More Launches and Cool Dev Tools Websites

If you're gonna write markdown and it's just text, you have to be really good at writing the text. I actually think PlanetScale did a really good job with it, where they focused on differentiation on the first line in the beginning. And then they talked about some things like bite and things like that, that maybe you don't know what they are. But if you scroll, there's actually a really good explanation on what they are, how they are different. They have like a diagram and stuff like that, that really explained the product and the pieces that the product had, which I think is what the developer wants.

Everybody's launching. There's a million launches. I love all the launches, personally.

I don't. I think we should stop talking about launches. We talk about them every week, but we'd love to get your opinions.

Give me one more week on launches. I love launches. And this is the launch season. So the first launch to talk about is the Browserbase takeover. I had, I'm an angel investor in browser, Browserbase. So I had some inside info as like this went along. I'm curious, first, your perspective on the launch, cause you saw it from the outside.

I like the launch a lot, but I feel on one side, all of the launches are looking the same. Like you're recording it in a studio, it looks pro, or you're making like a fun video that is also like recorded like a publicity or an ad. But the main thing that I did like about Browserbase, that I think they did really well, was the plane. I love the plane. I saw it all around on Twitter. And to me, what was fascinating was that I don't think people in SF saw the plane at all, but in the end, I think what matters was that the few people that saw it actually tweeted about it. And those tweets became viral. So it was more of the virality of the tweets rather than the plane itself being seen, because I saw some of the videos and they had to zoom so much. So I don't know if it was flying very high or not, but it was interesting. I would love to get your thoughts as the insider.

Yeah, I mean, I didn't even know about the plane coming up, but that was like an interesting hack. If you see a plane, you have to like take a picture of it and tweet it if you're on the streets of San Francisco. And so that was kind of a little hack, probably an expensive hack.

In Miami, there's the things where you are in the beach and you see the ads that are like electronic on the beach. They should do the same, but in hikes in San Francisco, maybe, or something like that, just because everybody's a hiker in SF.

Hmm, that's not a bad idea. All right, we'll see if people start doing that on your suggestion. Yeah, so first off, I love that they paired a fundraise announcement with an actual release and there was open source software, there were features in the monetized product and there were security updates and the fundraise. So especially getting the security updates and the fundraise, which I typically consider more boring news to launch on their own, paired with all this other stuff. It's great, because then you're getting people interested, but you're also checking the boxes. So that was fantastic. And then what Paul did was he told all his stakeholders, like me and just everybody who ever talks to him, hey, on this day at this time, I'm doing this tweet, we're dropping a video and this release and the fundraise, and you must retweet it and or comment and like it. He was like, I need the engagement. And that got them way more algorithmic attention than they normally deserve in quotes. So that was probably the best thing was the upcoming, like the coordination leading up to it was really well done.

Neon did also launch the database and they launched an event called Neon Deploy Online, and they launched a big feature called Authorize. On that event, for example, they didn't get that many views. So I don't know how worthy of the investment it was for them. Having said that, I think the Authorize feature is incredible and it should have got more attention. So in my mind, for example, linking to what you just said, this Authorize feature lets people basically use the database with the same credentials as the user that is logged in by sending a JSON web token. And that works with Clerk, with Auth0, a lot of authentication providers. They should have engaged with stakeholders, similarly to what Browserbase did, because if they would have maybe engaged with the founders of Clerk, the founders of Auth0 or something like that, and they would have said like, we're launching at this time, you are our partners, tweet with us or do something with us, I think it would have been much better as well. I always talk about that from Supabase, like Supabase started, I think, these launch weeks and they were really successful, thanks to actually always pushing features that work with partners, so that everybody talks about it at the same time.

Yeah, if you're gonna collaborate, actually collaborate and coordinate, and that's key. Builder.io also had a launch that was similar to Neon. Like they built up to it, they put a lot into it. It didn't get that much attention. I think it did really well with their enterprise audience, so maybe they're getting the ROI there. But what's interesting is you contrast it to when they, this is an update to an AI feature they have. If you contrast it to some things we just said, where you said, oh, everybody's doing studio videos, their CEO, Steve, launched a year ago this feature with a Loom video, a 10-minute Loom, but he coordinated really well with Figma, and there was really good coordination leading up to it. And it caught more views than Next.js Conf like the same week, and it was like crazy impressive. So like the contrast from the same company getting totally different results is interesting. I think it comes down to these little things, actually, these things that are so easy to forget.

The devil is always on the details. Prisma also launched this week. They launched a new database, which of course makes sense, like doing an open source ORM. It's not a good business, they need to make money. So they launched like a serverless database, which is basically a competitor for Neon and for PlanetScale. And in their case, I didn't like the launch, mostly because if you look at the tweet, they don't talk as much about differentiation. And in the blog post, they also don't talk as much about differentiation. When I read the blog post, they talked a lot about these cold starts, which is not true because like Neon doesn't have the cold start problem because you can set the auto-scaling and always keep a minimum. You can do the same with PlanetScale. So they are saying and arguing that serverless databases don't have that, even though that's a lie. So I don't know, I feel that people who are reading this and know the other database will be like, no, that's not true. And then outside of that, there was nothing unique. So my question is, why would I pick Prisma? And if the answer is pricing, I think pricing is always a fight to the bottom. I think you liked it, didn't you?

I, well, I mean, I only saw what you showed me. I didn't dig deep into it. I don't know about what they're lying about or wrong about, but I do like if a company can add feature parity to someone else. And then if their existing customers can now easily be further monetized on a thing that maybe they don't care about, the differentiation very much. My example of this would be Pinecone came out hard with all this cool vector stuff. And Postgres was like, yeah, we can add some vector stuff. Like, is it differentiated? No. Is it all you need? Yeah. And if it's all you need, it's all you need. So that would be the question. For people who need the better stuff, they'll go to Neon, PlanetScale, whatever.

I agree. Last thing to mention on this, one thing I think is important as well for launches, besides the stakeholder that I've seen some companies do is if you're a multi-product company and you see that your customers as a use case could use a lot of your features or some features together, I think sending emails to your existing customers about how they can use their features and specifically about how it can benefit them, I think makes a big difference. Working now on this with Appstash, they launched a new feature that is really useful for some of their AI companies that are their customer. And for example, what we're doing now is actually using AI to research their AI customers and see exactly how they would use the new feature. And we're creating emails that are basically written by AI, giving them actual recommendations on how these new product would serve them. We haven't shipped them yet, but I think that this could be something that really works. So thinking about customization with AI now for launches, I think could be a big deal as well.

Let's come back to this in February. We're gonna try a bunch of this stuff with some Q1 launches at Laravel. And I'm very excited. Like all the stuff you're talking about, that's a whole episode into itself. Okay, one more launch though. GitHub universe, one more launch. I think they are the pinnacle of in-person experiences for dev tools. They call themselves the world's fair of, what do they call it? Developer conferences or something. So they're trying to brand themselves stronger and stronger in this sense. But their releases were pretty meh. The one that was interesting to me was GitHub Spark. And it's just because all the AI launch, we've already talked about like bolt.new, it looked very similar to that or v0, but it's not an actual product that you can get your hands on. So it's kind of a wait list page and it felt kind of muted. And I felt like for the most part, while they were doing a good job tweeting, they didn't really have much material to get out. So it was kind of the marketing team making a best effort with what they had.

And I think in these cases, I would rather wait. I know they couldn't because they had the date, but in my mind, if you're shipping something that is really new, for sure, you can actually talk about it and then launch it later. But when you're launching a copycat of v0 and bolt.new, and you have a wait list and a blog post and you're not shipping it, it's like, what the fuck? Just for talking of something that is a copycat, it makes no sense. And the other thing that they did was they basically made, they created this plugin for VS Code that works very similar to Cursor. So now Cursor has a new competitor. But I felt that from an online perspective, it was shit. Like it's entirely in-person experience as you were talking about before. And I've heard that the in-person experience of GitHub universe was also not as good as NextConf. Because if you think about it, NextConf has all front-end people. Even if you're not next, maybe you go to meet and talk with other front-enders. GitHub is so broad that there's not that many groups or camaraderie or these people that know each other. And NextConf has the thing where if you are a dev rel or a YouTuber and you work with Next or front-end, you have to be there. So we create this necessity for people to be there and connect, but I think it's incredible. GitHub universe doesn't do that. So it seems more like a corporate event. Like, I don't know, I don't think it was that bad.

Yeah, there's kind of a sweet spot on the size and like niche for a lack of better word. And yeah, I think that's a good insight. And you know, maybe it's too big. Open source is too broad as a category. Like you can't, you're not guaranteed at GitHub universe where everyone you talk to is working on something similar and has similar tastes and interests. So that might be something special about a Next.js Conf. Okay, finally, we can give you a break from launches and let's talk about something else that we're always excited about is websites.

Yeah, I'm always excited about websites. And we talked today a lot about Neon. We talked a bit about Prisma. So I think this topic actually started with PlanetScale. So PlanetScale shipped their new website where I have thoughts, but I know that on purpose, you haven't seen the website on PlanetScale. So Hank, feel free to first watch the website and give me your thoughts and then we can take it from there.

Okay, I just opened it up. Yeah, when you said you wanna talk about it and it's really different, I said, oh, I'll look at it blind. So my first thing is, wow, they don't have a hero. They have a, you know, a very monospace-esque font, not quite monospace, but it feels very docs, doc-like. And there's ASCII art on the page explaining things. I love image, you have to have images that actually explain the product on your page and they're doing it with ASCII art. That's fun. The, this is all just like markdown plain text stuff. They have, the asterisk is bullet points. This is really interesting. It's super simple, clearly targeting devs. I like this. This is very different. It's definitely steering away from the every website needs to look like Stripe or Vercel or Linear. I'm not seeing this very much. I like it. Different is good.

I liked it too. What I liked about it is that like it's markdown. It's literally just a fucking markdown. So I think in that sense, it's like the design is lack of design, but it's very developer focused. What I really liked about it, and I think it's really hard is if you're going to write markdown and it's just text, you have to be really good at writing the text. I actually think PlanetScale did a really good job with it where they focused on differentiation on the first line in the beginning. And then they talked about some things like bite and things like that, that maybe you don't know what they are, but if you scroll, there's actually a really good explanation on what they are, how they are different. They have like a diagram and stuff like that, that really explained the product and the pieces that the product had, which I think is what the developer wants. So I really loved the style, but I think it's really hard to nail it because you have to be very clear on your differentiation, what you're doing and very, very good at explaining it. But I loved it. I think this could be a new trend where everybody I think was copying from Stripe or Vercel on the website.

That was my question for you, was do you think this is a new trend where we're gonna have this back to basics, markdown style website? And you think so?

I think there will be more. Somebody replied to me on my tweet that I talked about this, that there were already two websites that were similar. One was turbopuffer.com. The other one was SF Compute. We'll add them to the notes, but both are actually very similar. And I think they also did a good job of explaining the differentiation and why. I specifically like more the turbopuffer than the SF Compute, just because the other one was a bit too complicated in the beginning. But this I think could be a new style that competes with this Vercel Stripe. I actually think the style was created by Nat Friedman. If you see Nat, he was the CEO of GitHub before. And before that with Xamarin, his website is marked out and he literally did a markdown and that's it. And I saw a template for Next that is Nat Friedman's website. And this has the same vibe. So I think it actually came from him.

Very interesting. I mean, there've always been markdown or plain text websites. There's that one VC where that's a very AI focused VC by Sarah, what's her name, Sarah Guo. Her site is like that. So there's always like a niche group that's doing this. I wonder if it becomes a trend. Another question I have for you is, do you think they should, so they do have colorful images of the logos of their customers. Should they replace those with just a lit, like should they just lean all the way into it?

It's a good question. I actually liked that there were some things with color because they catch your attention more. I actually think customers are really important in general. So the fact they made them colorful, I think was useful. What I would have done is slightly different. So I would have used like a list similar to markdown, but instead of having just text, I would keep the colorful logos. That I think would have been smart because then you keep the markdown style, but you also call some attention to it. Similarly, I would like to see some, like, I don't know, I would do some ASCII art, like potentially you wanna do a GIF showing something, do an ASCII art GIF showing your product or something like that. Like you can even take it to the extreme, which I think would be interesting. The only other type of website that I've seen lately is MotherDuck. I don't know if you've seen MotherDuck website, but it's vintage and there's a fucking duck. There's a rubber duck that is just like flying around and stuff like that. And we copied our website from them. We really liked their style and that's why we copied.

We just straight up copied it. We wanted colors on codemarket.fm and something more playful. So it's, yeah, we blatantly copied it and it's great. I love it. I think it's so fun. The animations they have are really fun. It's a fun brand and something interesting about brands is if you want to have fun in your brand, you have to do it in the beginning. Otherwise, once you get corporate, you get your marketing and your PR and your people who just won't let there be fun anymore. And companies that do this, they'll stick with it forever. Like if you look at Reddit, they have their weird little alien snoo. That's such a weird thing. If they were public first and then they announced, oh, we have this little mascot weird thing, people would be so weirded out. So I like that MotherDuck has the ducks now. They're establishing the fun, weird parts of their brand and yeah, they've got some color and life in it.

I think MotherDuck takes it to the extreme in a good way where they actually have different type of rubber ducks that they give away in every conference, similarly to what GitHub did with the stickers. But what's nice about it is the rubber ducks are a gift for your kids when you're away. So I think for parents, it's fantastic because then the kids are asking for, I want my duck collection and stuff like that of the different styles, which I think is smart.

And it's funny that I brought up Reddit as the example. GitHub is an even better example. They got the weird Octocat. Imagine if they tried to introduce that now. Microsoft would never, they'd be like, what is this?

I agree. Did you see the Easter egg from Vercel, like the Halloween thing that if you went into the triangle, the spider went down? What do you think about it?

I didn't see that. This is, Jenny tweeted that. Oh, that is fun. And it's the, yeah, Easter eggs like this are necessary. And Vercel has other nice things about their triangle. Like if you right click it, it opens up all the brand files for download, which is just like a nice, it's a high utility Easter egg. This one's just a fun one.

I love Easter eggs in general. I think it's a fun way to get people to try more your websites. I remember I talked before to a CMO of Stripe and there was a section in the home in Stripe where there was a world that you could grab and just like spin around and see different things. And what they were saying is that there were some sections that developers hated on the text because it was like, oh, it's the global enterprise, but those sections had something fun so they could play with. So then what Stripe did that I think was smart was they had developer sections and enterprise section with some Easter egg or something for developers, which I thought was incredible. I've also seen other Easter eggs like the Konami code, the left, right, left, right, AB, where a dinosaur will show up or something like that. I love them. I especially love the console.log Easter eggs, where if you go, you'll see something maybe for hiding or something like that, but I don't know. Have you seen other Easter eggs that you liked?

None come to mind that you haven't mentioned, but that is very genius to intentionally put less boring designs and more Easter eggs around the boring parts of the site or messaging. It's a way to create engagement around something people might not be interested in, but is important to get across.

Exactly. Very close. Last thing to mention on websites. This is my exciting topic. There's a company called Jina AI, where they actually built a doc page that is specifically for LLMs, and I loved it. I'm all in for it. Like if you click on the link, we put it on the notes, we're showing it now a bit if you're on YouTube, you'll see that it actually says like, this website is not for humans, it's for LLMs. And they are teaching how an LLM can use their API and help a developer use it. Their idea is that if a developer is coding with Cursor or this new GitHub launch that is the Cursor competitor, they can actually be better at using Jina AI thanks to the fact that they're explaining to an LLM how to help a developer use it. I don't know if it's gonna be a new trend or not, but I love the idea. I think it makes sense because more and more, I think developers will use services in a way that is recommended by AI instead of them doing it themselves. And their tweet got so many retweets just because it was something original and different that I thought was pretty good.

I looked at it and it just made me think how many companies are gonna start doing this and how good our company is gonna get at implanting themselves into LLM training. And how important is it gonna be this new form of like SEO almost? It's LLM optimization, the LAMO, let's call it today. How good are they gonna get at inserting themselves into the training so that a developer who types a command in cursor, Copilot, V0, Bolt, whatever, Spark gets this suggestion that the LLM just thinks is a good idea because someone told it it was a good idea on multiple pages in their docs and GitHub issues and all these places. It's interesting, it's going to be, I think it's potentially dangerous in the short term as companies figure out how to sort out what is good training versus bad. What's almost, it's not corporate espionage, but you know what I mean, right? Like it's something you can manipulate just like the early days of SEO when you could hide the same keywords 100 times in white text on your white background and Google would be none the wiser.

I know what will happen. Two thoughts that I have on this, like one is a data point. Like one of the companies that I help is called AppStash. They're a database. We do a lot of first data attribution to see who's bringing the signups. The source number four is ChatGPT. They get so many users. Actually, that's ChatGPT is recommending them and they come through that. One company that I'm helping out now called Scrunch, they're very early, but they wanna be like the Ahrefs of LLMs. And for them, for example, you put your company name and they will automatically create questions that should answer your company. For example, if the company could be in Vercel as an easy example, it's like, what platform could I use for deployment? What platform could I use for this? And they will do all of the questions for you and tell you if your brand or your competitors are being mentioned in the answer in the first, third, second, or last, or where, and by which LLMs. So I think we'll see more and more of companies focusing and thinking about this as they should.

Yeah, this could be a good topic for our friend who we're gonna bring on as a guest in a week or two.

Actually, it's in a week. So that's something we're not mentioning who, but we're trying to bring recurring guests. One of our friends who is specialized on AI and being a CMO of AI startups is coming for three weeks to have thoughts with us on more topics related to AI as well.

Yeah. Well, I think we've done enough for today, right? You've done with launches. This is the longest episode we've done by far. Thanks for anybody who made it this far and good talking to you.

Thank you. So if you have any feedback, feel free to tweet either to Hank or to me on any feedback on this. Thank you.