Has anyone else noticed a significant drop in Bubble.io jobs on Upwork recently?
I’ve been working with Bubble for a while, and I used to see a steady flow of new Bubble projects on Upwork. But lately, it feels like the number of Bubble-specific jobs has dropped quite a lot.
I’m curious about what others are seeing:
• Are clients moving toward AI-assisted coding tools like Lovable, Bolt, Cursor, etc.?
• Are they choosing other no-code/low-code platforms?
• Or is the demand for Bubble developers simply moving away from Upwork to other channels?
For Bubble freelancers here: have you noticed fewer projects, lower budgets, or more competition recently?
Would love to hear what you’re experiencing and where you’re currently finding Bubble clients.
Yeah, I have noticed the same as of late. My rationale is that a lot of the motivated clients that would reach out to a developer to build their project can now utilize AI and build the tools themselves.
I have seen a large trend of past clients and associates that are enjoying building their own tools without the back and forth of a development middle man.
I also think the rise of solo developers using AI to solve their bubble issues instead of going straight for a Bubble Certified Developer or freelance aid has also impacted this market.
If you have an issue, you can use the Bubble Agent, BuildPrint or Claude/GPT for a step by step fix on how you can solve it. No need to pay someone online for that type of service anymore.
…why need the junior developers? Just give a senior developer Buildprint…
Now, what we’ve also noticed is that people are using Buildprint badly in some cases and creating technical debt. So, the result is something like this:
the demand for average/junior Bubble developers is tending to zero, as is the same in traditional code, because of AI
the demand for talented Bubble developers is increasing, because they have to fix the mess created by AI, in applications nobody really understands
NQU has tripled in size over the last 12 months, despite everyone saying ‘oooh Bubble is dead’ or ‘oooh everyone’s leaving’. Like, no. It’s just that people aren’t offering a service people need and 80% of Bubble developers aren’t delivering sufficient value when compared to an AI that’s $49/mo plus a Claude Code subscription.
I’ve from the beginning specialized in complicated implementations in Bubble, and the need for that has not decreased. I guess the people hurt the most were making basic things that they could do over and over again for new clients.
EDIT: I think OP still has a point though. When I started, there were frequent job offers on the bubble forums. Now there’s a lot less, and there’s more developers now then when I started.
@georgecollier@jonah.deleseleuc I don’t disagree with you both, but that wasn’t really what @amer was asking about. This wasn’t a bubble is dead but what happened to freelance bubble work on Upwork.
The main type of client that would go to Upwork/Fiverr to find development help in Bubble was not typically the sort looking for
From my experience they were solo developers that either started a project and hit a roadblock or wanted someone/something on the cheaper side to assist them in either learning how to build in bubble or create an app/feature they were unable to.
Now, those can all be solved with the emergence of AI agents (Bubble agent including). Where for the low price of $20 you have in your pocket the best bubble developer at your command.
I am not saying Bubble is dead, but the type of development that was frequently requested on sites such as Upwork and Fiverr has diminished and will only continue to as these tools like the Bubble Agent improves.
I’ve actually haven’t really looked into the “upwork ecosystem” of bubble freelancers.
For those of you that have used the platform, mind giving me a TLDR on the typical experience? Things like: most common app types you see requested, typical budget ranges, how often are the leads “good” vs people just wasting your time or not really serious about building a good app, etc
Yes, less jobs posted and more supply. Crazy to see how much money is spent on connect points to bid for jobs. Saw one whose job posting got 4 bids for 300+ connect points each which is like $45.
It’s also clear Bubble Forum jobs are much less. Used to be a few every week, now I think might be 1 a month.
@fede.bubble any insight on numbers for Bubble RFP system? Are numbers down there? Are leads good or bad and typical budget insights?
I would say @georgecollier is right now that AI has impacted the Junior roles. That much was clear 3 years ago it would happen, clear 18 months ago it already started and clear now it is not reverting back to pre-AI, but a real Senior Bubble developer doesn’t need AI. AI is just terrible overall for implementing anything well. Not sure how that compares to a junior developer though, if they are any better, but sad truth is, AI is cheaper and faster than a junior developer.
What is happening is overall demand for Bubble is shifting. Not as many MVP builds because AI can do it, and I’m really happy Bubble picked up on that and leaned into in their marketing by saying Bubble is not for building prototypes. But Bubble is losing in my opinion large enterprise new builds, because they can pay the $50K a year to one guy with AI to do what Bubble does for them.
Overall demand for Bubble is down. Doesn’t matter Junior or Senior level developers, there is just less demand. I sometimes look at overall Bubble Agency count and last I saw it seemed to have dropped from 300 down to around 260. It seems as if less agencies are operating now, which would be a reflection of demand overall, and not demand for junior or senior roles.
However, on Upwork nearly all posts state looking for Senior Bubble Developer but the price point is for Rookie Bubble developer.
I can’t comment on Upwork because I don’t use it, but I used to generate a lot of leads through Reddit and LinkedIn organically. But those have dried up significantly.
What I have also noticed is that where enquiries do come in presenting as Bubble enquiries they actually end up being something else - projects coded using AI rather than No Code builds. Given I can provide full stack and DevOps this ends up being more attractive to founders because ultimately they want ownership.
As good as Bubble is (and I like it a lot and there are still a lot of benefits from choosing it) they are decent alternatives for rapid development now. And if your background is as an old fashioned developer from back when you literally wrote the code, then you are in a powerful position to choose the best thing for the client. And in many cases it makes sense to use Codex to build and then engage an ongoing maintenance and support agreement rather than paying the Bubble infrastructure fees/subscriptions. There is great flexiblity and initially the cost can stay lower for longer and you can optimise as required. And the client has independence.
For mature projects I have long worked on Bubble remains the first choice. There is no point migrating.
But for green field sites, starting from scratch feels like a slog, when you can get the first 80% done in 20% of the time very quickly with Codex/Claude. (I rate Codex much more than Claude).
I am not so much as fan of Lovable, Bolt, Cursor - they seem once removed from the metal for me and I’d rather just use the agent direct, feel more independent and not end up paying margin to the use the same thing through a 3rd party. But I would say V0 is worth it, once you have it hooked up to GitHub and you have Codex working to the same. It all just hangs together.
You can even get Codex to set up the infrastructure and then manage your deployments and monitor your infra for you.
So is there a need for Bubble? Hell yes. Not everyone has full stack skills.
Will Bubble stay relevant - that remains to be seen. It could end up just being another platform like Lovable/Cursor/Bolt once you hook up an agent to it to drive the development - the No Code USP hook suddenly loses it’s appeal…
I feel like this is pretty key.
Once Bubble lets you kickstart projects quickly and efficiently (a better in-editor agent combined with full MCP support so you can bring Codex into it), how does that change the equation for you?
Yeah, I think it does change the equation. But then I am comparing it to all the other tools (Cursor/Bolt/Lovable) and Bubble just becomes a proxy for something I can do directly in code and host wherever I want at a price that suits my clients, even with a maintenance agreement.
That is Bubble’s challenge.
The USP was No Code when Coding was hard. But that is not the case anymore. I can now code in any language I want - even Shopify templates - without any experience of using those languages before - because I understand how it works and how to build code.
An agent being a proxy for a NoCode development tool introduces an abstraction I don’t need. So why would I choose it?
I’m not a freelancer or getting work from sites like this…
but I do know this isn’t JUST a Bubble thing. It’s across the board in freelance work in this field.
Sometimes I’ll check out Bubble agency sites to see what they’re advertising.
I always wonder why they say things like ‘We build Bubble sites’ as their hero headline.
The average customer doesn’t care where you’re building an app. They just want an outcome.
It’s like if another agency advertised ‘We build React sites’. Yawn.
Do your marketing around the outcome.
Yes, it’s easier than ever to throw together an MVP, but that didn’t make getting users any easier.
Also, what is with some agencies having 1 landing page? That is like trying to throw the whole kitchen sink at someone.
Do a landing page for each thing you specialize in and run your traffic to that landing page depending on your advertising.
Bubble can be a plus for a lot of your customers after you’ve hooked them on what you can do for their business outcomes. I mean, they can change the color of a button without going through 13 messages with a developer or trying to prompt AI to make the change without it changing 9 other things.
Bubble has a lot of pluses. Use the pluses after you’ve hooked a customer.
An app is not automatically a business. You can show them how to make it one
I think you are both touching on something that’s actually a pretty uncomfortable conversation in this new age of “DIY it with AI”: customers that have absolutely no idea of how to build a webapp are giving it a go before connecting with professionals. That’s hurting lead gen for them. It also means there’s an immense amount of noise. Anyone can now put a variation of “AI-assisted developer” in their profile and try to get freelancing work.
The key here is that AI doesn’t make beginners less beginner. They don’t know what they don’t know. You say you can build anything with Codex so why Bubble. But if you saw the kinds of places where beginners get stuck, you’ll pull your hairs out. It looks basic to us, but not to them.
Yeah. But that argument is as old as a the hills. Everyone has encountered a shoddy builder (I mean houses, construction etc) .
It’s no different. Good developers will market themselves well and win the work. As @senecadatabase says, if you market and sell yourself based on outcomes then your experience will show through.
And it means networking, going to events, meeting people, building up referrals. It takes time. Just as anyone can build an app they still need to take it to market. Selling skills and services is no different. It’s just a different part of the same argument.
And for proper business with real needs and a good budget, the business is going to realise they can’t do everything themselves if they also want to dedicate time to business development, marketing, people management and whatever. So they either delegate, employ or contract out the building of the solution - if they come up with sketches, some sort of prototype or the presentation of what they are trying to achieve then all the better for the conversation.
The rules are still the same as they were. It’s just the competition includes more of those who create noise.
Bubble needs to compete just as much as anyone else and has to stay relevant as a solution of choice for their target market.
