Life in the (not-so) Fast Lane
Photo by bluedeviation

Life in the (not-so) Fast Lane

I recently read in the stats section of the http archive the current average page sizes for various web sites across the world. (Being a performance tester I'm a sucker for stats.) As of July 2015, the average page size for the top 100 websites in the world was 1,440Kb, top 1,000 sites' page sizes averaged 1,999Kb and across all measured sites the average was 2,162Kb.

I've seen lots of stats over the years concerning performance optimisation, web site load times and various top-100 lists, most of which refer to US-based response times, but seeing these page sizes got me wondering about end-user response times for Australian consumers, out here in the backwoods of the world wide web.

A couple of quick Google searches later, I see that average web download speeds in Australia currently look something like:

  • Average 4G download speeds are 24.5 Megabits per second, (3 Megabytes/sec)
  • 3G speeds are generally about 3x slower, you could safely reckon on somewhere around 8 Megabits per sec, (1 Megabyte/sec)
  • And the average home broadband speed is somewhere around 7.5 Megabits per sec (0.94 Megabytes/sec)

It's not a huge leap to link the two sets of stats to come up with a quick overview of what this means for Aussie Joe Average surfing the web or shopping online.

  • He'd be able to load an average page from the top 100 sites in just under half a second on his 4G handset, a fraction over 1.4 seconds on a 3G handset, and 1.5 secs on an average broadband connection
  • Going slightly less popular to a top 1,000 site would see a slight increase in response to 0.64 sec on 4G, almost two seconds (1.95) on 3G and a whisker over 2 seconds (2.08) on broadband
  • Treading the path less trodden, he'd see his response times increase again, to average 0.69 secs on 4G, 2.11 secs on 3G and 2.25 secs on broadband.

At first sight these don't seem too bad, especially when you factor in the tyranny of distance and realise that most of the sites he'd be surfing are half a planet away, but when you consider these response times in the context of how people perceive wait times, things start to look a little less rosy.

Various studies on people's perception of time consider "instantaneous" response to somewhere in the region between 1 and 100 milliseconds -- so the magic number and holy grail of response times is less than 0.1 of a second.

The magic number and holy grail of response times is less than 0.1 of a second.

The next level up from this is between 100 and 300 milliseconds, where people detect a small delay, but things are still "fast." -- between 0.1 and 0.3 of a second. Try to beat a 100ms reaction time on your stopwatch and you'll get the idea.

When things start to take 300-1,000 milliseconds to respond, between 0.3 sec and 1.0 sec, people generally consider this as "the machine is working." It's not instantaneous, but with a noticeable lag between user input, the machine doing something and the result being presented, people notice it.

For our web-surfing Aussie Joe Average, this is the fastest response he's getting on any of the sites he visits using his state-of-the-art 4G device.

In human perception terms, anything between 1,000 and 3,000 milliseconds (1-3 secs) allows the viewer time to mentally context-switch. Unfortunately, this is where the majority of the web responses are in the above examples. Any illusion of interacting directly with his handheld or computer locally is lost when Joe waits for a response.

Obviously anything above 3,000 milliseconds (3 secs) allows him more daydream time, and when things start to get to 10 seconds, his abandonments rates would skyrocket.

Obviously these figures all relate to the average page sizes and average download speeds here in Australia, so if you're getting response times better than this, pity the poor outliers that are getting worse than you.

Google often states the magic 100 millisecond response time as their goal, but given our remote locale and our average download speed, the sub-300 millisecond "fast" category could be some way off for Aussie Joe Average.

Until then he's going to see a lot of "machine is working" responses and plenty of time to wonder when the NBN will arrive, during his mental context-switch daydream time.

Chris, Are you taking into account the extraneous advertising etc that gets pushed in when you view a web page? And do ad blockers improve performance?

赞
回复

要查看或添加评论,请登录

Chris Jones的更多文章

  • Performance is not horsepower alone

    Performance is not horsepower alone

    I’m always banging on about performance being fit-for-purpose rather than just outright speed. At AccessHQ, we’re…

    3 条评论
  • Why do we never learn in IT?

    Why do we never learn in IT?

    I honestly can’t think of any other major industry which consistently over-spends, under-delivers and repeats the same…

  • Think you'd make a good Performance Tester?

    Think you'd make a good Performance Tester?

    “Some men aren't looking for anything logical, like money. They can't be bought, bullied, reasoned, or negotiated with.

  • Performance Is Not Just Speed

    Performance Is Not Just Speed

    A lot of people think performance in IT systems is all about speed. Questions like; “How fast does it go?” and “What’s…

    2 条评论
  • Performance Testing - are you peering into darkness?

    Performance Testing - are you peering into darkness?

    Within big IT projects, often performance tests run with few infrastructure monitors in place, if at all. It's not…

    2 条评论
  • Performance Testing Averages, 90th percentiles or Avg-90%?

    Performance Testing Averages, 90th percentiles or Avg-90%?

    As a performance tester, my role is to ensure my clients clearly understand how their systems perform under load. To…

  • Performance is not just reliability and availability

    Performance is not just reliability and availability

    There's no truer maxims in the world of complex IT systems than Murphy's Law, and its corollary Finagle's Law. These…

  • Why Testing is like Book Publishing

    Why Testing is like Book Publishing

    Testing software is a complex and difficult thing. There are so many opportunities for issues to arise - from major…

  • Performance Testing Cheatsheet - Diagnosing Server Congestion

    Performance Testing Cheatsheet - Diagnosing Server Congestion

    When faced with diagnosing performance issues with windows-based servers (especially when perfmon stats are easily…

    1 条评论
  • Open All Hours. 23? x 7; 357 Days a Year

    Open All Hours. 23? x 7; 357 Days a Year

    In today's online market, to coin a phrase, time is money. So, availability and performance are king, right? Well…

社区洞察

其他会员也浏览了