Pod Insights a16z Podcast a16z Podcast | Culture and/of Design About this episode "Mobile-first" (and now too AI-first) has been a mantra of sorts in design, but what does that mean at a company, product management, and competitive level? Especially when someone in company X will always say "we should do what Y did" -- even if they have no idea let alone data why Y did it.
And while designing for screens is "like growing a carp in a bathtub" (will inevitably grow to the size of the container), what do design constraints mean in an increasingly screen-less world -- one where everything will eventually become an input ... and even an output? What does it mean to design for a mobile world where "an app isn't really an app" -- and the very definition of apps are themselves evolving, including cross-culturally?
From the age-old question of whether there are design universals to the age-old dynamic of bundling/unbundling, the guests on this episode of the a16z Podcast -- Luke Wroblewski and a16z's Connie Chan (in conversation with Sonal Chokshi) riff, hallway style, on all things design in practice. And on why startups may have the ultimate design advantage.
Transcript Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:00 Sonal? Hi, everyone. Welcome to the A6 and Z Podcast. I'm here today with Luke Roblousky. He's currently a product director at Google, which acquired one of his startups. Previously, he had another startup acquired by Twitter. He worked at Yahoo. He's had quite a diverse background and basically is an expert on all things design for mobile.
0:16 Speaker 01 I'm gonna try to live up to that, but we'll see.
0:19 Sonal? Joining him for this conversation today is our partner Connie Chan, who is an expert on many things, but especially covers China. So I thought today we could actually just launch right in and talk about what it means to have this concept of mobile-first design, which I think was the title of your book.
0:35 Speaker 01 Yeah, so just a little background on where that came from. I was at Yahoo at the time. And Flickr put out this iPhone-only website, because back then you couldn't actually make native apps for the iPhone. You had to make these web sort of pseudo apps. And what I was struck by was like how boiled down to the essence that product experience was. If you looked at Flickr on the desktop, there was like four dropdown menus, each with like 20 items in there. And then there was like submenus. We're talking about like 60-something things. And when it came time to build the mobile experience for Flickr, we got it down to six things. And it was like the things that people do the most. And this is what the core essence of the product was. And it stuck in my head that the small screen and the constraints really forced you to focus.
1:20 Speaker 01 And at the same time, we had stuff where we could use your location. Whereas before, we knew with 99% accuracy, you were in the US. Yay. Now, we knew down to like 50 meters where you were. So there was all this new opportunity. And that got me thinking, these constraints that focus you and these opportunities that open up new things you can do. You got to a very different place when you did that than if you started with the desktop website, traditional view of things.
1:43 Sonal? It's almost the opposite of what is playing out in China. Because when you describe and show some of these apps, Connie, it's almost instead of stripping it down to the basics, it's almost crowding everything possible into a small screen.
1:53 Connie Chan? China has been mobile-first for several years now. And because that way of thinking is so prevalent, a lot of companies will start out with only a mobile app. They don't even have a desktop version. And so all the features and functionality for that company need to be fully surfaceable in the mobile version.
2:10 Speaker 01 The way I think about it on the desktop laptop kind of world. Is you have this big screen that you can fill with anything you want. And as a result, people fill it. It's like the carp that's in the bathtub. You know, you keep feeding it. It will grow to the size of the bathtub. And there's just this natural tendency.
2:25 Sonal? It's like a vacuum of space that just, you know, you must, it must be filled.
2:29 Speaker 01 But it also makes, uh, making decisions really easy because you actually don't have to make a decision. So like, where does Bob's marketing feature go? Oh, I just cram it in the bottom left corner of that site. And there's, you know, 50 million things on it already. Whereas with a mobile device, like there's no room for Bob's feature. Talking about the Google Photos app, there was these things that were out of sight and out of mind for people. And when we made those features visible and obvious to people, all of a sudden, usage of that skyrocketed.
2:56 Sonal? Can you describe that, actually? Just kind of walk us through visually?
2:58 Speaker 01 Yeah, so, uh, we used to just have the stream view of your photos. And, uh, that was all that was on the screen. And there's like the little three-line hamburger menu and everything is there. Right? And so what we did is we put a little bottom bar of navigation, these like tab bars on, uh, both platforms, Android and iOS, that bring the core features of the app to the surface.
3:17 Sonal? Wouldn't that be like an obvious thing to do?
3:19 Speaker 01 Design is always obvious in retrospect. And, uh, you know, like, remember the lawsuit Samsung had against Apple about the iPhone? Like, oh, well, the design that's a black rectangle is so obvious. Well, if it's so damn obvious, why did nobody do it?
3:33 Sonal? That's true.
3:34 Speaker 01 Before then, right? But once somebody puts in all the hard work and all the thinking and applies all the processes and they come up with something, if it's good, yeah, it feels, oh, yeah. That's the way it should have always been. I think that's the part that gets under-recognized in design is how much happens under the covers to get you to quote-unquote obviousness.
3:53 Connie Chan? Yeah, I would use a twist on the word obvious, which is more that it's intuitive. And a lot of people criticize Chinese apps for looking too cluttered. I don't think they're cluttered. I think they're dense. They're very dense. There's a lot of information above the fold for any app that you open that's designed in China. But I think of it as higher ROI because I get more data, I get more information without having to... Skip over to another menu or without having to scroll down. So simplicity is all about being intuitive, not necessarily being minimal.
4:20 Speaker 01 So let me push back on the word intuitive just for a second. Because I don't believe there's anything intuitive in the world. Everything stems from an understanding of the world around you and you build up knowledge and then you apply it. If you look at a lot of the environments that people live in in Asia, they are naturally very dense from an information standpoint, period, right? So there's just a more familiarity to it. There's an author by the name of Edward Tufte. He used to, in his seminars, gauge the quality of a website by counting how many links he could find on the screen. And if there is more links, that is a better interface because the information density is much higher. It's very interesting to see now Apple going to a much lesser density in terms of their design.
5:00 Speaker 01 If you've seen like the new music app, like there's two boxes on the entire six inch screen. So you know, the pendulum has swung a little too far over on that front. But I think there's a balance between those two things. And it's not that things are cluttered. If the things all make sense to you, like when you're scrolling through an Instagram feed, they're all photos. You can just scroll forever. And people, no one scrolls, no one goes below the fold, get all, no, that's BS. If you have an understood context, people will keep going. So in the case of these Chinese websites, if there's an understood context, like all of these are shopping links, all of these are whatever, then the density can actually become a good thing. When everything behaves differently and there's 15 million things that function in different ways, all crammed together,
5:45 Speaker 01 that's where the density starts to become a bad thing. Because people can't wrap their head around, what does this do? What does that do? Or was this different than this one?
5:51 Connie Chan? The users can guess how to find this particular feature. And the fact that the user can figure that out without having to look it up, I think, is what makes great design in Asia.
6:00 Speaker 01 Yeah, it's interesting. We had, I used to work at eBay before I worked at Yahoo. This is like 2002 to 2004. And we made a big push to China at the time. And we had all these conversations of, oh, well, this doesn't work in China. And I remember the local managers arguing with me, "Top-aligned labels in web forums don't work in China. In China, users want labels left-aligned." So we tested it. We put the top-aligned labels above the form fields, and we had a 20% increase in conversions, or something like that. I'm always biased to, "Well, let's put it to the test, and let's compare and contrast what actually works or what doesn't."
6:38 Connie Chan? I think that it sometimes works in that case, but sometimes the design actually is different of what people are used to in China.
6:45 Sonal? What are some of the differences, in really specific differences that you don't see in US apps? I'm thinking of the constellation, for example. I think that's a really interesting trend.
6:54 Connie Chan? Yeah. In the US, a lot of internet companies will have multiple apps, and will choose to keep them separate standalone apps. As opposed to in China, Alipay, for example, would very unlikely create a separate app for Sesame Credit or for some of its other products. The idea that everything lives in one app is actually seen as a good thing in China, because it's less friction at the end of the day. So one very, very small detail to point out in the Alipay app is the app, when you open it, looks like a grid. And each grid has a logo and text underneath, and clicking into it takes you to a bunch of new features. You can completely tailor and customize, just like your home screen on your iPhone, which items you want on the first row, which items you want on the second row.
7:36 Sonal? Didn't people do that with... Remember Google Homepages? And they shut that down. I used to love it. I used to have my own port. I Google.
7:44 Speaker 01 I'm calling out all the classics.
7:45 Sonal? Let's go retro today.
7:46 Speaker 01 I worked on MyIglo for years.
7:48 Sonal? Oh, yeah. So you've seen this. Because that actually was exactly the idea of a portal, but now it's just transferred to the phone. What you're describing is a constellation of apps within an app, which is a very Asian design quality. But if you treat the home screen as that equivalent, then you can do that, as you mentioned. It's a constellation of apps, like literally a cluster.
8:06 Connie Chan? Yeah, so for example, when Uber launched Uber Eats, it's a separate application. In China, those Uber Eats would live inside the Uber app.
8:14 Sonal? What's the rationale for that?
8:15 Connie Chan? One is you don't have to re onboard payment credentials. You don't have to create a new login system. A lot of people in China don't understand why Messenger is a separate application. There are benefits and consequences separating it, of course, but one consequence is, for example, Messenger doesn't have a lot of the social sharing features that they would have if News Feed or some kind of feed was tied to it.
8:36 Speaker 01 I redesigned the Yahoo homepage back in 2008. That was super bundled. We shoved everything together in one place and it was for the convenience of the user, right? You go in there, here's all your weather, here's all your stocks, here's your local news, here's search, here's your mail, here's Messenger. Facebook used to have Messenger too, integrated on the desktop, right? They still do. So on the desktop, all that stuff is shoved into one product inside of Facebook. So it may just be a maturity angle.
9:02 Connie Chan? It could also be a PC versus mobile concept. Maybe it doesn't work as well.
9:06 Speaker 01 To use your example of Facebook, they had a super bundled app. Something made them start to swing the pendulum in the other direction and separate things.
9:13 Sonal? And you're just saying it's cyclical? Like, is there something concrete that drives that?
9:17 Speaker 01 We, in our bundle unbundled dance, at least at Google, kind of look at what are people doing in here? How can we create more value and how can we make it even better for them? And if the answer is we need a standalone dedicated experience, yay. If the answer is we should integrate something, then yay. I don't know if anybody has like a firm rule of integration rules or separation rules. There's, to your point, there's pros and cons of both, right?
9:41 Connie Chan? Well, I also think it's different if you're talking about tools versus services and products.
9:45 Sonal? Interesting.
9:46 Connie Chan? So for example, photo editing tools. It makes sense for a photo editing company to break that up into multiple apps. Because from a user metric standpoint, I will download four or five of them. But for a company like Facebook, I only want one if it's the same exact network of people.
10:00 Speaker 01 I think this is a much deeper conversation because what does bundling actually mean? Apple comes out at WWDC and is like, hey, you can make apps inside of our Messenger app. Hey, you can integrate things into Maps. Hey, you can integrate things into our share sheet. You can make custom key. We just launched Gboard, which is a keyboard integration that has Google Search inside of it, bundled inside of the OS that Apple makes. Right, so like, what does bundling really mean in those kinds of situations? An app isn't really an app. It's a set of notifications. It's could be a custom keyboard associated with it. It's a share extension. It's an integration in a messaging service.
10:37 Sonal? But certain newer, quote, newer type apps like Snapchat, which is really not intuitive for a lot of users from a usability perspective. There's this notion of here's a classic app that you see, and it has, you know, the ability to post selfies and filters and browse stories. But then there's sort of Easter eggs that are hidden in it. And this is not the classic notion of an Easter egg, where it's like a little teeny surprise that you just get by accident and you kind of giggle. There are actually parts of the app that are entirely intentionally obscured or seemingly intentionally obscured from view.
11:04 Speaker 01 And so your parents don't find out how to use it. As soon as they start making it usable, everybody runs.
11:09 Connie Chan? But that also allows them to launch these delightful little new things here and there. And I think that's quite important.
11:13 Speaker 01 Yeah, but you know, in the, I don't know, early 2000s, like MySpace is ugly, Craigslist is ugly. They're both successful. Successful apps are ugly. And we should make ugly apps. And like similarly now, you know, Snapchat is hard to, and I hear this, right? They're like, oh, Snapchat's hard to use. We should make the app hard to use. Really?
11:31 Speaker 01 Is it just because something is successful?
11:33 Connie Chan? I really hope no one's ever actually said that.
11:35 Speaker 01 They do all the time.
11:36 Sonal? Is it sort of that these just come in waves? Or that you have to just kind of go with the flow? Is there no principle?
11:40 Speaker 01 No, I think you should figure it out for yourself.
11:42 Sonal? I don't think that argument would work with a big company, though, in the design team.
11:45 Speaker 01 Yeah, here's what happens. Like, Facebook did this. Oh, we should do this. They know what they're doing.
11:50 Sonal? Right, exactly.
11:50 Speaker 01 A classic example is to have this like nav system, this hamburger nav system. And you know, when their iOS app, they literally had like an HTML, web view, JavaScript rendering, pseudo browser, making the feed. They didn't write it in native code. And the thing would crash, and it was slow, and memory was just a zoo. And like, they had so many bigger fish to fry than figuring out navigation, because they had to rewrite the whole news feed inside of Objective Z to make it work. But as soon as Facebook finished that rewrite, they went and built an A/B testing platform. They figured out what works, and they changed it. And now it's that bottom tab bar. Right, so you just run a dangerous game when you say, oh, Google did this, Facebook did this, Snapchat did this, because you're not in there. You don't know why they're doing it.
12:35 Speaker 01 You don't have the data. And without that information, you're just kind of guessing that it's a good thing.
12:40 Sonal? After your piece last year, Everyone's suddenly coming to you like, "How do I make the next WeChat?" in their existing app, which is exactly the wrong question, it seems.
12:48 Connie Chan? Yeah, and especially when they don't have the context of why it works in a certain country or why a certain decision was made. They're copying the wrong things.
12:56 Sonal? What do you think about this Snapchat-ification and WeChat-ification of iMessage that's going on? So is that a case of copying? Or is that them just evolving? Like, how do you kind of tell the difference?
13:06 Connie Chan? I hope that's a case of Apple realizing that chat needs more emotion built into it. So, for example, a lot of the apps will have things that will look very colorful and are meant to provoke a lot of joy in the user. But from a Western standpoint, a user might look at that and say, "Oh, that app looks childish. Why does it have a cartoon character dancing around? Why is there a big flash, like, spark of light behind this image, right? Why is it making a cha-ching noise when you're earning money?" They will look at that and think it's childish or gimmicky, but it makes a feature or a use case go viral.
13:39 Sonal? I was just thinking it's sort of like the Lost in Translation effect. I call it the Lost in Translation effect when I think of that movie with Sofia Coppola. It actually, that movie bothered. Bill Murray, yeah. Yeah, I mean, it was a good movie. It was a well-made movie. But it bothered me because there was a certain element of everything is so garish and gaudy and exaggerated and a caricature of itself. And I really resisted that. And what you're describing sort of reminds me of that. One of the things that you guys have both brought up now is this notion that seemingly obvious things are actually not that obvious, that seemingly little things can have big impact.
14:10 Connie Chan? So, for example, one feature I love in the WeChat version of stickers is that you can save a custom folder of your favorite stickers. And the muscle memory here of her is actually very important. So I know, within just a few taps, how to get to a specific sticker every single time. It's almost having like these autoresponder sentences built in to my keyboard. But they're stickers in visual form. Can I find the exact same sticker in two clicks? Or do I need to scroll through a list of different stickers again? Can I change the order that the stickers are displayed to me so I can put my favorite ones in a specific folder, right?
14:43 Speaker 01 Is that really unique to Asia?
14:45 Sonal? Apple added the whole predictive emoji thing, right? That was a big deal.
14:47 Speaker 01 Yeah, exactly. So like, it is hard to find something in that very large set of things. So helping people find that thing in that large set of things is good. And I absolutely think there's cultural stuff that influences how people perceive the world and what they value and how much they value it and all of that. Like, that totally plays into adoption of apps and services and things like this. But when you get down to this feature usability level, that's where I say it's a lot more universal.
15:13 Sonal? What you're kind of both indirectly dancing around here is this tension between AI first and mobile first. Because Luke is describing like, but you can, if you have them in a very predictive way and they can really kind of further your conversation efficiently and really facilitate a lot more, then maybe emojis become a lot more expressive because of that predictive nature. On the flip side, what you're describing, Connie, sounds like a little bit of a mobile-first approach, where you can have an entire sticker replace a whole thread of written text.
15:39 Speaker 01 Input is hard for certain languages, right? Like, character-based input is very challenging on a typical keyboard. That's why you have a lot of audio-based input. That's why you have stickers, because, hey, it's a lot easier in many cases for me to say a lot with a picture than it is for me to actually go through the process of trying to type it in a traditional way.
15:58 Connie Chan? A lot of people use that as an excuse, though, for why people don't like to type Chinese on a keyboard. But I don't think that's actually the case for a lot of the top-tier cities there anymore. Typing and pinging is not that much slower than typing in English text. Another example that I see in some Asian apps is the idea that you can use your finger to hold and press an area of the screen, and it would change to another button, much like on the PC when you're using your mouse and you're clicking on something. Or if you're right-clicking on something, you had a different menu, right? On both apps, such as WeChat and also Alipay, there are areas of the app, if you click down, it will change to a different option.
16:34 Sonal? Frankly, if it just was a layperson, I hate it when shit moves around. One of my biggest pet peeves, because I'm a dedicated Android user, is how you have the most frequent calls, people you call, the order keeps shifting, so you can't even have a permanent sticky list of like, your dialpad, one, two, three, four. Drives me batty.
16:51 Speaker 01 So it was a nice example of how little things can really quickly add up and make you... I mean, your example of the stickers is similar to this, too, right? Like, you were asking about the scale question. I think this brings it up. At scale, something like an extra tap, if you're doing it 50 times a day, can really piss you the hell off. And if you're in WeChat 24/7, even a tiny little shortcut like that long press makes sense to you because you're using it so often that you would appreciate that. Every step that you can save people on something that they do free, and I think it's related to what you were talking about. If your most commonly called contacts keep rearranging on you and you are calling people like five times a day, that starts to really, really irritate.
17:35 Sonal? Yes.
17:35 Speaker 01 And wear on you.
17:36 Sonal? Well, this is where I think there's also another AI problem in mobile where this is tension between, on one hand, the whole point is to have something predict and be intuitive and you love it when it works. You absolutely hate it when it doesn't. And that's when you want to be able to specify intent. The problem is that those two things are not the same for every single person. True. So it's really tough to tell.
17:54 Speaker 01 It's part of a broader trend. Everything is turning to software. Everything is turning digital and everything is digitalified. It's really fricking hard to find anything because everything is up there in some cloud, something or other-y thing. And so you need some predictive, you need some guidance, you need some assistance. And the smarter it can be, the better, but nobody knows your brain like you do. And I don't believe in a world where we get away completely from affordances. A lot of people have had this chatbot conversation.
18:24 Sonal? Including us.
18:25 Speaker 01 Yeah, but it's not really you talking to the thing. There's buttons that show up, there's images that you click. There's actual visual affordances of things that you can interact to and make the world behind the quote unquote bot visible in a graphical and interactive kind of way. I've had my Alexa for years and I love the thing, but like, I don't know what the hell it does. Right? Every once in a while, they send me an email, it's like, Alexa is now able to tell you the stock market quotes in Taipei. Just say do, do, do.
18:52 Sonal? Yeah. It's actually funny because I get that weekly email from Amazon as well. And it's the same header, what's new in Alexa? And they never give you a preview in the subject header. They literally violate every best practice around email.
19:03 Speaker 01 But you still read it.
19:04 Sonal? Because I have to find out what's new in Alexa. Well, let's switch gears for a moment then, because so far we've been focusing on mobile. So what are some of the other interesting, because I think what you bring up with Alexa, which I find incredibly fascinating, and I've written about this as well, is that it is about design for a screenless world. And so what are some of the new things, like trends that are coming up that could change a lot of the design discussions we're having? Or would you argue that these universals are always going to apply? I mean, what are you saying?
19:28 Speaker 01 I think the constraint on all of this is the real world and the human body because we're just not evolving that fast. And our society doesn't evolve that fast either. So I don't know if you're familiar with like architectural shearing layers?
19:41 Sonal? No, what is that?
19:41 Speaker 01 So architectural shearing layers is like things like the foundation, really fricking hard to change. Things in your room, really easy to change. So it's like slow, quick, quick kind of models. And there's certain things that just move really slow. The size of our hands are not going to change and evolve. And so dealing with the ergonomics of the hand, you've got to have certain sized touch controls. You have to have some sort of accounting for what kinds of gestures you can do and so on and so forth. And I think that is a boundary in some ways. And then the other boundary is like, people talk about voice all the time, but if we were here all talking to our phones while we're doing this podcast, like it would not work. There's like a time and place.
20:22 Sonal? Yeah.
20:23 Speaker 01 For different inputs. And when we get too myopic of screens are going away, well, no. I mean, AOL is still here.
20:30 Sonal? This is the argument I make about Alexa all the time. I mean, the fact is like people tend to treat voice as this homogenous thing, when in reality, it is completely carved up. In the social space, you're really unlikely, although who knows? Because people do surprise in their actual use of tech and we never know what people end up doing. You're right, you're very unlikely to be sitting here all talking to your phones verbally in that same exact space. But when you're in a contained.
20:50 Connie Chan? I'm not seeing Asia.
20:50 Sonal? So interesting. So what's happening in Asia there?
20:52 Connie Chan? People use voice for a lot of things in Asia. When I'm sending a WeChat, a good chunk of my WeChat messages to people in Asia are all through audio.
21:00 Sonal? And this is from public context? Because the point that I'm making is that people.
21:04 Speaker 01 Would you do that in a meeting?
21:05 Sonal? Right.
21:06 Speaker 01 In Asia.
21:06 Sonal? Or even in like in the hallway or like in the vestibule of a restaurant or a lobby?
21:11 Connie Chan? Oh, yeah.
21:12 Sonal? Oh, interesting.
21:12 Connie Chan? People will do this during meetings, they'll do it in the hallways, they'll do it in the restaurants. You're using audio input all the time. And there's another, almost like a Siri type product that lives inside WeChat right now. So you can send it an audio message and it will show you a visual response.
21:27 Sonal? Interesting.
21:27 Connie Chan? Which can be a map, it can be text, it can be all kinds of things.
21:30 Sonal? See, that's so interesting to think about that cultural context, because one of the theses that I think about a lot is that the reason Alexa and Echo work so well is because they are in a contained, constrained space. Shawn Madden wrote this op ed for me where he talked about how you can have a smart environment and a dumb device. You can have a dumb environment and a smart device. So if you're in a stadium, for example, you're going to have a very self-contained environment. You can have a Disneyland with magic bands. It's a contained environment and you can have the smart device. We're going to get to this world where we're going to have our smartphone, which is smart. Smart devices like Alexa or Echo. And then you can have like a dumb environment. And so what happens when that entire environment becomes sensorified? Where you have a smart environment and a smart device? And we're not quite there yet and the rules change.
22:12 Sonal? So I can see us then maybe evolving to a world like what you're describing in Asia. But culturally, it seems like that's a sort of a leapfrogger for kind of creating that.
22:20 Connie Chan? In Asia right now, it's very common to see someone walking across the street and just talking directly into the microphone in their phone. That's totally okay.
22:28 Speaker 01 I go back to this Richard Farson quote, which is, "Nobody smokes in church, no matter how addicted." Our real world has norms and expectations and they may vary from culture to culture. But I think that real world perspective sort of drives integration of input and what does or doesn't work in an environment.
22:46 Connie Chan? Before people were getting notifications or messages on their phone, you might say, "I would never check my message or my email during a meeting." But a message from your wife, for example, is something you might look at during a meeting and just very quickly reply to text or put away. But the fact that you might look at it during a meeting is something you might not have said that you would have done.
23:06 Speaker 01 Depends on the meeting, right?
23:08 Sonal? It also depends on the device. There's new devices that are designed for actually being screenless. We can't ignore that because as things get more intelligent, you can be screenless. Example, I'm thinking of like Christina with Ringly. You don't even need to have a phone when you have like a bracelet or a band or a ring on your finger that can notify you for the most important phone calls. You know, your husband, your child, you know, an Apple Watch. You can have a glanceable design internally. And both, you know, in Asian companies and in a company like Google and even what just... An individual big company, you've worked at some of, both of you guys have worked at some of these different big companies. How do design decisions play out? Tencent was able to do so many things because they were very orchestrative and architective, like they're the architects in the matrix.
23:52 Sonal? And in this case, like, you know, Google does not strike me to be that way. Like, or is there a point of view?
23:57 Connie Chan? I think there's a lot of tension, right? Because you have people coming from the business case saying, I need to push this thing that drives revenue. And you have this other group that's saying, I need to protect the user experience so that the overall platform maintains user trust. And so there's a huge tension.
24:11 Speaker 01 It's bigger than that. I mean, you have lawyers that are also involved in those conversations.
24:16 Sonal? Why are lawyers being involved with design conversations?
24:18 Speaker 01 The laws differ between all the different countries and there's things you can or can't do. There's privacy people that get involved as well. There's marketing people that get involved. There's also what can you do on the engineering side? How much work is it? How little work is it? There's hundreds of signals at any point in time that you need to balance and strike the right, it's not, what's the word I'm looking for?
24:41 Sonal? You're not looking for compromise on purpose.
24:42 Connie Chan? Tradeoff.
24:43 Speaker 01 Tradeoff's not the right word either.
24:44 Sonal? Because otherwise, both of those suggest a regression to the mean. That seems like that'd be the worst thing you could do.
24:48 Speaker 01 Yeah, you've got to find the sweet spot. So this is how I liken a product manager's job. It's you have a very large soundboard and there's like 100 knobs. And here's the legal knob, and here's the design knob, and here's the, you know, like revenue knob. And there's just thousands of knobs. And you're kind of trying to find the right mix so that the song actually sounds good. This is why startups do well, because startups can hear that mix much more quickly, because they can launch more quickly, they can rev more quickly, and they can monitor stuff more quickly. If it takes you a year before you hear that mix, like this is our traditional software waterfall department, it's going to take you a long time to figure out whether or not it sounds like shit or not.
25:27 Sonal? Yeah.
25:28 Speaker 01 And that's very, very painful.
25:29 Sonal? I hear what you're saying, that there's this mix and you have to get to that moment to hear that, tune into that mix much sooner than later. But is there sort of an algorithm almost that you have in mind for how much to weight one of those qualities versus another? Like, if you're doing a linear model, like 10x for the weight of users and 2x for the weight of management? Or, I mean, it seems silly to try to come up with that.
25:48 Speaker 01 I frigging wish, right? It would make my life 100 times easier. It's the same thing, you know, like startup advice, build a great product. Well, yeah, not really. There's just the, there's all of these like palpatudes that, this comes all, let's, hey, let's do this. That's easy. Right, but the reality of the day is it's nothing like that. It's a combination of a number of factors, some of which you understand, some of which you don't understand, that play into what makes something work and what makes it not work. Understanding all that as much as possible and trying to get feedback from that when you make those sorts of changes is the only way to really survive.
26:21 Connie Chan? What's really interesting about this topic is the role of what a PM can do and the level of power they have or the level of influence they have.
26:29 Sonal? You mean product manager, not prime minister?
26:31 Connie Chan? Yes, yes, yes.
26:32 Speaker 01 Yeah, some, I actually-
26:33 Connie Chan? Product manager.
26:34 Speaker 01 Somebody has said this to me this morning, like, "Do you mean prime minister?" I'm like, "No, no, no, product manager."
26:39 Connie Chan? But that role and their level of influence is so different across software companies, right? Like, for some companies, it's very engineering-led, some it's very business-led. And some places, the PM's extremely empowered, and some places, they're not. And they're more just organizing timetables.
26:55 Speaker 01 Yeah, they're more like program manager than product.
26:57 Connie Chan? Yeah.
26:58 Speaker 01 Yeah, so I put out a thing, "What does a PM do? And can I distill it down?" Well, three things I think a PM does. And immediately, somebody's like, "Ask 10 PMs what their job is, and you'll get 10 different answers." Ultimately, their job is to get products out the door, but that means totally different things depending on-
27:13 Connie Chan? Yeah.
27:13 Speaker 01 -where you work and how that place works.
27:15 Sonal? And there's no one right way.
27:15 Connie Chan? It completely varies by company. And some of that culture comes from top-down. Some of that culture changes. It's so different from each software company.
27:23 Speaker 01 Yeah, it's the stage they're in, right? It's the type of people that they have working with them. It's the type of product they're building. Product managers at Google that work on fiber are kind of in a very different world than the ones that work on robotics or, you know, ones that work on something like software.
27:39 Sonal? So I guess my final question for you guys then is around this notion of what are some of the other interesting trends that you think are going to change? Like, right now, I think we're still talking about what is, like, what's not yet here.
27:50 Connie Chan? Visual input, it goes beyond QR codes. In the US, we use the front of our phone. We don't really use the back of it, besides taking photos.
27:58 Sonal? Yeah, it's like what you said. We don't still use every part of our phone in this experience.
28:01 Connie Chan? Yeah, we're not using the camera in all the possible ways that-
28:04 Sonal? What about you? Are there any trends that you think are sort of-
28:06 Speaker 01 I think we're dancing this line with this bundle-unbundle thing again in a broad sense. And I would characterize it bigger than just apps, but like, how do services play in the OS? How do they play in platforms? How do they play in between those platforms? Right. You remember, we had the apps of the PC world and the web took off and that kind of changed the dynamic. Then we came back in the app stores and now we have this thing, Instant Apps, which is kind of an app inside of a web browser, but it's really an app inside of the OS. And so I think that's going to play out in interesting ways as people rethink the notion of a closed container in today's world and then kind of put things together. Bigger picture than that, the whole world is going to become input.
28:46 Sonal? Right, exactly.
28:47 Speaker 01 And to some extent, output as well. I don't know if you caught like some of the research that Disney put out a couple of years ago. You touch a plant, it does something, you touch your doorknob, it does something. And all these things can become input and really trying to design for a world where everything is input, not just putting your camera, phone and pointing it at a thing.
29:05 Connie Chan? And every surface can be an output.
29:07 Sonal? So that's actually related to a point because of the economics of Moore's Law and everything becoming so cheap. When everything becomes sensorified and connected, you basically have exactly that. Everything can become input and some can become output.
29:18 Speaker 01 We're already there in some ways in that, you know, just like it's cheaper to make a microwave with a clock than without it because of standard components. Like you have to do all this custom work to remove the clock from an oven. It's there. Similarly, the cost of adding Wi-Fi to things is approaching a similar thing. We're going to have to pay a lot more to remove the network thing and make all the other components not work with the Wi-Fi thing or network thing or whatever it happens to be. And at that point, when everything's got a radio and everything has sensor input, it becomes Wild Wild West.
29:47 Sonal? That's great. Well, thank you guys for joining the ACX and Zee Podcast. Thank you.
29:51 Speaker 01 Thanks.
Previous Company Building in Crypto: A Conversation between Brian Armstrong and Chris Dixon Next What's Next for Education Startups Page updated October 9, 2026 · transcript and summary are machine-generated