Forum index layout

[Justinian]

I would be interested to see the Analytics for this month and how NTW changed the numbers.

[Van Zandt]

I would be interested to see the Analytics for this month and how NTW changed the numbers.

A months numbers is not enough to make a good decision. Lets looks at at least a year or two and some other stats as well, such as number of views, posts, time on page etc.

Looking at one of Sims original statements,

10% of new visitors who hit the forum index currently just leave, not going to any other page. They probably looked for a few seconds, didn't see what they wanted, and went back to try another search result.

....there is no way from this stat you could know why the viewers left the forum, this is just speculation on your part. Another explanation is that they goggled "Napoleon", and ran across a gaming site instead of a biography....

[Augustus Lucifer]

....there is no way from this stat you could know why the viewers left the forum, this is just speculation on your part. Another explanation is that they goggled "Napoleon", and ran across a gaming site instead of a biography....

The likelihood that we would be in the Top 500 results for a google of 'Napoleon' is infinitesimal, much less that someone would click it based on the blurb. 'Napoleon Total War' we're on Page 2, and ironically it's the mod "Napoleonic Total War 2" for Rome, not the new game.

[Van Zandt]

The likelihood that we would be in the Top 500 results for a google of 'Napoleon' is infinitesimal, much less that someone would click it based on the blurb. 'Napoleon Total War' we're on Page 2, and ironically it's the mod "Napoleonic Total War 2" for Rome, not the new game.

I did a quick check, variations of "Napoleon" and "war" and "battle" brought up Total war links on the first page.

The point being there is no way you can know the viewers intent simply by reading the numbers. Only Sociologists would do something like that.

[Søren]

I did a quick check, variations of "Napoleon" and "war" and "battle" brought up Total war links on the first page.

The point being there is no way you can know the viewers intent simply by reading the numbers. Only Sociologists would do something like that.

Some bring up totalwar.com and NTW related links, but how many bring up twcenter.net?

[Simetrical]

I would be interested to see the Analytics for this month and how NTW changed the numbers.

Feel free to ask for permission so you can analyze the stats yourself.

A months numbers is not enough to make a good decision. Lets looks at at least a year or two and some other stats as well, such as number of views, posts, time on page etc.

Let's just not make any decision for a year or two because maybe it won't be perfect?

....there is no way from this stat you could know why the viewers left the forum, this is just speculation on your part. Another explanation is that they goggled "Napoleon", and ran across a gaming site instead of a biography....

I speak from the experience of being a web developer. I have read more than a few descriptions of controlled studies that were conducted by a variety of parties (e.g., Google) on how new users interact with websites. For instance, High-Performance Websites by Steve Souders (formerly employed by Yahoo!, now by Google) mentions some statistics about user response to page load time. I've also seen summaries of various eye-tracking studies by Google, etc.

It is not speculation that new users who arrive from search engines will leave if they don't quickly find what they're looking for, although you're correct that this doesn't necessarily mean that accounts for most of the bounce rate. The bounce rate is actually much higher for search engine results for the site overall, more like 25%. It varies widely by search term, too. For the top ten terms,

twcenter 12.00%
total war center 12.31%
third age total war 19.44%
tw center 12.86%
empire total war mods 15.23%
rome total war mods 16.02%
total war forums 11.89%
roma surrectum 13.92%
total war 22.48%
empire total war forum 10.93%

Notice how the lowest bounce rates are when they search for us specifically, or search for a forum specifically. Searching for mods gives a higher bounce rate, although we're surely the best place to find them: it seems likely people aren't finding the modding part quickly enough, and we could improve that. The highest bounce rate of the top ten search queries is "total war" itself, which makes sense, because our generic Total War content is fairly poor. It's unlikely that anyone searching for non-game uses of "Napoleon" or "total war" would even click our link in the search result.

It's possible to set up controlled studies using Google Analytics, although the setups it supports are pretty tightly constrained. It would probably require hacking vBulletin to get such a study to work. Sadly, I don't have the time.