Forum Moderators: DixonJones
we have a meeting with them on Thursday, and i just wanted to get a feedback from you guys so that i'll know better what questions to ask etc.
what are your opinions on the service they offer. is it really that good? is it really that much better then WebTrends? what are the advantages and disadvantages of their technology?
thanks a lot!
If you have access to the server logs you'll probably get better data than that supplied by a service like HitBox. For instance, you can't track search engine spider activity with their service. They will give you really pretty online reports to access, but there's only so much value you can get from those reports if you're looking to get information about all your site activity, not just those that trigger their proprietary code (usually JavaScript calling a 1px transparent .gif)
Websidestory, WebTrendsLive and a few others provide a real-time statistics service which requires that you paste some javascript into every page that you want to track. The javascript is basically a redirect which creates a hit in their server's log file. If you have a very busy site, this is sometimes the only way to go. A few disadvantages of this type of tracking:
Errors (like 404 - File not found) cannot be tracked;
Only html files can be tracked;
Bandwidth reporting cannot be tracked;
If there is an error in the javascript, you do not get a hit on their server, thus no stats.
Livestats and SurfStatsLive have real-time reporting programs but do not provide a statistics service. You would have to ask your host to install the program to monitor your stats for you. The server log files are directly analyzed in real-time which have the following advantages over javascript generated stats:
You can track bandwidth, errors, different file types, do not have to insert javascript in pages, have historical data (if your log files are on the server)and is probably also cheaper in the long run.
Others like Urchin, Analog, SurfStats and Webtrends have programs that analyzes log files on scheduled intervals but this is not a real-time solution.
Hope this helps...
I dumped their HitBox for SuperStats after I found that they only tracked search engine referral data to the home page of the site being tracked...
I suspect you were using their free tracker - the paid service - which I use for a client - has always tracked multiple pages.
Errors (like 404 - File not found) cannot be tracked
This may have been the case at one time; it is not the case today. Errors can be tracked. (after I paste the new code in today I'll be able to say that more authoritatively.;))
Only html files can be tracked
Overall, we are satisfied with the service, but like ANY tracking system I have seen, it has it's shortcomings. Unlike log-based systems, it is real-time. You won't have to wait until tomorrow morning to see what happened on your site today.
If you want me to address anythig particular (look to see if a specific feature is available, etc.) sticky me.
You usually need to set two cookies: one with a large max-age and another one which gets erased after the session is over.
That way you can track both visitors and their sessions.
The tricky part is restoring visitor cookie from a session cookie because many (especially new) browsers deny cookies with long expiration dates.
So if you track a hit and see that the session cookie is present but the visitor cookie is not - you got to match the visitor to that session based on the previous hits.
Otherwise, you'll eng up having multiple visitors for one session, which is impossible :)
Anyway, I guess I'm a bit late with this info for your meeting, but you might still want to ask them how they handle all that stuff.
Also, find out how they handle tracking of pages served over ssl, since the domain of the pixel would be different and that gets really tricky :)
Actually, you need both raw logs and pixel trackers to be most effective.
I would completely agree - just wish I could find something that meets my needs. It sounds like you are doing exaclty what I wish I was doing. What solutions are you using?
I dumped their HitBox for SuperStats after I found that they only tracked search engine referral data to the home page of the site being tracked...I suspect you were using their free tracker - the paid service - which I use for a client - has always tracked multiple pages.
This did happen a couple of years ago and things probably have changed since then, but up until 2000 all of that WebSideStory data about search engines is accurate only for click throughs that land on a site's "home page". I always cringe when I hear people refer to WebSideStory data in reference to search engines or keywords since then.
The stats are very comprehensive. You can set up groups of pages that correspond to the different directories, and track paths, downloads, SE referrals, uniques, keywords, and referrer urls.
Instant satisfaction pleases the client. You just can't beat the SPEED of the results, which can be downloaded to an Excel spreadsheet. Clients love INSTANT gratification. (shrugs)
The service doesn't provide bot tracking, which to me is very important. I like to know who was where doing what and when. Of course, clients don't care because they don't need this info. But as an SEO, I need it.
Perfect world scenario? Get the hitbox pro for the client, and purchase faststats for yourself and crunch their data for the information that you, as an seo, need.
Also, I'd like to mention that to get the FULL picture logfiles and trackers are not enough. As pointed out there are some data not available by trackers, and some that are not avaialable by logfiles - they each have something unique and usefull. However, you can also collect data on a newtwork level using network packet sniffing technology. This will give you access to even more data than what you have in your logfiles or trackers like HTTP STOP REQUEST (tracks when someone hit the STOP buttom - a very valuable information), true measuring of users actual connection speed etc.
