The Truth About Download Time
We hear all the time from web designers that they spend countless hours and resources trying to speed up their web pages’ download time because they believe that people are turned off by slow-loading pages. Their concerns have been amplified by experts like Jakob Nielsen who asserts that users become frustrated after waiting too long for pages to load. It makes sense that a slow loading page is unusable. We know that if a page takes 2 hours to load, chances are people will abandon their tasks. But when does download time go from too slow to fast enough?
Nielsen reports that the home pages of the most popular sites he studied took an average of 8 seconds to download, whereas the pages of the less popular sites took an average of 19 seconds to download. He therefore concludes that users will be annoyed or frustrated by pages that take any longer than about 10 seconds to load.
When we began our research, we thought we would find a strong relationship between page download time and usability: sites with faster download times would be more usable than slower sites. We also expected that users would be consistent in their ratings of site speed, and that these ratings would correlate strongly with the actual speed of the sites.
To test these predictions, we studied 10 different web sites over a 56 kbps modem. On these sites, we had users perform their own personal tasks; each user did something that was interesting and meaningful to her. No two users performed the same tasks on any site. For each of the sites, we had users rate how fast they felt the site was. We called the users’ measures their “perceived speed” of the site. Later, we watched videotapes of the studies and measured the actual download times of the pages.
We started by confirming one of our hypotheses: all users rated the speed of the 10 web sites consistently; they thought Amazon.com, REI.com, and L.L. Bean.com were the fastest and About.com was the slowest. Despite having performed different tasks on these sites, users were consistent in their reports of perceived speed.
Our other finding, though, took us entirely by surprise. 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. About.com, rated slowest by our users, was actually the fastest site (average: 8 seconds). Amazon.com, rated as one of the fastest sites by users, was really the slowest (average: 36 seconds).
There was still another surprising finding from our study: a strong correlation between perceived download time and whether users successfully completed their tasks on a site. There was, however, no correlation between actual download time and task success, causing us to discard our original hypothesis. It seems that, when people accomplish what they set out to do on a site, they perceive that site to be fast.
When we thought about these findings, they made a lot of sense to us. If people can’t find what they want on a site, they will regard the site as a waste of time (and slow). But, when users successfully complete tasks on a site, they will perceive their time there as having been well spent.
Jakob Nielsen tells designers to focus efforts on improving actual page download times on their sites. But what we’re seeing leads us to wonder if it’s worth the resources to make web pages load like lightning. Instead, we’re wondering: When users are complaining about the download speed of your site, what are they actually complaining about? Are you better off making the site load faster or ensuring that users complete their tasks?