All of a sudden I see my app way faster with no latency even when doing complex searches with advanced filters. Did bubble implement something to fix latency or something?
Is anyone else experiencing their app be way faster? Or is it just me?
Its always hard to tell, but I don’t see lcp improvement rather, total blocking time reduces by 400ms on everyapp of mine, to like 10ms, causing performance to go up.
I agree its always a hard metric to test against and could be a number of things - with that said, dunno if it’s just me but I agree with this statement. In general actually things are feeling a bit snapper e.g. editor usage, page load, searches - generally that lag, especially after a solid hour of editor usage seems to be gone (for the time being). Happy so far.
I don’t think so. They certainly would have announced an improvement like that. At least for me, the behavior is still exactly the same. I haven’t noticed any improvement in performance.
It remains slow; they need to improve database read and write performance, which is very slow by current standards.
It’s pretty fast for a lot of things tbh, I guess when indexed well & when the datatypes are light.
But can be slow when doing more complex searches. Also I think it’s slower because it always pulls in “all” fields even ones that are not needed at that point.
But that’s exactly the point: if it’s only fast with lightweight data, good indexing, and simple queries, then it’s slow compared to modern database standards.
In many Bubble applications, we end up creating redundant fields, denormalizing data, or restructuring searches just to maintain acceptable performance. If workarounds are required to achieve speed, then there is clearly a limitation.
These optimizations also increase application complexity and ultimately raise costs, both in development time and WU consumption. I’m not saying the database is bad, but saying that it is generally fast doesn’t really reflect the reality of larger applications.
Thanks for the insights, guys!
Huh? If your users are satisfied and your software works as expected. What’s the problem?
What’s the benchmark for “modern standards” of database speed? 
Considering the apps built with Bubble come in all forms and sizes.
Have you tried using Azure dashboards or Meta’s Business Suite?
In my case, I’ve received quite a few complaints about occasional slowdowns. With smaller applications, this usually isn’t an issue, but things change significantly once you’re dealing with large tables and higher data volumes.
That’s when we often end up denormalizing data, creating redundant fields, duplicating information, or restructuring queries just to maintain acceptable performance. Those are valid optimization techniques, but they also highlight a limitation of the platform in these scenarios.
On top of that, these optimizations make the application more complex, harder to maintain, and in Bubble’s case, they often increase WU consumption as well.
As for “modern standards,” I’m not comparing Bubble to products like Azure or Meta Business Suite—they solve completely different problems. I’m comparing it to modern databases like PostgreSQL or backend platforms like Supabase, where equivalent queries generally scale better and require far fewer compromises in the data model to achieve good performance.
That’s really my point. Saying Bubble’s database is “generally fast” is true for smaller apps or simple queries. But as datasets grow, that stops being the experience of many developers—and it certainly hasn’t been mine.
I was not complimenting Azure and Meta Business Suite. They are horrid to use.
I do agree that Bubble’s abstraction does take away some benefits of PostGres, AFAIK, it is built on. That said any good database optimization requires data modelling that works best for that database.
Any abstractions will have come with their own quirks. For example, 6 years ago I came into Bubble from Airtable. Both were abstractions and came with their own best practices for data modelling.
If working with Bubble’s best practices (be it database speeds, rendering performance etc) still offer my users a good user experience comparable to my competitors, is Bubble still “not as good” as whatever else? IMO it is the fault of the builder for not wanting to work with Bubble’s strengths and best practices, and working around it’s weaknesses.
In my app I have two main types of products one is the main thing (full vocals) and the second is like a side thing (vocal samples) two market places in one I guess.
The main product is using bubble search and the second one I built using coudflare D1 database as a “fast layer” (but with bubbles data being the source of truth).
I did this partly as an experiment but also becuase of rivals & expectations. When producers search for samples after years of using tools like splice they simply EXPECT lightening fast results. I felt I had no choice to match that user experience. I believe I matched it and even improved on it… Searching samples is now pretty instant. Searching the full vocals (bubble dB) is a little slower.
But the point is… both work. No one is complaining. The heavier files (full vocals) feel a bit more heavy & premium because of the 1 second search , and the small lightweight samples feel “right” that they come down the pipe instantly. I was originally intending to rebuild the vocal side of things with the D1 database as well, but now I’m not actually sure it’s worthwhile because there are trade-offs in complexity.
Actually not really sure what the point was, but hope you get something from this! 
I read this earlier (diary of a CEO book) and I realised it explains what I was intuitively getting at in my previous message:
"Flight, hotel and insurance aggregator websites understand that friction can create value. They found that faster search times on their websites often resulted in fewer sales. They now artificially increase search times and show all the sites they’re searching in an attempt to convince you that they’ve done a thorough search, so you don’t have to look elsewhere. This tactic has resulted in more sales, better retention and higher customer-return rates."