Wild Predictions

bolt.new and lovable.dev have both grown at an impressive pace, reaching 40M ARR and 30M ARR respectively in less than six months. Why? what is driving this growth?

Interestingly, their primary users don’t seem to be traditional software developers. Instead, product managers, designers, and entrepreneurs are adopting these tools.

It doesn't seem like we are seeing a huge inflow of new apps created by these people. Or maybe we are and I just don't see it. What are these people building that they are willing to spend so much money on?

Prototyping as a multi million dollar business

One possibility is that these users are primarily building prototypes—specifically, prototypes of additions to existing apps. Why is that worth so much money? The only alternative these people have is the "traditional" route: writing a spec, having a designer create mockups, and waiting for a developer to build it. This process naturally takes more time and coordination, MUCH more. My best guess is also that as a PM, if you hand over a prototype of an addition to an existing app, the developer will be able to build it faster and more reliably than if given a spec. So it’s win-win. This fundamentally changes how software is built.

Prototyping, I guess, is now a killer app.

How far can this go?

One way to look at this is that Figma has a 10B+ valuation. Do you need a design tool if you can just use a prototype tool? I don't know, but I see quite a few companies hiring design engineers rather than designers. Given the pace at which AI improves will how long is there even a difference?

Software creation is changing, how about software usage?

Build your own app

If product managers can modify apps, should users have similar flexibility? Consider a model akin to Notion, where a composable set of tools is provided, and users can arrange them as they see fit. Of course, you wouldn't provide a blank canvas, but rather a default layout with some guardrails and sandboxing. Now, if a user wants to move a couple of buttons around, change colors, or adjust labels, they can do so. AI would make quick work of such requests, no technical skills required.

Would people find it useful? I'dd guess yes.

For starters, different people need different things. For example, designing apps for elderly people is completely different from designing apps for teenagers, even if the functionality is the same. Elderly people generally need more contrast, larger text, and bigger buttons. Teenagers don’t want any of that.

Then there is the subset of people who just want to control everything around them. Those people would clearly value this.

Another perspective is that there are people making millions of dollars selling templates to Notion. In fact, there are large marketplaces for templates in WordPress, Shopify, Google Sheets, etc. The list goes on and on. These marketplaces tend to make the apps they support better and more profitable.

Yet another perpective is that say I create an app that does something really useful, but I am a terrible designer. If I just ship my really useful app with reasonable composable features, then I can let the marketplace take care of some of that.

Another example is internationalization. You can have AI translate your app into different languages. But maybe parts of it are off. Let the users decide what the right label for a button is.

Existing apps are in a similar situation. Think about how difficult it is to make design changes to existing apps. It’s a huge amount of work, and sometimes the results are not great. Snapchat, Reddit and Twitter have all had to face stiff opposition to design changes.

Compare all of this to shipping a single static app that can’t be changed. You are essentially betting that whoever you have hired is better at doing what they are doing than the rest of the world combined. That seems like a very unreasonable proposition for the vast majority of companies.

This is all theoretical, though. Time will tell.