Here’s an interesting thing to consider next time you’re looking through your web stats: the popularity of Safari 4 and later (the desktop version, not the iOS one) probably looks way higher than it actually is.
Safari’s “Top Sites” feature checks sites to see if any content has changed since you were last there (and also to draw the preview images), and it does this reasonably often. The way it checks is to load your website behind the scenes, which means your server gets hit, and your stats show Safari’s more popular with each refresh.
Perhaps Apple should make the background thread use a different user agent string so that we can differentiate real requests from the Top Sites updates?
1 thought on “Safari skews web stats”