Perceptions of page loading speed

It is fairly amusing when people (clients or otherwise) demand that a page load in under x seconds (where ‘x’ varies by which guru article they were reading).

Pages should load quickly, absolutely, no question. Sometimes, for various reasons under you control (large page assets, application load etc.), they won’t. Sometimes they won’t load quickly for reasons out of your control (the network traffic, or the user’s connection).

Applying a number to page loading speed is always suspect, and I’ve generally applied the metric of making sure that people can continue their task within 10 seconds on a reasonable connection (56k+). That doesn’t mean the page has loaded, but that enough has loaded for the person to start reading, navigating or interacting.

This hit home today when David Hyatt (lead developer of Safari) highlighted people’s mis-perceptions:

I just saw this article on Digg about how to “speed up Safari” by reducing the “page load delay” in preferences. In the comments are many testimonials like “Oh my gosh! Safari is so much faster now!” This just goes to prove how inaccurate people’s powers of perception are when it comes to measuring the performance of browsers. I say this because the preference in question is dead and does absolutely nothing in Safari 1.3 and Safari 2.0.

Although talking about browsers, the same (mis)perceptions has to apply to pages as well, being what browsers load.

If you still aren’t convinced, try Christine Perfetti’s research:

When we looked at the actual download speeds of the sites we tested, we found that there was no correlation between these and the perceived speeds reported by our users., rated slowest by our users, was actually the fastest site (average: 8 seconds)., rated as one of the fastest sites by users, was really the slowest (average: 36 seconds)…

…when people accomplish what they set out to do on a site, they perceive that site to be fast.

I’d much rather a site seem fast than be fast. Unfortunately, little truths like this don’t tend to stop the checkbox mentality.

5 contributions to “Perceptions of page loading speed

  1. well, as long as every page is only a maximum of 2 clicks away from your homepage (regardless of complexity of your site) and you don’t have 7 +/- 2 links in your navigation, you’re golden anyway. oh wait…let’s apply those golden rules to amazon for a minute…

  2. lol, I usually hear the 3 clicks and 7 +/- 2 navigation elements, which gives you about 27 pages?

    That’s not usually to hard to overcome in people’s minds any more, there being so many obvious examples where it can’t be the case (and research on it).

    Thanks for the link Andy, it looks like a good overview of quite a few aspects of people’s perceptions.

  3. 3 clicks by 7 +/- 2 works out to 7*7*7, or 343 pages…which is mostly irrelevant. : )

    As bandwidth to the end-user increases, I’ve seen the actual .htm file increase in size…I used to try to keep the HTML under 5kb, and now I run across non-multimedia sites whose front pages are around 100K, and the images and CSS are another 200K+ on top of that.

  4. Surely you get the factorial effect (where each page also links to previous pages) so it’s 7*6*5? (Aiming for the smart-arse award 😉

    You’re right though, I still have friends on dial-up, and for them the web is turning into rather a walled garden experience, there are many sites they won’t visit.

Comments are closed.