I recently launched a social web app built on Bubble (it allows users to sign up, enter some details about themselves, message each other and view a filterable list of all users on the network).
We gained around 150 users in a 1 hour period after a facebook post. This prompted us to upgrade plans to the professional plan (we exceeded maximum capacity on personal plan). After upgrading, with ~40 users online we were at 60% of the professional plan capacity (without adding reserved units).
From my observations, this means that Bubble is incredibly costly to scale especially for a semi-complex social network of this nature (all users performing searches of other users and frequent reads + writes to the database).
If we took off and had say, 30,000 users, 3k active at any one point --> we would need literally 30-40 reserved servers constantly - a bill of >1000 AUD/ month. The equivalent AWS/ firebase bill if we decided to use traditional coding methods would result in a way lower cost…
Is my app just incredibly inefficient (we have followed best practice guides throughout our development) or do other Bubble users feel the same way about Bubble pricing? I understand that the idea behind the pricing is that Bubble shares in the success of its users. However, this leads me to question if Bubble is really only viable for high revenue per user SAAS applications, internal tools or simple applications with very little complexity (in this third case you could argue wix, squarespace or wordpress may be better alternatives…)
Genuinely curious to hear your stories regarding pricing per active user as you have scaled / at what point is it better to just use traditional coding methods?
There’s a lot od thing consider from how you design workflow to the fact that you are on shared capacity on personnal plan and a dedicated capacity on pro plan. The need for capacity is not linear. Also, if you have 30000 users at the same time… dont expect to pay 100$!
Yep with ~40 online users we were at 60% of max capacity on professional plan. No workflows running in the background, but a pretty involved search on the index page (i.e. search for all users according to 10 filters - no advanced filters though).
Sorry Taiheta I can’t link the app easily (you need a specific email to login with verification). In regards to the searches, there is a repeating group on the index page that displays users by searching for all users where the users meet 10 filters (no advanced filters) with “ignore empty constraints” selected. This seems to be the bulk of the log.
At this rate, I think the only logical way forward would be to rebuild the app using traditional coding (react + firebase). Bubble is amazing in that it really does seem like the future of coding (10x cheaper, 10x easier), but until someone develops a visual editor that is open source (where you can also see/ at least export the underlying code) it simply won’t take off.
Like I say, this is not exponential. There’s also a lot of way to reduce Server request (and sometimes, search can also be reducing. You can find some tip on the forum).
I also hope that some people that have built app with a lot of users (because I know they exist), jump in this topic to give you a better idea. (And sorry I miss that you said 3000 active user, not 30000 )
What is the basic functionality or usage? Is it mere commenting or is it heavy on media files sharing?
Am doing some cost analysis for my app. I am also expecting around 3000 users… with 300 peak login every 5 mins , 100 every 1 mins… mine will be just text and number submissions… or call it chat comments
)