4
11 Comments

Website loading speed - How slow is too slow?

I run a free email notification service for slow-loading websites ( https://cwviq.com/). At the moment, I am sending out alerts anytime the loading speed is slower than 5 seconds but would like to check with the community here what they think is a slow-loading website - is it 5 seconds? 10 seconds? Or more than that?

What loading speed do you consider as slow loading?
  1. 3 seconds
  2. 5 seconds
  3. 10 seconds
  4. 11+ seconds
Vote
posted to Icon for group Developers
Developers
on December 26, 2022
  1. 3

    10+ seconds. However, this can change if it's just other elements on the website that take more time to load. For example, if a section in the footer takes 30 seconds, that might not matter because it'll take close to a minute before I even get there.

    Also, does your app also alert users if it's slower on phone vs desktop, or depending on the browser? Would be interesting to get more segmented data.

    1. 1

      Thank you. 10 Seconds is where I started, but then I started wondering if someone would actually be happy with, say, a 9 or 8 second loading time. So, have it set tentatively at 5. Maybe would leave it for the user to decide in future.

      Right now it's rather basic - crawling the sites once every few hours. I plan to make more customized crawls based on device, multiple geographies, etc. once I gain some traction.

      For example, if a section in the footer takes 30 seconds, that might not matter because it'll take close to a minute before I even get there.

      Valid point. So, right now, I am capturing data related to page loading (in a general sense) as well as the Core Web Vitals (CWV) - so the user can dive in deeper and fix things, if required. I should probably add more intelligence into the system in future so it can give deeper insights instead of just the data.

  2. 2

    For me, I'm skeptical of a website if it doesn't load within 2 seconds, or so.

    1. 1

      This is my feeling as well. In most cases I will wait for 10s, but everything above 2s increases my skepticism. So I'd say that near instant makes me trust a site, waiting makes me trust it less. And when it's at 10+ seconds I'm either on a government page or spam, I would never input into into it unless forced.

  3. 2

    Overall page can take more time, that's not a problem.
    But initial part of page should be very quick, around 1sec..otherwise it is considered as slow.

  4. 1

    Firstly, the average attention span of internet users is short, and they expect websites to load quickly. Therefore, if a website takes too long to load, visitors are likely to abandon it and look for alternatives. Based on this, a website that takes more than 5 seconds to load can be considered slow.

  5. 1

    I'm working on something similar, a product to monitor web performance.
    According to google the response time should be around 200ms.

    I think more then 1s is already too much like it had been said here...

  6. 1

    When you are coming from outside and it is an unfamiliar website… 1 second is already too long. Especially for the younger generation. There are studies that show a majority of (young) people hit the back button when the page hasn‘t loaded after 1 second.

  7. 1

    Aiming for under 2 seconds is best but the tolerance changes depending on the amount of content and functionality the site contains. As the complexity goes up, I think people are more understanding of longer load times and it becomes about how gracefully you load the content. YouTube doesn't do great on performance tests, but it has to load a lot of content and does so with good loading states and without having much layout shift, so it still feels good.

  8. 1

    Anything over two seconds is a problem (time to Interactive).

Trending on Indie Hackers
From building client websites to launching my own SaaS — and why I stopped trusting GA4! User Avatar 76 comments I built a tool that turns CSV exports into shareable dashboards User Avatar 70 comments $0 to $10K MRR in 12 Months: 3 Things That Actually Moved the Needle for My Design Agency User Avatar 67 comments The “Open → Do → Close” rule changed how I build tools User Avatar 49 comments I lost €50K to non-paying clients... so I built an AI contract tool. Now at 300 users, 0 MRR. User Avatar 44 comments A tweet about my AI dev tool hit 250K views. I didn't even have a product yet. User Avatar 40 comments