Cambodia Expats Online: Forum | News | Information | Blog

http://www.buzzfeed.com/mylestanzer/why ... ie-already

Really establishes how the # of views is fairly meaningless. Time spent on each view is much more crucial, as well as low bounce rates on a site like this one. Analytics has added some great new metrics over the years, including one called "engagement" where you can see if you're actually engaging users (do they click on resource links like e-commerce, do they comment/review/post on things, etc). Engagement, to me, is the most critical aspect of a site... getting 1,000,000 people a day to land on it for 6 seconds and go "what's this? shit, not what i wanted." and then leave (even if they click to another page to be sure, which would nix them from the bounce rate stats even though they are still really a bounce) means very little. Getting engaged users who frequently revisit the site and participate (not just for forums, but other site types as well.. browsing a catalog or doing a search and finding the results they were looking for are engagement as well) are where the real values are.

Anyway, enjoy the article if you're a geek on stuff like this as I am.

As an example, for today we've had 64% of our total user session be over 30 seconds.

60% were over 1 minute.
51% were over 3 minutes.

But even with 26% being under 30 seconds, our bounce rate is still 8.5%... which isn't really an accurate reflection of usage in my opinion. so they hit 2 pages... in 27 seconds... then left. That's still a shitty user (nothing personal, only from a statistical view) and shouldn't be counted in positive usage stats.

Unless you get a lot of traffic a handful of people leaving their browser windows open would skew the session length results, wouldn't it?

giblet wrote:Unless you get a lot of traffic a handful of people leaving their browser windows open would skew the session length results, wouldn't it?

Generally not... it doesn't just sit there and keep reporting back, it waits for actions. I guess one leaving it open then coming back an hour later and clicking onward might. But it would still be a minority percent easily removed by eliminating the ones that are in the "1801+ seconds" group (that's over 30 minutes, to save you the math)... which make up 14% of the total. That is hard to factor into bounce though and it's likely not 100% of that group if you count people frequently checking back but keeping it open.

But the likelihood of someone leaving it in their browser being a bounce is pretty low. It would likely be left that way after a few interactions.

It has a session timeout as well, if there's no activity, but I'm not sure how long it is.

You need to bribe vlad to divorce 440.

Oh hell no, he'll post over here twice as much, then!

StroppyChops wrote:Oh hell no, he'll post over here twice as much, then!

Maybe I am being pedantic but you offered your services to tutor the girlfriend of a poster here. So tell me professor, why do you have that second comma in the quoted text. Was it an autocorrect problem or are you only semi literate at best?

I tend not to care but when someone is trying to flog off their services as an English tutor shouldn't they also be trying to at least display a general understanding of the language?

Or is it unnecessary when all you are teaching is passages from the Bible.

This amount of passion for what you believe to be a misplaced comma. You really do have wood for me, as OD suggests. That's nice, but I'm not looking for any more bromance right now, my dance card is full.

,,,,,

OrangeDragon wrote:http://www.buzzfeed.com/mylestanzer/why ... ie-already

Really establishes how the # of views is fairly meaningless. Time spent on each view is much more crucial, as well as low bounce rates on a site like this one.

I'm not even reading the article, but I've been in the game long enough to say that it depends on the website.

More time spent on a page is not more crucial, it depends on what the viewer is searching for. Sometimes you need to look at a page for less time, because you need just a little information from it. Sometimes you need to look at it longer because the article is interesting.

(For forum websites I'd say "Time on Site Per Day" or something like that would be the best metric, but I can think of other cases where "page views" would be more important than that and can think of cases where "page views" is meaningless).

Bottom line, depends on the type of site.