I have been using bubble for a year and a half. Speed was always an issue. I came back after a couple of months and the situation seems worst.
I tried an experiment to see if it was only an impression. I made a new that it is a blank page with a text. No plugins installed. Nothing but a text. These are the results from PageSpeed Insights:
@JohnMark post updated for more recent stats: Airbnb shows score of 35 and Uber shows 5. PageSpeed Insights / Lighthouse metrics are not a good reference point as you can see major tech companies scored low which is inaccurate for page load times experienced by users. That said, we are working on optimizing for this score.
@lucarenga1 We are making as much progress as large tech companies See post above. In our experience, this score is not strongly correlated to ranking as most apps are able to rank high with the built-in SEO settings. Email our team if you have further questions about optimizing for your app.
In my experience, speed is an important factor when it comes to seo, but I have no problem changing my mind. Can you share some of those bubble-app that rank high?
I think for whatâs offered, bubble isnât that badâŚ
name other sites that you can do just as much or more (while retaining no-code friendly UI/UX) that out perform bubble.
If you can name a few then maybe Iâm on the wrong platform.
Right now, I am considering performance (speed) as the only variable. And when it comes to speed, bubble is probably the slowest no-code builder I ever tried, by far. Like it or not, speed is something you canât ignore in 2020.
I have no doubt that it is one of the most powerful, but thatâs not the point here.
I agree with your argument regarding performance. However, like @neerja stated, metrics by PageSpeed Insights are usually misleading and inaccurate - so if youâre basing your appâs performance issues on that premise, you may want to reconsider. I have Webflow apps that score much lower than your score.
Below are PageSpeed Insightâs scores for Uber and Airbnb respectively.
I have built and sold two Saas on bubble. In addition, I built a marketplace that was pretty complex.
I said that this was a blank page, just to test if the performance issue where on my side. If a blank page performs so bad, that I can easily think that the problem is on bubbleâs side.
I am not evaluating the app performance based on PageSpeed Insights only. I take into consideration those stats for SEOâs purposes, because Google does the same.
I know that there are lots of tech companies whose website sucks, according to PageSpeed Insights metric, but they are so big that: 1. they donât rely anymore on SEO. 2. they have so many backlinks and so muc domain authority, that can easily compensate for the bad PageSpeed Insights metrics. The average bubble users canât say the same, unfortunately. The average bubble user needs to optimize everything, if he wants to rank in google for decent keywords.
I found Google PageSpeed Insights to very subjective at ranking your real website speed.
I usually use Pingdom to really figure out whether there is something we can\should improve.
Hereâre the example result of Airbnb on Pingdom by the way.
Not mind blowing, but not as awful as it looks like on PSI (1sec page load is not too bad).
I understand that it does not help the perception of Google of your website, which is what you are saying
IMO the real problem with Google is that it just does not like dynamically rendered websites.
A lot of Insights from PSI are about removing this script and that script, or changing something in the way they load\work.
Google favors good old static html websites that sit still while it crawls them.
All that being said - I fully agree that Bubble has a lot to improve in terms of speed, and I also agree that there is visible progress over the last couple of years.
I also ran a initial test on a Zeroqode template (Iâm not picking on Zeroqode here, I just wanted to use a template that has a pretty complex and element intense frontend.) Note: this is on the Hobby free plan, so possibly throttling page load:
With both Bubble sites within the waterfall of page requests I have noticed a few things that are adding relative delays to a full page load, not sure if these are normal:
Now Iâll be the first to admit Iâm not 100% sure what I am looking at and how these variables are calculated to process the end results, but thought Iâd share on the subject of testing tools and page performance under live sites.
Also for general context for averages for insight:
Iâll need to do some speed tests between using the integrated Google Fonts API and using custom fonts, loaded over Cloudfront via this technique - but it might be a quicker way round (the latter technique).
Iâll do a couple of tests over the weekend if I get time and report back with findings.