Monthly Community Update -- December 2025

Announcements
1

Hi everyone,

This is the December community update. You can read November’s update here. This month I’ll be covering the usual product updates, but I’ll also be talking about what we’ve heard from you via the Ideaboard and a big shift that’s been happening behind the scenes on our team (more on that below). If you want to know more about anything in this post, join Emmanuel for his live AMA tomorrow at 12PM ET. You can RSVP here.

Now onto the updates.

What we shipped in November

We’re kicking off a builder-centric initiative called LaunchLab, a program designed specifically for builders getting ready to launch. We’ll partner with you to showcase your product to our global audience. If you’re gearing up for launch in the next month or two, read more about how to apply here.

We just launched AI expression generation in the Bubble AI Agent, the first of the major updates we promised at Bubble Tour. The Agent will automatically create dynamic expressions when you ask it to build elements that require them, and it can also edit existing expressions. This means you can ask the Agent to do things like “Generate an expression for this repeating group that searches for a list of tasks completed by the current user.” You can read more about this update on the forum here.

We’re also continually implementing core AI Agent reliability updates, like latency improvements and reducing false “success” claims (that is, the Agent telling you it did something it didn’t actually do).

To improve the mobile deploy flow, we introduced a validation step when connecting your Bubble project to the Apple App Store Connect and Google Play Console developer consoles to ensure the pasted keys and IDs are correct before your first deploy.

We’re still focused on mobile reliability and stability work, and the team is continuing work on bug fixes. We resolved an issue with OTA updates and real time updates on dedicated apps, fixed display issues with horizontal lists and webviews, addressed a camera access issue on Android, and made the interaction between new deploys and the update app screen more reliable.

Lastly on the mobile front, we just completed a pretty big upgrade to the latest version of React Native and Expo. This not only improves performance, but fixes a handful of bugs and makes us compliant with Google’s 16 KB page size requirement. You can read more on the forum here.

The new brand guidelines page is live! You can use it to access assets like logos, hex codes, and more, which is especially useful if you’re a community member who hosts events. Check it out here and read about the process behind it from our Design team lead on the blog here.

For a complete list of additional updates and small bug fixes, check out our release notes page.

What we’re working on

  • Property editor: The property editor alpha is underway, and we’re targeting mid-December to start releasing an opt-out beta to our whole user base. This update addresses several points of friction like displaying color variable names in the color picker, giving users the option to name conditionals, and drag/drop to reorder. It was also built from the ground up to make it easier and faster for our team to iterate on moving forward. You can check out the preview on the forum here.

  • AI Agent: Now that the Agent can create and edit expressions, we’re moving on to having it create data types after initial generation. That means the Agent will be able to make changes that address the front- and backend of your app.Once that work is complete, we’re aiming to release workflow generation, which will make the agent truly full-stack. Currently we’re planning for the end of January.We’re also working on better element deletion. Right now the Agent hides elements, which isn’t as tidy of an experience.

  • Mobile: Outside of performance and reliability, in-app purchase functionality is the team’s top priority. We’re aiming to release the ability to process subscriptions through Apple and Google’s in-app purchases by February. In case you missed it, our mobile PM @nick.carroll shared a demo and a timeline update on the forum. You can check it out here. On the performance and reliability front, the team is focused on addressing a handful of rendering issues, like missing icons and data, which tend to be more common on Android.

  • Academy: We’re wrapping up the first phase of our Academy updates this week. This redesign phase will make it easier to find courses. Phase two will make it easier to track completion.

  • Experts: Alongside that work, we’re making it easier for users to find coaches by adding a rating and review system. Users will also be able to book with coaches directly, so you can better manage payment and booking with anyone who hires you. Those updates should be in your hands by the end of this week.

  • Typifying the editor: Behind the scenes, we’ve been migrating the editor’s source code to Typescript to make updates easier and more stable for our internal team to execute on. That work should wrap up in December.

Ideaboard recap

Since we updated the ideaboard in September, we’ve been digging into your submissions. Your feedback impacts what our team tackles, and even if we didn’t address an idea you had this quarter, we’re still paying close attention and have more improvements planned very soon. Here’s a recap of some of the most common themes we heard and the work we’re doing to address them:

  1. Native mobile unlocks a lot of possibilities, but there are some feature issues that are blocking more people from feeling ready to launch: We shipped a mobile-only password reset flow, deploy settings validation, mobile logs, and AI mobile app generation (beta). Next up is in-app purchases, continued performance improvements, and the full rollout of AI mobile app generation.

  2. You want more intuitive visual controls when modifying web and mobile app layouts and styles: We made improvements to the table element (and moved it out of beta) and made it possible to generate and edit designs/new pages with the Agent. Next up, we’re redesigning the property editor.

  3. Some new users find the data tab and API Connector confusing: We made UI updates that make it easier to start generating data types with Bubble AI. We’re working on the ability to generate data types directly from the Agent and improvements to the API Connector’s visibility in the editor.

  4. You want simpler, safer security defaults and clarity around best practices: The security dashboard is now easier to access directly within Bubble, we updated privacy defaults for API workflows, and we added a security check for leaked API keys. We’re continuing to integrate the dashboard more thoroughly into the editor and optimize its performance.

  5. You don’t want to worry about platform stability and want to feel in the loop if and when outages do happen: We updated our status page to improve real-time incident communication, made improvements to internal observability, and removed outdated services that handle Bubble’s traffic and database. We’re reworking our deployment pipeline to be safer and faster.

  6. You want to solve latency and debugging issues when your app is scaling: We improved our database index management and introduced a faster, more scalable data service layer. We’re continuing to rework how we store and retrieve global metadata, which paves the way for further improvements.

What else is on our minds

I don’t typically share many team updates beyond new hires in these posts because my focus is the product and community. But a few weeks ago, Emmanuel and I pulled the entire team together to lay out our vision for the next five years and the changes we needed to make as a team to bring it to fruition. We shared some longer thoughts on the blog which you can read here. If you’re a regular reader of these posts, or have ever commented asking about the “why” behind product updates, I think you’ll find it interesting.

ICYMI

We have quite a few community-focused initiatives happening right now.

  • The Bubble Community Awards are back for 2025! Vote for the winners of categories like Agency of the Year here by December 8. We’re announcing winners alongside BubbleWrapped later this month.

  • Congratulations to the four Immerse finalists who took part in Demo Day. They pitched their AI apps. Check out the recap and see who took home the top prizes here.

  • Congrats to the top 3 winners of our AI hackathon with Anthopic: Tribe AI (Best Design), ChartPanda (Best Use of AI), Taskville(Most Traction). Watch the recap short here.

  • We launched #BubbleIsGlobal, an interactive map that shows the top Bubble apps in different countries, starting with Europe. We’re adding more regions very soon, and you can check out the map here.

New hires

We had five new hires join us in November, and we’re going to be hiring for a lot more this month and in the new year. Welcome to:

  • Jessica, account manager

  • Sage, software engineer II

  • Steve, software engineer II

  • Emily, marketing and events coordinator

  • Victoria, executive assistant and people ops admin

That’s all for this month. Wishing you all the best headed into the new year.

— Josh and Emmanuel

20 Likes
3

I’m curious what you did by giving them a choice. Did you switch them all to salaried workers and expect more hours from them? Not really clear what the choice was from reading your post. Were some of the 5% that left the company a large part of it? Engineers?

And then we asked them to make a decision.

We asked everyone to reflect on whether they were truly confident and excited about the future of this company, and whether they could commit to doing what it takes to make Bubble everything it can be.

If the answer was yes — if they wanted to be part of building the future of software development here with us — then let’s do this together.

If the answer was no — if their career priorities had shifted or this level of commitment wasn’t the right fit anymore — we accepted that with respect and gratitude for their contributions. We provided financial support so that anyone who chose a different path could find something better aligned with where they are in their career right now.

This wasn’t about forcing anyone out, and it wasn’t about reducing headcount — in fact, we’re going to be aggressively hiring in 2026. It was about ensuring that everyone who stayed was genuinely choosing to be here.

2 Likes
4

My guess, and strictly my guess from being in business…

the 5% are usually the workers who have been around a while and don’t want to realize or deal with changes.

I would say 5% is pretty low; there could be those who are just staying on for a job.

Bubble has no choice but to pursue AI.

They have to do it to stay relevant.

If they don’t, they’ll lose.

I think Bubble is on the right track, and I honestly believe their AI could be the best out there…especially given the fact that they also have a visual editor.

Just my thoughts

5 Likes
5

The one thing stopping me from the mobile native version is the complete lack of plugin editor, and the near total lack of documentation. At the very least, I would expect working plugins, since it’s a key part of the ecosystem but without it - I can’t even get the basics of my app on mobile native.

Otherwise, an interesting move forward on the platform. I have tried the AI at several stages including recently, and felt everything it created was so bland and dated, I gave up and did it from scratch. I think the biggest issue was that creating a new page needs to be “boxed” into a category, such as landing page, social feed etc. - but I know exactly what I want, and it isn’t any of the options.

I’m a little confused at the API connector changes though - it’s an advanced feature, and usability was never going to be the strong point. Functionally, the API connector fits almost every requirement, it’s just the interface and the core that is buggy AF. As a plugin developer, I spend way too much time debugging issues solely caused by long-standing bugs in the JSON→bubble thing conversion code.

And my perpetual frustration - documentation. For plugin developers, it is near non-existent (everything I learned is from dissecting existing plugins and rantings from others in the forums), and for the mobile native, it is more like a pitch than anything else, covering a whole lot of nothing. If you’re planning to improve ANYTHING, please make sure the final gate to release is comprehensive documentation that covers every aspect of the feature!!!

3 Likes
6

Genuinely curious how you’re able to measure this $1.1B mentioned in the blog post? Doesn’t seem like something you can measure in a meaningful manner.

Today, 6 million builders have created on Bubble. This year alone, over $1.1 billion in transactions has moved through apps built on our platform. These aren’t prototypes gathering dust. They’re real businesses creating real economic value.

7

Thanks for the update again Josh, can’t wait for the new property editor!

I want to share my two cents about the AI aspect of Bubble. Some things might be oversimplifications and maybe wrong but just wanted to share what’s in my mind for so long. I run an agency and we use couple of no/low code platform beside Bubble. We used almost every popular new AI tool or updates of existing platforms (Airtable, Framer, Webflow, Lovable, Replit, Bolt, Webstudio, Spline, Figma Site/Make etc.) for the last 1-2 years and on track to try upcoming ones too.

I get excited when new things added to an existing platform that can make my life easier. It gives some sort of inspiration when working with it. Also trying new tools which makes something different then my existing methods is exciting. Seeing new ways tickles my brain.

These feelings was almost non-existing for AI tools tho. Problem was, it didn’t make my work easier most of the project phases, or gave me a relevant inspiration. Output was mediocre at best, and not enough to use without modifications, which made the process longer. It sometimes provided the solution that I don’t have any or little knowledge of, and I trusted that it was the right solution. Edge cases of this solution mostly failed and needed manual work. This issue is obviously different for each user knowledge level.

Below are my simple breakdowns of these levels with observations why Bubble’s approach can be better. These breakdowns are simplest versions and doesn’t separate coding, designing, and product architecting experiences, which actually shows the topic is not that simple. Behaviour of these user levels are almost the same for other platforms too, it’s not Bubble, it’s the whole AI movement.

No Knowledge: Have an idea, and have no idea how to bring it to life. Have energy and motivation to start learning new things to achieve their goal. Most suitable level to use AI features of Bubble. Issue is, they try it and leave after some time most of the time. Large portions of these new users are here because of the hype, and when they see a new tool, they depart without thinking. If they stay a bit longer and do make some progress, they realize developing and designing a product is not the only thing they need to tackle. Only small portion of this cohort is staying. Obviously Bubble is focusing most of its efforts to this level, ergo implementing AI features to make it easier for newbies is the priority. But this level is financially unreliable in the long term, and probably won’t return the investment made for their ease of onboarding. Bubble should make this level of users to go one level up easier, instead of giving them the idea of shortcut to success.

Beginner & Mid-level: This level is the biggest user base in Bubble. They have a goal and have some knowledge how to achieve it. They can’t create an app ground up without help so AI features are pretty useful to them. But they never build an app ground up with AI too, because it fails on deeper levels of the app features. This user base is actually probably the most profitable. Some gets bigger and switches to higher plans or maybe even to dedicated server. So prioritizing AI features for this user base is not logical, since they are using it as a sidekick. Instead, improving existing features or adding new most requested ones will help this level to stay in Bubble. So there is a higher chance of getting back the investment made.

Senior & Full-stack: This level can build a Bubble app ground up with all the bells and whistles without needing a single AI help. They maybe use AI as a utility tool. Convert csv to json, write a simple js, check this PDF for user credentials etc. AI has no significant effect on their pipeline. Prioritizing AI features for this user base is not logical too. They care about platform reliability, scalability and speed of implementing new features which in return earns Bubble large apps with maybe some unicorns in them.

TLDR; Bubble should lower the priority of AI implementation to third of fourth place. First place won’t satisfy the financial and commercial expectations since it depends on the flimsiest user base. Instead improving existing features and adding most requested ones will earn Bubble a more reliable user base.

9 Likes
8

They have contacts at Stripe and Stripe has metadata of where requests come from (as you’d expect). It’s not going to track anywhere close to all revenue, but would certainly measure a good chunk of it.

4 Likes
9

Niace. I eagerly and nervously await the rollout of the new property editor.

At the same time I shed another tear for another update with no hints of updates to the plugin editor.

2 Likes
10

Partial AI is the sweet spot. This is my condition I want to add x is y in the middle of the condition can you do that or check this option set, I am giving you a list just match them add on attribute part. or I have done this in desktop but it is very messy on mobile, can you reorder without breaking workflow to be compatible in mobile. If it does that major time savor. But I don’t know it feels too advanced

2 Likes
11

To be very honest, I’m becoming pessimistic about the current state of mobile features. I understand that AI has become a top priority, but essential components like dropdowns and other fundamental elements are still missing. On top of that, the lack of offline database capabilities severely limits the launch of many applications — especially internal management apps and apps built for countries with regions of low network coverage. Until these features are released, I don’t feel comfortable launching mobile apps using Bubble.

8 Likes
12

Please, team — we really need more attention and development resources dedicated to Native Mobile. There are several recurring bugs, and support often responds with: “Sorry, our team is small, hope you understand.” I do understand, but we also need clarity on how long these issues will remain unresolved.

A very simple example: input fields on native mobile are not formatting correctly. If I set the content format to “percentage” or “currency,” it simply doesn’t work. Support asked me to be patient, but it’s been over a month with no updates.

@nick.carroll please help us with this. My customers are complaining a lot about this issue — it’s a “small” detail technically, but visually it affects the entire user experience.

3 Likes
13

I’m curious how much Bubble will need to charge users for AI usage once the AI companies are forced to adjust their own pricing. AI companies lose money currently, as some estimates show an example of a $20/month OpenAI subscription can actually cost upwards of a couple thousand dollars a month in compute costs for highly active users.

As these AI providers need to adjust pricing to reflect realities of their “businesses” (I use quotes because I believe loss making enterprises are not really a business), how is that going to affect companies like Bubble that sink so much of their resources into building on top of the AI service provider infrastructure. Is bubble going to be profitable at $30/month starter plan for an app being built by AI? I don’t believe so, unless they charge for AI usage.

They say 95% of all companies that have implemented AI have not seen a return on investment and that’s now, at the stage when AI companies are basically giving away the product just to get people/companies hooked. Are AI companies taking a page out of the playbook of crack dealers; first rock is FREE, then you’re hooked.

7 Likes
14

I agree.

15

AI companies can increase WTP, decrease computation cost, get ads to show their users,

Back in 1970’s 60’s people where buying equity solely for dividend gains, cash holldings and health of the company. And a young guy came and bought stocks like disney or heinz which is not giving out dividends and dont have that much cash or in some cases burning cash. So a guy came and bought them believing they will make cash in the future. So basically free cash flow idea but implemented in 60’s, the guy called Warren Buffet. I am not saying he is well educated at the end of the day he is a “Classical Economist”.

Then the “Creating shareholder value” book by Alfred Rappaport is written which defines modern finance practices in 1992 I believe. So In practice 60’s and academically 92’ is when cash burning companies begin to considered to be real businesses. So, your idea of not considering them as actual companies is 75 to 50 years outdated.

So real problem here is In todays world U.S government is spending ( creating from thin air ) too much money, Main contribution is being 1) Interest rates 2) Military spending. And that is going on since 2021, it tends to create bubbles, and continuation of those bubbles, like Bitcoin, like AI, funding never stops, and we are not even at the peak of the bubble, due we are in the beginning of the credit cycle (people or companies getting in debt is the other main form of money creation from thin air), overall interest rate reduction will reduce the gov spending but this trend will keep on going for at least 1-3 years.

So the real question is when the bubble pops, do we have tangible AI products, the AI products will have the sufficient funding? and will be able to transition to low cost, high wtp products? and how long will be the duration of the downfall (2008 bubble downfall lasted till 2014, 2019 corona thing downfall lasted 2 years ) the duration interval shrinking but it is overall a concern. And how Bubble AI investments will be affected by that?

16

We can’t allow ourselves to be distracted by inadequate, albeit flashy, challengers.

We must not underestimate the power of “flashy".

When a user is looking for any piece of software to help them with their task, they’ll initially research & test out options. In these “first impression” stages, the full power (and flaws) of a platform won’t be fully apparent - but what does make a big difference - is how the app looks & FEELS - the UX. If it feels slick, polished, well designed… that really does make a big difference & gives confidence in the company behind it, even if unconsciously.

I truly love bubble but from the UX perspective it does ‘feel’ clunky and looks dated - even a mess - in places. (Sorry)

It’s great reading about all the new features but I do feel like you need to give the UX a massive priority boost, especially the little bug fixes. Make it slick, modern, lightweight, snappy & stable.

It may be seen by many as a lesser important aspect, but in terms of competing with the ‘new tools on the block’ & attracting new young users I feel it’s essential.

1 Like
17

How is it that you are paying close attention but do not regard the first and most voted idea in the Ideaboard? A PDF generator is a basic functionality, and the most requested one by your users. Right now we have to pay for basic stuff like this besides the Bubble plan.

AI has de prioritized many basic improvements that should be done:

  • PDF generator just mentioned
  • The expresion editor can’t do basic stuff, such as editing from the beginning of the expression
  • Workflows have a low timeout
  • API workflows can’t return values

It feels like you have stopped prioritizing features that matter to most users in pursue of AI features that bearly serve beginner users

6 Likes
18

Great read on the blog.

19

thank you all for the comments. I can help answer some of them:

Yeah, the Native mobile editor team went through a lot of changes and transitions in the second half of this year and that significantly slowed down progress in our roadmap. We’re hiring more and moving internal resources to get Native Mobile back on track to deliver on this front

Like @georgecollier mentioned, our partners at Stripe shared performance data coming from the Bubble ecosystem. Obviously it was anonymized and other measures were put in place to protect everyone’s privacy, but it still gives us an idea of how “productive” Bubble apps in production are in the real world.

In principle, totally agree. However, because this is a new surface area for bubble, we need to learn how to walk before we can run. Fundamentals like being able to explain why the AI agent built something in a particular way, decreasing latency so engaging with the agent feels snappy, and making sure consistency and reliability are super high are all necessary. If we don’t do those we can’t hope to be useful to mid-level or expert bubble developers. But again, agreed in principle that an AI agent that helps users further down their building journey is a superpower we want.

It will roll out to ~25% of users in beta late next week and you’ll have the option to opt-out of the beta experience if it’s not there. But it’s a beta so the plan is to collect feedback so please share if you get in that first wave.

What I mentioned above on Native Mobile editor team/roadmap setbacks. Hoping to get better soon.

Obviously we will have to look into this eventually but for now the team is focused on bringing to market the best AI agent we can.

Fair. I think the new Property Editor rolling out soon is a good example of modernizing the UI/UX of Bubble better for everyone. It’s not just that the frontend got reworked, but also all the work the team did under the hood to re-stack the tech behind it, how it handles data, modernizing the libraries, etc. So it’s faster to load, easier to maintain and better to upgrade in the future.

Obviously I can’t speak for the product teams and how they plan their sprints, but typically a larger initiative will get prioritized that then impacts all the smaller asks, grouping them in a cohesive way. For example: a PDF generator didn’t get built this time around but maybe next quarter the product team will decide to work on the plugin editor, and as a side effect, built a bunch of in-house plugins or functionality the users want. Hope that makes sense.

thanks all

keep the questions coming!

9 Likes
20

Yeah, that’s what I was wondering. Stripe is not the only source of transaction movement through Bubble apps. So that number is actually much bigger.