat different modes that you'll go into. If the beginning is very messy, exploring the space, figuring out what you want to say and how you're going to say it, I don't really shift into execution mode until that's taken care of. I much prefer—maybe it's just because I spent so much time in a technical setting—I love when an idea forms solid enough...
In the other to really work through those problems, it's a lot more loose exploration. I think maybe the reason I care about it so much is I've just observed in my time that you can—it's hard to say it without making it sound manipulative, but it is a little bit like manipulative has negative connotations, but if we just take the literal definition of the word 'manipulate,' I do think you can kind of manipulate people into seeing a thing a different way via the delivery of the idea. Not to say if the idea is bad they'll think it's good—that would still be a failure state. It's more like they'll see things about it that they might not have, and help them feel like it's a good idea. Just the energy transfer that would happen in something as small as a Loom video: if people can feel that you're excited...
Back then, that was not necessary because we were all on site all the time. I could see the entire video team at any given time, which is kind of wild to think about now. So making a video was purely additive to our comms channels. But then during the pandemic, that became the only way to communicate because your teams are distributed. Now you're async, you still have a lot you need to say and a lot you need to show, so tools like that became absolutely required to communicate certain ideas. I was really happy about that because I was like, 'Wow, this is the thing I was doing already, and now it's a thing we kind of have to do.' But the trick is, all the same rules still applied. For example, if you made a video that was over, I don't know, 2 minutes long, you can look at...
The point is that all the same rules around posting to YouTube and making it engaging and thinking about how you're going to hold people's attention apply in corporate communications as well. I think there's a mistake and assumption that people will read everything you write and watch everything you post, and that's absolutely not the case. So you have to kind of work against it.
H
Host15:34
You mentioned the videos for Facebook. I recently shared the three-dot chevron newscast that you made in the newsletter, and people loved it. So maybe you could talk a little bit about why you think that video stood out, but then also how often are you doing this kind of thing? Are there other related examples that you can share?
A
Alex Cornell15:53
Very often. I think the icon for 'more actions' should be different. I don't think it should be a chevron, I think it should be a three-dot. There are a thousand ways you could attempt to make that change at a company. You could write something, you could say something, there are all kinds of ways. Maybe you could say, 'Well, one way would be to make a video,' and you just make a video and you talk about it. In my mind, I'm always thinking about how to take it not like one or two turns further, but like seven turns further. At a certain point, once you take it that much further and make it weirder or more intricate, the fact that you did that almost becomes itself an object that people react to. So even if we're not looking at the ideas in the video...
Then cut to that other camera, and it looks like that guy's turning to his right. It's all done live. The diagram for that video was so crazy because I remember thinking about it on the bus: 'If I turn to my left, though, and there are three cameras, it's going to look like this guy is turning to his left and the other guy is turning to his left. That's not going to work.' So they have to be mirrored. I was drawing it out, and just the fact that that happened is enough for people to be like, 'What the hell is going on? Why would anybody do that?' So that becomes its own little thing. I think sometimes that's the key that lets people in. They're like, 'Well, this is weird as hell, I'll maybe listen to it for a second.' And then of course it's important that the ideas being communicated, once you've got people, are good. I spent a lot of time writing that.
I'm reading something, but you can't tell that I'm reading it. I think that's why it resonated. There are plenty others. One that I can share because it specifically didn't have any real details in it on purpose: I created a fake feature, a hypothetical feature, and sort of filmed a thing which presumed it had been shipped. As a result of it being shipped, there was some controversy, and different people felt differently about it. The video was people arguing about whether or not it should stay that way. The nice thing about it was it allowed me to say everything that I wanted to without incriminating anybody or making anybody feel bad, because I'm just talking about this thing I made up. The thing doesn't even exist.
My favorite part of it: I isolated myself with the same perspective and everything, then cut a keynote, isolated that, and replaced it on the video. Then I composited all that together so it looks like I'm talking to the audience with this deck on the screen. I could intercut between me and the keynote. I would often just cut to the keynote because I have full control over that. My audience loop was usually 15 seconds long, so I couldn't stay on that shot for very long. Then I had a close-up of myself as well. If you just looked at that in the newsfeed, it looks like I'm presenting to an all-hands audience. That was enough for people to be like, 'What the hell is going on?'
There was a fake promo video for a thing I thought we should do in live, and I just posted it to my profile. I got a message from legal almost immediately: 'We haven't seen this. You can't ship things publicly without going through all these processes.' I was basically getting reprimanded for this thing. I had to write back and say, 'Oh no, nothing about this exists. Not even the app icon. Not a thing about this is real.'
H
Host20:45
I saw some questions from Twitter on this topic, and one that I really liked was from Kyle Barber. He said, 'What does your process look like when you have a product or a demo that you're getting ready to tell a story?'
A
Alex Cornell21:11
Just an upside-down triangle. I think whenever you're shifting from having made an app to explaining it, it's really important to start with a fairly broad message that's relatively not that controversial, that people agree with. That's the math proof element: if I want to convince you that C equals D, I would start with something like, 'Do we agree that A equals A?' You'd be like, 'Yeah, obviously.' Then, 'Okay, great, A equals A, so 2A equals 2A.' You just layer one new thing at a time to eventually get to where...
That's crazy because they're already kind of with you. The alternative approach, which sometimes can work, is to just start with that and be like, 'Here's a surprising idea you've never thought of before.' That can send people one way or the other. They could be like, 'Whoa, that's really crazy, I love this idea,' or they could immediately be like, 'Wait, what? C doesn't equal D. That doesn't make sense.' So I think a lot about what style is going to work depending on what it is. The hardest I ever felt this was when we worked on Cocoon, which was the company my friend Sain and I started after working at Facebook the first time. That was probably an app that had the closest overlap to my skill set.
It was a problem that I really enjoyed our solution for and really wanted other people to experience. When it came time to communicate that to people via video or whatever, that was by far the hardest I ever felt it. Maybe it was so close to me or I cared so much about it that every idea I had of how to communicate it, I would shoot down for one reason or another. That was an interesting experience because I had never really been stumped in that way or for so long. Funnily enough, the thing that broke me out of it was TikTok at the time. I know it's thou-shalt-not-be-named anymore, but at the time...
The floodgates were open. Before that, one of our early employees had told us, 'Guys, I'm messing around with TikTok and it's crazy.' So we were like, 'Alright, let's give it a go.' I made maybe six or seven TikToks about Cocoon that opened up a whole new world for me of how to communicate the ideas I wanted to. It was amazing. I hadn't had that experience in a really long time. I don't know what it was about the media; maybe it was just that it was sillier. I think that was probably what it was—just less pressure. It was pretty weird because it was definitely not my normal style. I'll sometimes post them because they still make me laugh. What I really enjoyed about them was that I...
'What this app does, would you like to use it?' That was not the right delivery.
H
Host25:21
Can you give a little bit of background as to what Cocoon was for people who are not familiar? Also, it's clear hearing you talk that you cared deeply about it, so maybe you could even touch on why was this such a big deal for you and ultimately how'd you get to the point where you were like, 'Yeah, I'm going to leave Facebook to start working on this.'
A
Alex Cornell25:39
We actually left without an idea. Sain and I had become friends working together at Facebook, which is an experience I hope for anyone in the field to find a working partner that inspires them. When we started working together, it felt...
We knew what we were thinking about but we didn't have an idea. Then before they looked at the application, we thought of the idea. I remember we reapplied almost before they looked at it, which was a good forcing function. In that process, we would spend a lot of time talking all day, like six hours of talking, about everything—all of our experiences and what we care about. We came to a similar place: we felt kind of distant from our family in particular, and friends also, but mostly our family, despite all these advances in communications and Zoom and iMessage and everything. That insight eventually led to Cocoon: why is it that our most important relationships are...
The same style of message is delivered in the same zone. On one hand, it's like, 'Wow, cool, what a versatile tool. It can handle all these things, and nobody wants multiple apps. We want everything in one place.' Totally get it. But also, there are a lot of very specific things about talking to your family that are different from talking to your DoorDash or whatever. So we started from a place of, 'Okay, let's build an app for you and your family.' The core insight was that we pivoted very quickly to allow you to have more than just one group, because we started in a somewhat crazy place: you could only have one group and there were no notifications. If you want to set your app up to not grow, that is a great way to do it. It was actually a really good learning because we felt...
All of our early users were like, 'If there's anything I want a notification about, it's when this group messages me. I don't want all the other crappy notifications, but I want these.' That was very obvious in retrospect, but it was a good reminder not to get too caught up in the nonsense noise out there, because I think there's always nuance. We eventually opened it up to be more than one group, and it worked a bit more like a traditional messaging app. We had a little lull where we almost started to replicate a lot of the features that existed in other apps, and it was less clear why you would use ours. Where we ended right before we got acquired was the product I wish we'd launched with, because we had gotten to a place where...
There are tons of examples, but it was a really fun product to work on. I really did enjoy it. I actually worked with the same iOS engineer on that app as I did at Substack, because we got acquired, and then we also ended up at Linear together. We worked on three apps together in five years or something. It's hard to give advice because sometimes it's like, 'My advice would be to find that person in your career,' but you can't really manufacture that. When you have that kind of relationship, especially as a designer finding an engineer who you trust, who can execute your designs in the spirit and maintain the intent to where there's almost no lines...
N
Narrator30:11
Framer, I made the switch recently and I'm never going back. Now, when it's time to animate my designs and showcase my work, that's when I'm going to use Jitter. It's like Figma only for motion design. Every step of the way, I'm going to be using Raycast constantly. It's the cornerstone of how I use my computer and the way that I stay in flow while I work. Now, I hand-selected Framer, Raycast, and Jitter to partner with me so that I could do these interviews full-time. If you want to show them some love, it would mean a lot. Okay, now on to the rest of the episode.
H
Host30:44
Let's move forward in the story a little bit. So Substack acquires Cocoon, you're now tasked with taking something that's working on web and going kind of zero to one on their new mobile apps. One of the things that I...
A
Alex Cornell31:11
It was a unique experience for me from the get-go to take a product that has fit in the market on desktop email and translate it to a mobile context. Before that, I'd almost always been trying to find product-market fit. With Cocoon, we were iterating to get to that state; we weren't translating a use case to a different form factor. I really enjoyed the specific nature of the Substack app: it's a reading app. You're focused quite a lot on people's experience reading on the phone. That hadn't really been something I'd thought about a ton before. In Cocoon, we did use a custom typeface, but the reason we were doing that was mostly for character...
We accepted it because it felt just as legible. But in Substack, it was the first time in a long time—I went to graphic design school—that I hadn't really thought about typography. How many characters are on this line? What's the line height? What's the x-height? Are we worried at all about these serifs being a little bit less legible when the contrast is high? All that stuff brought a lot of my early career back flooding to the forefront, which I love. It's actually really depressing to me where we are as an industry now, where we pretty much mostly use system fonts. When I say 'we,' I mean most of the apps I use. On one hand, that's amazing because I love...
Point types. I always use 16, 14, sometimes 12 or 11 maybe if it's all caps. I don't think about it: 16 great, 22 or 24 line height, and I'm off and running. I know all my colors, I have all my hex codes memorized. I don't decide on these things anymore. The weird thing is, when you start out as a designer, especially in traditional graphic design, you probably spend 60% of a project worried about the specifics of the typography: what's the typeface, can we get it licensed, we bought one variant, we're going to use this one variant. You're dialing in things to such a level of precision that you...
For a while, I think one thing I found when I was selecting the typefaces was that depending on who you talk to, there are a lot of feelings around how objectively good or bad certain typefaces are for reading. That's true: there are things that affect legibility, and there are lots of little facets that can be turned to push that in either direction. But I do feel like at a certain point, especially with modern typefaces, the decision is really a feeling decision. There's a set of typefaces that have all achieved certain objective things, and now you're just picking what feels good. I love how hard that is to articulate. Why do I like, let's say, my favorite monospace typeface is JetBrains Mono?
I just prefer it in a way that the only analogy I can draw is maybe if you're trying to argue why something tastes better or not. There's a baseline of 'this food is edible and it's healthy,' but now all this food on the table is edible and vaguely healthy. If we're going to argue about why you like certain things better, you're not really going to win an argument. I love bell peppers, my wife hates them. I will never convince her that they're good. It's just the way it is. I probably would be that way with somebody suggesting that Tinos is the best serif typeface. Other people might be like, 'You're absolutely wrong. I like New York,' which nobody would say, but maybe somebody would. New York is a great example.
New York is in the subset cap, but so is Tinos. When my app would sometimes flip back, they're close enough that you wouldn't notice right away, but after reading two sentences I would be like, 'What is going on?' I cannot describe it. If forced, I could give all the technical reasons, but those aren't the reasons why I don't like it. Anyway, I love that experience. I love putting in the little type selector and making that a big part of the app and putting a lot of care into that. I also really like doing it the proper way, licensing everything, type foundries and the whole thing. That's one of my favorite areas of the field to support, because I just can't imagine doing that executing...
H
Host37:11
This was kind of your first time working from a place of established product-market fit and then translating it to a new medium. It's kind of interesting: you went from doing that for the first time to then doing it immediately afterward with Linear. So are there common threads between those two experiences that would shine a light on how you approach your practice as a designer?
A
Alex Cornell37:33
Yeah, but I'd have to dig for them because they're such different products and such different teams that the experience felt dramatically different. The thing that was the same, I think, is that there's a really weird instinct that designers in particular have, which is to look at a problem space that's been explored already. That would fit the category in both of these, where solutions have been established, and I...
I'm like, 'Why are we even doing this? Should we step back and figure out why we're doing this?' It's very tempting, and I understand that. I felt it too in both times, where I would look at the architecture that I'm moving over to mobile. In Linear's case, for example, there were a million iterations of it, which I hope to one day be able to share. I was just caught up in, 'What if it worked a different way? What if it was organized a slightly different way?' It was funny because for somebody who is so allergic to that instinct, when I experience it directed at me, it was kind of weird to do it myself.
See if that makes sense. I think it was probably one of my favorite experiences in being allowed to explore. It's actually what every designer asks for: 'What if I could work somewhere where the CEO and leadership cares deeply about design, they let me do whatever I want, they let me work with the people I want to work with, and the product is cool and interesting and profitable and an incredible product?' You could not draw up a more perfect situation. It was really interesting for me because, similar to what I was describing before with Cocoon, it was hard for me in that perfect scenario to figure out how to articulate some of the ideas that usually come so naturally. It was similarly difficult at Linear to decide on...
It was a very interesting experience: having all the circumstances be totally ideal and then finding that it's not so much that my skill set wasn't there—I felt like it was—it was more like I almost had too many ideas and it was hard to pick one.
H
Host40:28
Can you talk a little bit about how you're bringing people along for the journey while you're navigating that idea maze? At some point, you do have to check off some alignment boxes to ensure that you're moving forward towards shipping something. So what did the process look like outside of just having this freedom to explore like crazy?
A
Alex Cornell40:48
Linear is fully remote, so it was a lot of asynchronous video communication. I...
Just like, 'This needs to be a video because you need to see my screen.' That's it. So I did a lot of that. One thing that I would try to do there that I think is useful anywhere is not just display the prototype and the idea and show a thousand iterations, because I think sometimes that can overwhelm people. 'It could work like this, it could work like this, it could work like this.' There are a hundred options, and people don't know when you came up with those options, what you had in your mind, what you were trading off, what framework was guiding you. The differences are just going to be visual at first. When they see the different things, they're going to be like, 'Okay, well that one the menu is on the top, that one it's on the bottom, that one there is no menu.' So I think it's very important to establish and make very clear the...
Understand what is distinct about this idea, why it is even being included in this combination, and where it sits on this little map that I'm going to draw for you. So now you know all the things that I was trying to balance as well. I think that can be very helpful. I just did it yesterday, actually. Communicating something where I get frustrated how often I feel like I'm describing option one versus option two versus option three. I'm always like, 'This or that, this or that, this or that.' But sometimes it can be useful to kind of whittle something down to a scoped decision that I can fully describe all the background for. As a result of all that background, you could kind of either go this way or that way, and here are the reasons why you might go one way or the other.
Now you're referring to these things as named entities. Sometimes it's even worth giving them a name: 'This is the no-menu option, this is the gestural option.' I found that to be pretty helpful, especially in that environment when you're remote. If you're just sharing a whole bunch of work and saying, 'What do you guys think of this?' without making it clear what you were really toiling over, you're going to get feedback all over the place. I like the idea of thinking of it less as options and more as directions. If you put things on a spectrum, you're not just picking option A; you're picking this chunk of the spectrum, and this is the axis. It reminds me of something that Kevin Wong said to me—he's the head of product design at Webflow—and he...