Wait, you're still here?

I pop in once or twice a week to see what’s new. Without fail, there are always a handful of recent threads where ‘we’ are just letting the Bubble team have it; really airing our grievances.

With yesterday’s outage - among other recent issues (not to mention a hilariously timed price increase) - I’ve made the decision to move off of Bubble.

I’m not asking anyone to talk me out of it - but it did get me thinking… why did I last so long (over three years)? Why is everyone sticking around? Would I be leaving if not for the Cursors of the world?

Anyway… you’re still here? Why? Curious.

3 Likes

I know how the platform works and I’m not interested in having to invest the time to re-skill myself in a new platform.

12 Likes

Got to agree… Bubble is the AOL of website builders. You could probably replace any site built on Bubble with AI coding tools now in a few weeks. Some projects just a day or two.

I sometimes check back on the Bubble forum - i used it a lot in past years. Now I check back to see if people are still using it… And the forum still looks busy (though perhaps a slight dropoff). For people who are clearly smart and technical (Bubble IS technical, so if you can learn Bubble you can learn basic coding), I’m surprised.

It makes me wonder how many are stuck in their own bubble, excuse the pun, because a google/youtube search will show all the cool new tools now available. I encourage people to try them. (Sorry Bubble team.)

1 Like

Just one more model bro I promise the next model will make this happen just a few more weeks.

— Slop Builders every week since 2023

To answer OPs question, Bubble has a 1 year shot clock. If they can deliver loops, native listshifter functionality (mapping etc.), fix JSON parsing, add websocket support, and generally overhaul the platform so that the core functionality can accomplish anything an AI app might ever need, I’ll stay.

Some of the current limitations are just unacceptable and there’s no excuse for them. A glaring example is that Actions in a Backend Workflow always run sequentially and can’t run in parallel. This does not belong in a 2025 web dev environment. Even automation tools like n8n can do better.

If they continue to build mobile and AI stuff without making changes to the core “programming language,” I’ll just bite the bullet and learn javascript then migrate off.

5 Likes

This is just ignorant, I’m afraid. And what makes you think your site built with Bubble is so special and doesn’t look like ‘AI slop’?

I switched my very complicated single page app from Bubble to code-first in December last year - it performs way better it’s laughable to think the same performance / development speed can be achieved on Bubble. And you should see what coding tools can do now…

I don’t want to make those on Bubble feel bad, just urging people to explore other options…

2 Likes

Tell me you haven’t worked on large scale production apps without telling me

For technical users, Bubble is even better than AI-assisted coding.

The demand for senior Bubble developers has never been higher. I know because I’m trying to hire all of them, and the good ones are already busy with work. So, one could anecdotally claim that Bubble’s dead in the water, but the evidence wouldn’t really support it.

17 Likes

This :backhand_index_pointing_up:

Complicated large-scale production apps can’t be easily migrated off of Bubble, and frankly I don’t see a need to. If your app on Bubble was suffering from poor performance, that’s on you. With best practices, performance isn’t an issue. The recent outages on the main cluster are unacceptable, but if you’re scaling, our Enterprise-plan app has 100% uptime over the past 90 days. I’m hoping that Bubble will adopt Xano’s model soon, so that each app can have its own instance without that much-too-expensive Enterprise plan. It would solve most of main-cluster-app headaches.

6 Likes

It would be interesting to see the successful app with lots of users and great MRR that you’ve made, so we can compare that to the tons of successful apps built and used every day on Bubble.

This way, we could see firsthand your success using an AI builder.

So please share your site. It would be proof of you actually building something just as successful as you could on Bubble, and would help us see firsthand your reason for leaving

5 Likes

Man! This is depressing!
I have been working with bubble for over 4.5 years.
Happy with so far. By so far, I mean until I hit a road block with the mobile editor. Many functions and plugins are not Mobile ready. I cannot finish my Mobile version without them. I am talking about : file upload (video, audio) for example. Radio buttons anyone? Lists are long to load…
It could also be that I need to learn better ways to do things.

Because people compliment the design and feel of it.

Your app is held together by tissue paper. You don’t know what you don’t know.

Codex/Claude Code weren’t even around in December of last year and they still aren’t good enough now.

I still use Bubble for my prototypes, but not for production apps that rely on uptime that I cannot justify a dedicated server for.

I’ve been here since 2017 – The editor is in a better spot than ever.. but these slip-ups with uptime, after all of these years, is a sign of deeper tech problems. We were having downtime just as often then.. Then, it was more understandable, the platform was still in its infancy.. Once the $100m investment came in, that was supposed to solve all of these problems.

It’s a shame because I really enjoy using the platform. The editor is awesome. If you can’t justify a dedicated server, which does have significantly better uptime, then you’re better off building production apps elsewhere. (I did look into their dedicated servers.. It’s a solid offering if you’re saving the cost of a real development team.. My payroll is $30k/month USD for full time software devs. If 1 person can be the developer.. and build with Bubble, the $2.5k-$5k/month dedicated makes sense)

Edit: On the note of AI development.. Yeah, it’s getting pretty good.. But having some underlying dev skills can be a make or break. We’re using it now to supplement our work in VSCode via the Codex plugin.. For existing large code bases (200+ files, 1,000,000’s lines of code), it’s pretty solid actually. But it can lead you down the rabbit hole if you’re not careful.. It’ll put down a lot of code.. But you have to read it.. Otherwise you will eventually end up being in a spot where you fix one thing, break 3 others (just like real software development lol!).. So the note above from @georgecollier about large scale production apps is indeed true.

3 Likes

Don’t forget Bubble is launching AI in a couple weeks. I’ve seen it, and it’s very good at documentation style questions.

2 Likes

Not clear if this is a response to my OP, but if so, I’ll say upfront that I didn’t make my point clear.

There are a lot of assumptions baked into your comment (not to mention your classic passive aggressive tone @senecadatabase, never change :slight_smile: ), but here’s the quick(ish) answer:


TL;DR: I’ve built two real companies on Bubble (~$2M ARR with enterprise clients, and a practice management platform in healthcare). Bubble was great for getting started, but in both markets I hit ceilings:

  • Enterprises demand custom security, SSO, audit trails, and on-prem options Bubble can’t provide.

  • Healthcare requires compliance and infrastructure control that goes beyond Bubble’s HIPAA hosting.

  • Investors and acquirers see rebuild risk, since most engineering teams won’t work on Bubble.


I run two companies that wouldn’t exist without Bubble. One is an enterprise tool (~$2M ARR in 2025, with customers like Amazon, JnJ, and Stellantis). The other is in healthcare (a B2B/B2C practice management platform for therapists).

Enterprise: Perception aside, large companies expect custom security, SSO, audit trails, and on-prem deployment options. You can only go so far on Bubble before those requirements force your hand.

Healthcare: Even with HIPAA-compliant hosting, Bubble doesn’t cut it if you want to sell to large clinics or hospitals. They need proof of compliance, infrastructure control, and trust that goes beyond what Bubble offers.

Raising money: Investors are reluctant to back companies running entirely on no-code platforms. Some reasons are perception based, but others are very tangible. For example, if you build a big business on Bubble and a competitor wants to acquire you, their engineers will quickly realize the platform has to be rebuilt from scratch. No CTO is going to ask their senior Python developers to learn Bubble instead of continuing to build on a stack they already use. That is a deal killer.

General: I saw @georgecollier say Bubble developers are in super high demand. In my experience, high demand doesn’t always mean healthy growth. It can also signal scarcity (think: more non-technical founders are starting Bubble projects, but not more developers are actually learning Bubble).

Final point of clarity: I have nothing against Bubble. It’s just not the stack I’ll scale my companies on. That said, I’m (sooo) glad I learned it after Covid and I will continue to use it for prototyping and side projects.

1 Like

Well, you asked why we’re still here - because there’s a never been a better time to be in the Bubble developer market :slight_smile:

Yeah I mean, if you’re trying to build healthtech, then Bubble is a non-starter for you, so it’s understandable you move away don’t use it for that.

Our clients have B2B SaaS sold to some of the largest companies in the world and have got through third party audits/pentests etc. It absolutely can be done!

2 Likes

Interesting reply.

Though I’m a little surprised it seemed to trigger you.

I thought my question was straightforward: you mentioned leaving Bubble, and I asked to see a successful app you’ve built outside of Bubble so we could compare. That doesn’t feel passive-aggressive to me, just genuinely curious.

All 3 of my Production apps that bring in revenue would take me weeks/months to rebuild anywhere else, not to mention porting the data over on top of it.

I’ve been a software developer for 12+ years now, I’m pretty aware of how long it would take me to rebuild everything elsewhere. Sure AI exists, but I would be spending most of my time fixing everything it changes over and over again. Once an app gets past basic features, AI still can’t get it right.

Bubble is just very easy to work with. The data is here, the authentication is here, new features take me much less time to implement vs coding it out. I wasn’t thrilled by the price changes with Bubble, but just like me they have to turn a profit too. Right now as I type this, all 3 of my production apps are currently over the WF limit for the month, but the extra costs me nothing compared to the revenue they bring in, so it’s fine.

Now the outages are an annoyance, but the outages are a result of improving the infrastructure and my customers know that. A few years ago I lost quite a few of them because of it, but now I have a different client base and they are very pleasant to work with and I’m very lucky in that regard, however the outages are temporary.

Again, I really think Bubble should implement their own payment platform and take a % of revenue from that to make more revenue on top of the current system, but that doesn’t seem likely any time soon.

9 Likes

And not a real possible solution because of API connector and custom plugins. This is already an issue in some Plugins that instead of using the Bubble payment solution, use their own payment solution…

same here… i’m holding out hope that i can weather the storms, and not lose customers in the process

I’m here for one main reason: my time is short these days.

Spiritual health, physical health, family, children, work, family entertainment, personal side projects, and a few other things related to adult life. Try to maintain all of these with an adequate level of dedication, and you’ll see that your time is the most important thing in your life, that it never comes back, and that it becomes scarce as you get older.

I’ve been at Bubble for about 5-6 years. I have a programming background. I build my own things with code when necessary. I enjoy challenges. I wish I could do more, but again, time is short and life flies by.

Maybe if I were younger, living with my parents, being supported by them, or single with not much to do, I might invest in testing every new AI tool that emerges (many jump from tool to tool, become mediocre at them all, and never focus on anything) or even delve deeper into traditional programming. But… that’s not my case. At the end of the day, for the vast majority, it’s not the tool that counts, it’s the solution.

Therefore, the lack of time prevents me from focusing on delving into all the areas required in traditional programming, such as security, software architecture, databases, scalability and performance, code quality and maintenance, version control and collaboration, monitoring and observability, infrastructure and deployment, APIs and integrations, and so many other areas and their subareas, just to be able to deliver a robust, scalable, and, above all, secure system. Of course, in traditional programming, this is a team effort, but that’s where other problems arise. It’s not just the cost of a team, but other issues that come with working in larger teams.

And don’t get me wrong, in Bubble, even though we don’t need to delve as deeply into these areas because Bubble already does a lot for us, we still require some knowledge and the ability to put these things into practice. So, it’s not as if we don’t know or don’t need it; we just don’t need to delve into the deeper layers of it all.

Furthermore, while many people these days believe that AIs are creating applications from scratch, completely secure, scalable, etc., they will realize over time that at some point things won’t be as simple as they seem. Technical debt will arise, and they will need more technical people to fix the mess, if there’s still time. We’ve been seeing numerous discussions about the problems of interacting with AI tools for a long time. They work well up to a point, but then, with each necessary iteration, while one problem is solved, three more are created. If you don’t understand the code it’s writing, or only understand it in parts, how can you identify what’s wrong? Even some expert programmers get lost in certain concepts created by AI and the way it “thinks.” Who knows, maybe things will change in the future; by then, I plan to be retired.

Bubble, like many other low/no-code tools, has helped many start from scratch; many have changed their lives because of these tools. They all have their applications and limitations. Just as the solutions created with these tools have their distinct types of customers. And like all tools, Bubble also has its problems, its flaws, and its strengths. Many things could be improved, and others that, for some, don’t even really need to be implemented (I highly doubt that Bubble’s experienced or long-time developers are using its AI, at least for now), but hey, you can’t please everyone.

I see a lot of people complaining about Bubble’s AI. The point is that everything has a simple and flawed beginning before becoming something big. But I think the main problem is that people want to compare AIs out there that have access to the entire internet, can use any library, any block of code found in the darkest part of the internet, in other words, a completely free environment.

With Bubble’s AI, which operates within a controlled environment, that is, Bubble’s own architecture. In other words, theoretically, it can’t create a component from scratch unless it exists within Bubble’s “framework”. Both approaches have their pros and cons. For non-technical people¹ (which is why Bubble was initially created), which approach do you think would be safer?

Is it safer to let a child play freely on a street, full of risks and dangers, or to let them play in a controlled environment, supervised by specialized people?

¹ Non-technical people doesn’t mean people without knowledge. These days, it takes a lot of knowledge to create a truly functional and secure Bubble app.

I also get pissed off sometimes, with unexpected bugs, features requested years ago that haven’t been implemented, this and that, and of course, the main thing that affects everyone: crashes (the ones that happen in production versions, because we can still deal with the editor crashes). They’re not always recurrent, but they’re the kind of problem that, when they happen, makes most people angry and forget about Bubble’s strengths.

It’s like a marriage, where your spouse has strengths and weaknesses, and you live with them, but sometimes something happens that makes you forget their strengths and become extremely irritated. Some are more patient, they relieve them, they know there are good and bad days, while others already think it’s a reason to break up.

Well, I think that’s it. That’s my opinion, given my reality and life story. Nowadays Bubble is my job; it allows me to work remotely from wherever I want, have flexible hours, be present in my daughter’s growth, spend more time with my family, travel whenever I want, among other things. And I intend to stay here for a few more years, trying to help improve the tool, criticizing when necessary.

I understand those who have clients who depend exclusively on the solutions you’ve created and who get incredibly upset when something goes wrong. Or those who want to maintain a certain relative dependency by not being on Bubble or other locked-in tools. Ultimately, we’re simply changing our dependency source, sometimes they’re just not very explicit.

And finally, I’d like to share a small example that everything outside the “Bubble” isn’t perfect either. Everything in the world is imperfect; the question is what bothers us.

13 Likes

A clarification: I'm asking genuine questions here, not intended to be passive aggressive at all :) I am currently happy on bubble but my app requires bridging outside of bubble in select areas, so I have been contemplating the merits of moving fully outside. I am not a trained SWE, but I’m pretty capable at reading and writing Javascript when needed, so I may have a "higher" starting point than some folks for moving outside the ecosystem

Do you think any of this has to do with how you are working with the AI tools? I tried a little while back to “one-shot” rebuild of a piece of my bubble app on bolt and within maybe 3 prompts, I got stuck in an unbreakable error loop. I 10000% agree that trying to ask AI to just “build” a complex system on its own will fail quickly.

However, I have pieces of my app that live outside of bubble on custom javascript (specifically, a webscraping microservice that required architecture bubble can’t support), and I use AI in a very focused manner to help with writing and modifying code, and I have had solid success with it. Basically, with good coding practices where things are adequately DRY, I can give the AI only the single function or set of functions related to a given task and work in that directly, rather than the entire codebase at once. Like working with a real engineer, I might also have to explain how these pieces fit into the larger system for context, but I keep it focused on the particular task at hand

The AI absolutely still hallucinates from time to time (I find asking it to link to docs for each function it uses will help reduce that), and I often do need to read the code and spot issues / make small changes, but it is able to spit out working code / requested changes really rapidly. When I consider how my own dev time is spent, I am personally starting to debate whether bubble is still “faster” than working with AI in this way - for instance, working with “only when” conditions can be miserable and take hundreds of clicks in bubble when I need to make changes.

This aspect is the main thing that has kept me on bubble - I just focus on app functionality while Bubble is handling everything else. But, I myself am contemplating whether a bit of upfront cost to have a senior dev build the core architecture would pay-off in the long-term by allowing me to start using AI for feature changes, like I describe above.

Similar to observations above, would this not be the same problem if you just handed an engineer 6 months and an unlimited budget to build whatever they wanted? If you’re not along for the journey, of course it will be hard to understand what has been produced. But if you work on narrow pieces with focused context, you’ll be better able to understand what the engineer (or AI) has produced and work piece by piece to fix it?

——

If bubble does actually add an AI editor (not just page generator) in the near term that I can use to more quickly update existing systems in bubble, that might change my tune.

When you say “documentation style questions”, does that mean being able to explain to someone how a bubble app is working in natural language (rather than needing to decode the bubble workflows visually)? That would also significantly reduce my eagerness to leave the platform - a primary problem I have now is that working with developers on adjacent systems can be really hard since they go blank in the face when looking at Bubble (where with “normal” code, they could drop it into an AI and ask it for help understanding things)