Why is run.js so huge ? Everything else is queued and this is breaking my Google SEO scores

emir.ozgun · 2025-04-09T11:48:48.392Z

Yes, so this costs them a lot. Not good for very simple pages, not good for SEO, not good for very complex apps. I couldn’t recommend this product to my friends when they ask me, just because they would get slow page load, 60 score, rather they can get almost in an instant page load, optimal SEO from shopify and of course built-in payment system

rafiqhasan · 2025-04-09T11:52:04.389Z

I don’t think the issue came in today, some of the performance related threads are going back to 2022 ( ex: Bubble.io SEO ... Problems and Solutions - #23 by stuart4 ). Hence the solution should have been there till date, at least some guidance or some light from Bubble’s team

For a platform that markets itself as a No Code, I think Performance / Architectural optimizations should be in-built.

As @ademiguel mentioned above, a simple starter app with nothing brings in 823 KB of unknown / un-necessary code in run.js on Day 0. If nothing, then there should be at least some level of control available

emir.ozgun · 2025-04-09T11:58:17.334Z

BTW, I want to point out that all the ship your MVP fast is not the Bubble’s vision, that is something that community has given to Bubble , not even close. They have literally written a manifestation on how is not a mvp builder any more but something that replaces code. Thats why they are getting so much backlash currently, not delivering on promises but always chasing new ones.

rafiqhasan · 2025-04-10T08:41:28.562Z

Tbh, I am clueless on what can be done now ! I wish I knew all of this before I started investing my time and money

I am thinking of giving the whole website download to Claude and just help me build put this up with React

ihsanzainal84 · 2025-04-10T09:03:47.322Z

This is the simplest option.

manipaje · 2025-04-10T10:11:48.748Z

I shifted to a webflow SEO facing site and a bubble app on the subdomain. SEO has shot up on speed alone. It was worth doing early due to user login session cookies being a problem later down the line (moving to app.mysite.com make the existing bubble login cookies fail, users had to manually clear cookies). If you want a super fast front, I’d layer your tech now.

your experience may vary and this is just mine.

randomanon · 2025-04-10T11:04:37.133Z

How complicated/easy is this from a domain registrar perspective? Is it the same exact thing? Just copying a bunch of records?

On the Bubble side, for the domain name, do you literally enter app.domain.com?

I’m only asking because I’ve heard that domain.com/blog (on a separate site) and domain.com (on Bubble) is complex, but the subdomain seems easier.

manipaje:

SEO has shot up on speed alone.

Also, did your SEO actually meaningfully improve immediately after? Or just the core vitals/performance stuff?

rafiqhasan · 2025-04-10T14:09:54.770Z

I unfortunately can’t do this completely because I would like to SEO also on my product listing dynamic pages

I run a commodity search and listing site which needs a public facing dynamic content as well. Let me also try your idea if it helps me out

emir.ozgun · 2025-04-10T14:43:41.622Z

What is your performance score on light house ?

rafiqhasan · 2025-04-10T15:53:57.143Z

image1510×1154 59.8 KB

these are the scores and LCP Is 18s with 14s just as delay

emir.ozgun · 2025-04-10T15:56:56.066Z

With a huge SPA you should be gettin 60-70, in bubble you can get 70 at most on blank page, this is something else. It is about 2.8 secs to 3secs, do you use any bubble built plugins ?

rafiqhasan · 2025-04-10T17:45:47.713Z

Like I said in the first screenshot message, run.js takes almost 9s just to load and that stalls loading everything else ex: large banner on my website

I do have plugin but check the response from @ademiguel where he posted the same behaviour on a blank page application

manipaje · 2025-04-13T18:08:43.340Z

I use namecheap, and it was as easy as changing the ‘www’ DNS records from ‘www’ to ‘app’. Then I switched the bubble domain to app.myapp.com and waited a few mins, for DNS records to update. I did it at a very quiet time, so to minimise the offline.

A few links needed to change, while I put the new website live and ensured all the links to the login were to app.myapp.com/login

I also killed the index for app, as I didn’t want that to be the www index any more, so I ensured it did the most logical redirect for either logged in users, or not.

As for the SEO improvement, my LCP went from 16 seconds down to 2. That alone was worth it as I had an ad campaign running where nearly 0 users would wait 16 seconds, so I was paying for 0-second exits, not a good investment.

Could I have sped up bubble? Maybe, but do I love my more performant site - yes. All in, it was easier (and the better decision for me) to break them apart.

boston85719 · 2025-04-14T03:20:24.633Z

Don’t worry so much about performance because once you do what is in your control it wont go above 70 on bubble, which is not a make or break type of score. I highly recommend you get the other 3 scores to 100% because on bubble you can and they are in your control.

randomanon · 2025-04-14T07:55:30.734Z

manipaje:

I use namecheap, and it was as easy as changing the ‘www’ DNS records from ‘www’ to ‘app’. Then I switched the bubble domain to app.myapp.com and waited a few mins, for DNS records to update. I did it at a very quiet time, so to minimise the offline.

A few links needed to change, while I put the new website live and ensured all the links to the login were to app.myapp.com/login

Sounds pretty painless, thanks!

manipaje:

As for the SEO improvement, my LCP went from 16 seconds down to 2. That alone was worth it as I had an ad campaign running where nearly 0 users would wait 16 seconds, so I was paying for 0-second exits, not a good investment.

Oof wtf, mine is around 3. Not sure it would be worth going from 3 → 2. But I’m also not sure how much the other stuff affects SEO.

rafiqhasan · 2025-04-14T10:34:02.193Z

I don’t understand that why does Bubble.io not offer simple static pages or maybe something like SEO focused fast loading static pages ( with no WorkFlow option ) ? What is stopping them from having this simple yet powerful feature ?

There should ideally be no need to turn from one website solution to another, not sure what is the Product roadmap of Bubble

randomanon · 2025-04-14T10:36:27.600Z

Their roadmap is AI Builder/Mobile Builder then Loops/New Workflow Editor. But yeah, a static prerendered page setting would be amazing.

Bluebird · 2025-04-14T11:25:33.556Z

kind of funny, but i figured out you could put code in the page html header to crash run.js and inject your own static page content

rafiqhasan · 2025-04-14T15:03:08.367Z

yes I think too !

Do we have any PMs / Dev Leads who would care to have a concerning look at this issue and perhaps think of putting something like this on the roadmap ?

manipaje · 2025-04-14T15:48:00.334Z

Mine started fast but I needed quite a few plugins to get it built - this added the slow down for my app. I have another bubble app which loads more performantly (1.8s for LCP) but it’s super lightweight, nearly zero plugins.

I didn’t see another way forward to speed up, so I split the app. Maybe one day I’ll put it back together again but I’m happy with this choice for now.