> Unfortunately, it's still hard to tell how much of that is substantial engagement. A pinging stats provider, like Clicky, gives you great insight into single

> Unfortunately, it's still hard to tell how much of that is substantial engagement. A pinging stats provider, like Clicky, gives you great insight into single-pageview engagement. It's true that there's room for inaccuracy in the pinging measure. Some viewers with third-party (or all) JavaScript disabled won't be counted, viewers that leave the page open in a tab before they look at it may be over-counted, and readers who legitimately spend more than ten minutes reading a longer page are capped at the ten minute mark. However, averaging that across thousands of views is a very worthwhile measure of engagement. I had a relevant post about that here recently (which was castigated for semantics more than it was actually discussed, unfortunately): http://news.ycombinator.com/item?id=1294969 http://news.ycombinator.com/item?id=1294969 The Clicky pinging data was invaluable then too. HN visitors to that post show an abysmal 95% bounce rate and an average 14s time-on-site in Google Analytics. Meanwhile, Clicky shows those same HN visitors spending an average of 3m29s time-on-site and truly bouncing from the page only 12.9% of the time. Even if you assume Clicky's numbers are wildly optimistic, it's nice to have that alternate insight.