Hank: Hey everybody, reply to this tweet… 

Gonto: Yes! hahaha…

Hank: …with your best name for an agent builder product. Like, go, that's not Clark. I don't know, something dumb like that. There, now we've got the clip. Gross. 

Hank: Gonto, are you launching an agent platform? Because everybody else is launching agent platforms. So when's yours coming out? 

Gonto: I should launch one. I was actually testing now. I was getting one of the transcripts from my advisorship calls and uploading them to a chatGPT to see if I can get chatGPT to answer like me. And then I don't work. 

Hank: Get one of these new agents to build you an agent builder. And then you can build your agent.

There are agents everywhere. So the main thing we're referencing right now is last week, two competitors launched agent tools within one day of each other. And that happens from time to time where things just converge.

They're close enough together that you and I already talked about this before. We don't think that one launched and the other did like, “hey guys, let's get this done right now.” They were probably already going to launch on those days.

There's a chance that Retool, Retool is the one that launched the day after. Superblocks is the one that launched before. There's a chance Retool was going to launch like this week and they were like, “We got to do it now. Let's pull it forward.” 

But what did you think when you saw these? 

Gonto: First thing was I was blown away that they actually both did it one day difference. For our listeners that don't know, both Retool and Superblocks help building internal apps.

That's their main focus that they've had. Superblocks is open source, but they have a hosted version. And then Retool was always paid.

The launches were very similar in the sense that they both had a main video. In both main videos, the CEO talks first, but then they show UI. They both have new cool landing pages on their website that have some images and some text.

They both did tweets from the founder and from the company. I would say they both were very, very similar, but if I have to pick one, I personally like more Superblocks. 

Hank: Agree.

Gonto: I like their positioning more. I'll explain more of it later, but which one did you like more and why? 

Hank: Well, the thing I don't like about Superblocks is they gave it a name, Clark. So you sent me that tweet and I was like, what is Clark? At first I misread it as clerk and we talked about clerk last week. And I was like, what is Clark? And it wasn't until we got on the call that you were like, “Hey, it's Superblocks thing.”

And I just kind of everything finally connected for me. Companies are constantly doing this. They're giving too many things names and everybody has a name for their agent builder now.

So that's their only miss. I thought, I mean, their video was better, honestly, mainly because the product UI is better. It's more to my liking, at least. Like retool’s, it's like a GPT interface. It's text-based, you prompt and like there's text responses. 

Gonto: But the end product, I think on how they think about internal is different.

Like, first of all, I agree with you. Naming your AI sucks. Like, please stop doing it.

Every company that I know has a difference. So now we have, Oh, I'm going to try Clark and Zoe and Rafa, like my kid. I don't know, like stop naming them because I won't remember your names.

It's hard enough to learn the names actually quick aside, Laravel has grown so fast that one of our engineers built a name flashcard quiz app so that we can just learn each other's names faster. It's hard enough to keep track of your coworkers names. 

Gonto: Exactly.

Hank: Like quit giving us more names, just call it Superblocks agent or retool agents, you know, and that'll be fine with us.

Gonto: Retool has shipped two products. One was called Retool AI. The other one was called Retool Agents.

Retool AI, I don't think it's that clear by the name, but Retool Agent is pretty clear. Superblocks did Clark. The other thing I didn't like about Superblocks is that they did this cringe thing where they say, if you reply to this tweet with Superblocks, we're going to give you for free the system prompt of cursor and these other tools.

First of all, there's no chance you actually have the real prompt from cursor and all of the others. Secondly, if you want to be cringe, go to LinkedIn, don't do it in Twitter. And if you're going to be cringe, yes, do it from the human, like have a human be cringe, like somebody on the team, not the brand, like making the brand tweet cringe…I think it's fucked up. 

So that's one thing I also didn't like all the Superblocks. Let's talk about the landing pages.

Which ones do you like more? Do you like Superblocks landing page more or do you like more Retool landing page and why? 

Hank: I'm stuck. I'm trying to think of what cringe thing I could say, like for the clip that we put on Twitter that like, hey, everybody reply to this tweet with your best name for an agent builder product. Like go, that's not Clark.

I don't know, something dumb like that. There, now we've got the clip. Gross.

Gonto: Did you check the landing pages? Like if you go to Superblocks website and Retool websites, you'll see the landing pages of each of the new features that they launched. 

What I personally liked was like Superblocks website’s landing page, I thought was really good in the sense that as you scrolled the image showing, what do you do first? What was the next step and why was very easy to understand.

And I think what they made was in the top, they mentioned the main things that they do different, which is this idea of like, we'll integrate your existing tools. We'll use your design system. We'll make sure that authorization framework is okay.

And all of that is really what matters when you're trying to build something like a UI builder from Superblocks. Retool also had text and images. So they were showing some stuff, which I liked, but I felt that the UI was a bit worse to be honest, than Superblocks one.

And the other thing was, it was hard for me at the beginning to understand what were the differences between the two things that were shipped. And I actually finally understood them, not through the landing pages, but through the videos. When I saw the videos, I understood how they were different.

And we'll talk a bit later about that, about why they are different and what's their positioning different. But on the landing page, it was hard. And when I went into Retool AI, I never expected that the idea was to add AI to your existing app.

I thought it was, oh, you're going to use AI to build something, but that's not the case. That's more of the agents. So overall, I personally like more Superblocks because they made the images and the show a lot bigger.

In Retool, it had sort of like similar weights, the image and the show to the text, which I thought was meh.

Hank: I agree with that. Yeah.

I'm looking at the pages now. The main page I looked at for Retool's thing was their blog post about it, which has a lot of text about it. It does have the video and the video was the most helpful thing.

Gonto: The other thing that to me is interesting is how they use AI is completely different. Superblocks tech is, I write a prompt and we’ll help you build a UI using AI. So they have like a coding agent, a support agent and stuff like that internally.

At least that's what they say. I don't know if that's true in the end, but they are talking about that. And what it does is by writing like a prompt, similar to V0, you'll be able to get a real app and the app, you can make sure it works because it's connected to your existing services.

It checks for authorization, it checks for bugs. Like they have so many agents that will help you with that. But overall, they build a real application that's working, connecting to your services with your design system and enterprise.

And they actually tell how V0 is more like you create a first mockup, but they don't do a real app. Retool on the other side, they take a different approach. Their take is now you do not need to build internal apps anymore because why do you need UI and to build internal apps? Because you need humans to manage them.

Retool's take is we do not need humans to manage those things. You actually need to have agents, maybe a sales agent, an accountant or something like that, that you ask it what to do and it automatically connects to multiple services and gets shit done. 

So Retool's take is you should stop building internal apps because with agents, you can now give commands and with those commands, something will be done and that output is done and you don't need a UI.

Superblock's take is we use AI to build UIs that then humans can use to do something. Which of the two do you think will win? But I love that they're so different approaches to building an internal platform basically. 

Hank: There's definitely place for both of them.

In the short term, I think AI to UI that you can use and manipulate, it's better. It's easier to adopt and for someone who's already a Superblocks or Retool user, it's a stepping stone toward, I think, that eventual end result. Let's say 10 years from now, yeah, I send a voice memo to my AI assistant and it does whatever I just asked it to.

We're not there yet and it's really hard for people to get to that point, right? We still have trouble with a lot of people adopting basic automation. Put aside AI. Hey, let's send an email to everybody who signs up.

People still do that manually and there's not been a need to do that manually for 10 years but that's still the expectation of a lot of people. So I think we do need those stepping stones of, all right, if Superblocks, it has the UI to build an internal app, yeah, just plug AI into that and then let me configure and use that. The pure text agent stuff, I still struggle with personally.

I've tried a few and I struggle with them versus hopping into Zapier or N8n. And maybe that's me but I think when you're trying to launch this product to the people who are already your customers, it actually reminds me of our conversation on our last podcast where, hey, if you have a shampoo and you want to launch another product, you could launch a conditioner. That will be a smaller market because not everyone who uses shampoo uses conditioner but there are other products that are related to hair care that might be bigger markets. Maybe brushes are bigger. Maybe hair dryers are bigger. I don't know. I'm not, this isn't, this is Jason Lemkin's metaphor that I've just adopted.

But to me, maybe Superblocks thing is more of a conditioner. Maybe it's like the easy sell to their existing customers and in the short term that will be easier. But I feel like it still positions them better for the eventual thing.

Eventually they can say, they can tuck that whole UI thing under a sidebar and then eventually even after that hide it. Whereas the way Retool is going, which felt from the video very pure text and GPT-like, they're going for a totally new audience and I don't think their customers are going to adopt it. And that's an interesting distribution problem.

Gonto: I agree. And I think now there's still a trust problem where people do not trust that AI will accomplish the task. So if I use AI to build a UI, I can make sure by iterating on it that the UI will eventually work and then I ship it.

If there's no UI, maybe sometimes it will work, sometimes it won't. But because of that, I think UI still matters a lot because UI gives people trust. So I think on that sense, Superblocks tech, which I think is more now, it makes sense.

Retool one, I think it's a lot more futuristic, but I don't think the market is there because as you're saying, Superblocks still targets developers, but now these developers or the junior developers who are building it, it's easier for them because they are prompting and doing other things. Retool now instead of targeting a junior developer basically targets the whole org or like, okay, you can just do this and ask for it, which I don't think will happen. 

What I would do if I were Retool, I think a step in the middle, if you want, is something where you ask a chat to do something and then you see if it works.

And once you see if it works, you can do something on convert this to a UI or convert this into an internal app. So now it knows the process, it knows the process has worked correctly, and it builds a UI on top of that because it's going to be a repetitive task. The problem with Retool one is that a lot of internal processes and things that I do are repetitive.

I don't want to ask on text every time I want the boot on an app where I can just do it if it's ongoing. So maybe something in the middle where you can chat with it, but you can actually convert a chat into a UI eventually could have been, I think, something in the middle where it's still more futuristic like Retool one, but still allows you to build apps. 

Having said that, very interesting that this positioning that we're talking about now, it took some time to understand.

Like I don't understand why Retool and Superblocks don't talk about this more. If I were Retool, I would talk a lot more about, “Hey, our idea is you don't need to build internal apps anymore. Our take is you can now directly chat with something and get the data directly.”

Superblocks take is different. But I think Retool messaging and positioning was actually not good because they didn't highlight their different take on the market and how that's changing what they are shipping. Superblocks, I think, was much better when they did the positioning than Retool.

Hank: Retool wrote a lot more words about their product, but they said a lot less. And you know what? I'm going to take us on a couple of tangents. 

First, small tangent. I'm seeing some em dashes in the Retool blog. You know what an em dash is? That's AI generated now. I don't care if you were a proud em dash user before.

If a human actually wrote this, I will be surprised. And if they did actually write it, then they did not think about the implication of an em dash. 

Gonto: And if you're shipping a feature, write your own goddamn blog post. You can use the AI for acquisition awareness one, for a launch blog post. Get a human. 

Hank: Yeah. It better start from, I mean, just take the video. He made a three minute video, so they wrote a script. Take that script and expand it with AI.

That's fine. But, okay.

Gonto: Exactly.

Hank: So that's the first tangent. 

The second thing is here, and again, relating to last week, because we talked about this is another set of companies going multi-product or expanding their product line. And so I'm just trying to think, what are the principles for the people listening to us right now as they think about products and they think about AI products? I think one principle we've talked about here is “what is your existing customer base desiring?” What's the next step for them that makes it easy for them to not only adopt, but to trust.

I really love that you use that word. I feel like Superblocks customers because it produces from a prompt, that's the new part, a UI that's familiar and they can verify, validate, test, and iterate on themselves. That makes their adoption curve better.

It'll build more loyalty with the customers they already have and love while also expanding their ability to get more customers and get rapid adoption. Are there other principles here? I mean, do you agree with that? Are there other principles here we should pull out for our listeners? 

Gonto: I agree with that. Like for sure. The other principle I think is you have to make your positioning on the future and why you're building your products a lot more clear. Like Retool, I had to guess by looking at their video on the landing page, what's their take on the future. They were not clear about it.

So I think making sure that you're clear about that, something very important. And as you said, thinking about who your customer is for what you're building, I think is key. 

Hank: Yes. I'll say one more thing. Super blocks did a neat little bit of counter positioning against all the main app builders. Yeah.

Yeah. V0, Bolt, Lovable. That was how they opened their video.

They're like, we've got 60 million bucks to do this. They show this iceberg. They're like, those guys are the tip of the iceberg.

Here's all the stuff they can't do. Great counter positioning. We're taught to be resistant to talking about competitors and that's usually smart.

But when you have a clear counter position, I think that's the way to do it. To clearly state, here's what the others can't do that you're missing and how we're addressing it. 

Gonto: And they nailed it. That's the question people have like, oh, yet another one. And they did it absolutely correctly. 

Hank: I'm, I'm interested in trying this now.

Whereas the Retool thing I've seen like, I've literally been pitched five of those exact thing in the last quarter. 

Gonto: I know. And more are coming. We'll see which wizard of those wins. I know of a lot more coming. 

These two products, Superblocks and Retool had videos that were professionally made to do this launch.

And if you look at all of the YC companies and actually any company now, they are literally all doing videos that are high production. Some seem like a short, some things seem like something else. I saw a tweet about this.

That was absolutely incredible showing like the video. And he was like, oh, the video is perfect. And then the product.

And he's like, yeah, it sort of works. 

Hank: Yeah. The horse.

Gonto: Exactly. 

Hank: Yeah. The horse image where the back of the horse looks super beautiful.

That's the video. And then, well, hey, however long ago, what, nine months ago, when we started the podcast, we were all for, hey, everybody should be doing video great.

So don't you love this? Don't you love all the videos coming out? 

Gonto: Well, my biggest principles on this is it's better to be different than better. So I was a big fan of videos because nobody was doing videos. And we saw some videos like, holy shit, these high production. It's very clear. It shows the CEO. It has some messaging, but now that everybody's doing videos.

Should they do something different? Like yet another fucking video. It's the same as everybody else do something different. But I don't know.

You are launching a video, I think soon. What's your take on videos? 

Hank: Yeah, I've got a video coming out right after this podcast will publish next week. Hopefully, if the product team delivers, we'll we'll put the video out.

I think videos. I mean, look, there are some things that become the new baseline. If you launch a product, you better send out an email.

You better write something about it on your website. You should also tweet and post about it. So you can't just say, well, because everybody's tweeting, don't even bother tweeting.

Gonto: That’s true.

Hank: But let's look at the, I think the most true thing you've said here. I think the thing that you have to agree with yourself on no matter what that you won't be fickle about is it's better to be different than better. Right.

So there's still so much differentiation you can do within a video. I think even going back to those examples of Superblocks and Retool, Superblocks, yeah, they did a very basic opening. I was deflated immediately.

They were like, oh, okay, you raised money. Cool for you. But then they did a little counter positioning work and they showed a really cool visual that, you know, there was a metaphor, the iceberg.

And guess what? I've seen the iceberg before. Me and Rouch G used the iceberg at Vercel in one of my videos. 

Gonto: I used it at Auth0 as well.

Hank: Everybody uses the iceberg, but they used it in an interesting way for counter positioning.

And that I leaned forward when that was on the screen. Right. And I think that's where you can think about videos differently, where you can think about your videos is how do you make them different and just better? Yeah.

Gonto: I do think that some of the videos you can do different than some are the CEO talking, some are more like shorts or something similar that explain an emotion or something that's going on. So in that sense, I agree you can do different type of videos. 

The other thing that you got me thinking was I agree that now you have to do a video launch because it's now table stakes, like everybody has a video, everybody has a landing page.

So then my question is before doing a video, what's the holy shit? This company is doing something different on their launch and doing something extra. That's awesome. And we talked about it because that was the extra thing they were doing.

So my question now is not everybody's doing videos. What is the extra thing you will do for your launch? What is the one extra thing that nobody else is doing, but you are? Is it an interactive game? Is it an experience? Like what is it? But I think adding something else that makes you unique is something that's really good. The problem is going to be that if you do it so good, people are going to copy you. And now there's one more thing. 

So product marketing launch list ends up being huge because now everybody's doing something different. But I do agree with you that you can make a video a bit different, but I would think about what are the experiences or what are the things can I do in the launch that not everybody is doing as well? 

Hank: Yeah, this is like basic product marketing stuff, but not every startup like thinks about this and develops their framework of what's your tier one launch, what's your tier two launch, what's your tier three and which items go into those, you know, for your tier one launches, there's a new baseline, you better have a video.

And you should try to have that something extra, you know, what's the extra bonus thing that has a chance, you know, most likely, you're not going to go viral, whatever. But you should always have something that could, if you have the time and effort. 

And if you don't have the time and creativity for that, you should think, well, at least we have the baseline, if we're going to do all the things, we're going to think deeply about the language, we're not going to write our blog posts with GPT. And, you know, with all its delves and em dashes. 

Gonto: And yeah, it's not delving that much anymore. Now it's em dashing more than delving, I think.

Hank: Yeah, I had to add it to my custom instructions of just like, hey, like this latest model just loves em dashes. 

Gonto: Yes. But if you ask it not to use them, it uses them anyway.

Hank: So I mean, you're right. Like, here's another lesson for the listeners. Once Gonto is sick of something, that means it's table stakes, and you actually have to do it.

And yeah, so you have to launch with some sort of video now, some sort of video content, and start thinking about how to make it different or do more. That's my strategy for next week. Our video, I think is going to be okay.

But we have a few videos similar to our cloud launch, you know, it's actually slightly smaller than our Laravel Cloud launch in many ways, but I like the idea of doing something different, besides the video. I'm not saying you shouldn't do the video, but it's more like, what else are you doing? 

Hank: Yeah. 

Gonto: But anyway, we're back to weekly cadence.

Sorry, we missed it last week because of the holiday. But we hope you enjoyed our Superblocks versus Retool, and our rant on videos for, well, my rant on videos for product launches. Thank you, everybody. Until next week.