@robert_harris8 We originally planned to build Bolt in February 2024, but the AI models available then weren't capable enough, so it was a lot of trying and testing new things. When we finally gained access to Sonnet 3.5, it was exactly the technology we needed which was super exciting.  

But the first two months were incredibly challenging. We certainly had our share moments that were "oh sh**, oh SH**” hurdles that, at the time, made us wonder if this was really possible (GPU shortages that jeopardized our ability to stay online, servers crashing at 3AM, severe product bugs, etc.). 

As I mentioned in a previous answer, it felt like the movie 300, where our 20 person team was surrounded by tens of thousands- and our team did the impossible!!

@steveb great question! I think our vision here seems to be distinct from others that our aim is to to build a "power drill" - a professional tool that most homes have & anyone can use, but equally powerful for professionals themselves. Most AI products right now either target developers directly (cursor, replit, etc) or completely non-technical users (lovable, etc), this is ok of course but the problem we see with being on opposite ends of the spectrum is that you inherently limit the user's power over the tool and exclude vast segments from using it. I think this is the main reason Bolt continues to be the #1 tool by DAUs, ARR, etc, so this is where we intend to keep focusing product vision wise.

As for spending the cash... Lambos of course! :P But in reality, hiring. Supporting the customer base we have even today requires dozens more people, across product, engineering, GTM, success, etc (and we continue to grow rapidly!). To provide the level of experience and support we think is required for excellence, bringing on great team members is critical. PS we are hiring: stackblitz.com/careers

@warrio4 amazing! That's so awesome to hear:)

Our general plan for this year is to really pave out the story around integrations & workflows for Bolt. In the same way that Slack is such a powerful tool because of it's ease of use and vast integrations, we think the same will be true for Bolt, and there's a near infinite amount of work to be done there ha!

@mat_sherman Great question! I think what'll happen is we'll see distinct lanes emerge based on who they're built to serve.

Rather than a winner-take-all market, I see it more as products finding their specific audience and use case. That said, tools that can serve both technical and non-technical users while maintaining power and flexibility (like we're aiming for with Bolt) have a natural advantage in terms of total addressable market.

The key differentiator long-term won't just be the AI - it'll be the complete experience, integrations, and how the tool fits into existing workflows. That's why we're so focused on the integration story for Bolt this year!

@nora_lam short answer: we put out a tweet. That was it (seriously!)

Longer answer: building a community of builders & developers the past seven years, who were following what we'd announce and release. And we released a lot of things over the past 7y! So when the time came to announce Bolt, we already had a trusted brand, community love and willingness to try our products.

Tldr- while you try and build your products/startups, always be building a community. You need that distribution channel when the prime time comes, and its something that cannot be bought!

@abod_rehman Yes! Very glad you made it. But i'll definitely take the story on how you almost didn't...

@abod_rehman @steveb great questions!

what was the biggest turning point that kept Bolt on track during the tough times?

Hard to pinpoint exactly one specific turning point honestly. In the first 2 months, there were probably a dozen or so moments that were "oh sh**, oh SH** everything is melting" sort of situations. Like, would've killed the product/company sort of things: not enough GPUs to stay online, servers getting hammered and going down at 3am, critical bugs in the product, etc etc.

It felt like the movie 300, where our 20 person team was surrounded by tens of thousands- and our team did the impossible. I'm so proud of them.

the story on how you almost didn't make it

Long story short... we'd been working on StackBlitz (company behind bolt) for 7 years at that point, and despite making insanely cool technology w/ much developer love, could not figure out how to make money in a venture scale way. So at the end of the year, we were going to begin spinning down the company if we couldn't figure out a way to make our technology commercially viable. We had one last bullet in the chamber though, which was Bolt... and we are still mindblown by what happened next. (And fortunately, we did NOT shut down the company! :)

@gabe Absolutely, we've had countless moments where we were like "wow, Bolt just did that 🤯". Even today, we're still blown away by what it can do. What we're building with Bolt is truly democratizing web development and fundamentally changing who can build and how they build and that definitely feels magical.

The product iteration was pretty wild- it all came together really at the last minute before the launch. Of course it was cool to use before the launch, but ironing out the bugs, etc you couldn't quite get the full feel of it. So we ourselves were pretty mind blown on launch day ha!

@owenoktay A key differentiator is going to be the entire user experience, from prompting, to integrations with other tools. We’re very focused on adding integrations and making the experience of using them very smooth 🔥

@albert_santalo It has really been community, community, community. As I mentioned above, StackBlitz has been around for 7 years. In that time we have been building a community of builders and developers through our open source initiatives and other community-driven efforts. We had built a community who were following our announcements and releases over the past seven years so when the time came to announce Bolt, we already had a trusted brand, community love, and willingness to try our products.

Having a community is essential if you want to build products/startups. You need that distribution channel when the time comes, and it's something that cannot be bought!

@jakecrump Plan out your app before starting: What I have seen work well is making a plan and having a clear outline of your app's structure, including which pages you want to exist and how users will navigate through them.

Then, break these pieces into manageable tasks. For example, start with the UI rather than throwing an entire PRD at Bolt in the first prompt. 

What we’ve seen work well, is building incrementally. Build your foundation first, then add the fancy stuff! Make sure you've got all the basics scaffolded before adding more sophisticated functionally.  

@admiralrohan We built WebContainers for 7 years, but couldn’t figure out a commercial application for it. 7 years is a very long time ha! So we were going to begin spinning down the company at the end of last year if we couldn't figure out how to build a meaningful business here. And fortunately Bolt ended up being exactly that (and beyond our wildest expectations)