| 1:49 pm on Oct 15, 2004 (gmt 0)|
Hanumanus, WebTrends' myriad line of software is getting to be so confusing (and expensive) even their own reps find it difficult to explain what all they are offering these days.
Going with the basic WebTrends 7 would require us to go with an Enterprise edition - which at $10 grand - I'll keep using the old Webtrends 5.5.
| 2:11 pm on Oct 15, 2004 (gmt 0)|
i recently decided to use webtrends on a couple of my sites, and i have to say, i wish i didn't. Most of my people who need to access it can't, and those who can have ridiculously long load times. Now i know this prob has to do with teh quality of computers i have people on , but please, they aren't THAT bad.
| 2:21 pm on Oct 15, 2004 (gmt 0)|
The pricing has to come down eventually. I mean com'on - processing log data into a report for 10 grand a year is outrageous.
| 5:54 pm on Oct 15, 2004 (gmt 0)|
Hanumanus - why would you want to upgrade anyway?
| 6:25 pm on Oct 15, 2004 (gmt 0)|
I recently did some research on appropriate log analyzers, webtrends being one of the candidates. We ended up going for an in-house solution, but the best candidate was Urchin. It's price is much more reasonable, it's price structure simple, it loads log files faster than webtrends and, best of all, you get a downloadable application that you can use for as long as you wish. Oh, and it's easy to use, unlike the webtrends test version I got my hands on.
NetTracker and ClickTracks are also interesting options. Give them a run for a couple of weeks and see how the stats they come up with compare.
The solution with best documentation was ClickTracks, but both Webtrends and Urchin have very good documentation as well. Nettracker had poor documentation, but I tested 7.0 and 7.5 came out a few days ago.
On a side note to that last comment: I ran the same log files in Urchin and NetTracker and got some VERY different results (particularly regarding how many sessions pr. day, there was a 60% discrepancy). It's a classic webmetrics problem, but might serve as a friendly reminder to newbies that the number, even from expensive suites, aren't exactly the most reliable figures in the world.
| 6:30 pm on Oct 15, 2004 (gmt 0)|
HOT TIP: all prices are negotiable.
I had Webtrends rep take their price from $15k down to $8k, so just remember that you have some 'wiggle' room when checking them out.
I agree that their old software versions are slow, but until someone comes up with Urchin's speed with Webtrends power, you are just going to have to use both or neither for log file analysis.
| 1:00 am on Oct 19, 2004 (gmt 0)|
Write your own tracking system. Server logs were never really intended to do what we want them to do. By writing your own you can capture pay-per-click details like keywords, engine, referring url, number of clicks per user. I write ever page hit to my databases so that I can analyze paths -- ie: I know the path of every user who purchases a product. I also do split testing which drives design changes. I have some code I can send (ASP). There are plenty of PHP examples on the web. ClickTracks and Urchin are okay for the money. Implment your own system - you'll be far happier.
| 1:29 am on Oct 19, 2004 (gmt 0)|
| 9:45 pm on Oct 28, 2004 (gmt 0)|
Yeah--webtrends is going down the toilet since M$ got involved. Lack of support for anything but Microsoft OSes AND the annual pageview limits (with fees for overage) is an awful decision.
new_shoes is right -- Urchin is a great way to go. Easy installation, great interface, good price, unlimited pageviews, and you can get reports on 100 sites with the base license. I use it through my host. Highly recommended.
| 3:14 am on Oct 29, 2004 (gmt 0)|
I setup Webtrends for the marketing people in my office, it was a nightmare to get setup and maintain. It broke when trying to do an update and their tech support was absolutely worthless.
We swiched to Urchin, the setup was very simple and I haven't had to deal with fixing it once.
Marketing seems to be happy with the numbers it produces as well.
| 4:59 am on Oct 29, 2004 (gmt 0)|
It looks like the fees for page views (by Webtrends) are accompanied by a price decrease, so maybe it works out. It's more or less the same pricing model that everything except the cheaper, simpler packages use. And I'll bet the decision to only support Windows was a way to keep their costs (and prices) under control. It's not so terrible to own just one Windows box, is it? If they need to do that to avoid having a second development and support team, then I'm happy to reap the benefit of lower prices.
Urchin is nice and I've used it off and on for a long time, but it's limited. It's good if all you want are basic and reliable report-card type statistics or you just don't see a payoff for site statistics. A lot of us are just trying to get visit numbers in order to be able to attrack advertisers. For real power at a middle range, I'd use WebTrends or the higher levels of HitBox or NetTracker. I don't think anybody should buy or complain on the basis of price without checking out what it can really do for you.
| 11:36 am on Nov 5, 2004 (gmt 0)|
Has anyone had much experience with IndexTools? How does this differ from Urchin/Webtrends? Is it just that they are log analysis and IT focuses more on data it collects itself? I would be interested to know what people think of it.
| 1:13 pm on Nov 5, 2004 (gmt 0)|
I don't have experience with it but it operates like HitBox, iWebTrack and other tools that use gif pipes (single pixel hits creating a second log). I wasn't too crazy about some of the scare statements on their web site, such as log files "often show 400% more page views than browser based tools such as Indextools"
[edited by: cgrantski at 2:05 pm (utc) on Nov. 5, 2004]
| 1:33 pm on Nov 5, 2004 (gmt 0)|
what features do you look for extra compared to cheaper products such as Web Log Explorer?
| 3:09 pm on Nov 5, 2004 (gmt 0)|
I dont know why people charge so much for these things, it doesn't look that hard to make.
I might make one in my spare time, i'll let you know how i get on :S
| 4:44 pm on Nov 5, 2004 (gmt 0)|
On a related note, which is the most reliable way of analyzing traffic (ie., calculating accurate numbers of page views and unique visits)?
1) A log analysis program which you run on raw log files on your desktop computer
| 10:28 pm on Nov 7, 2004 (gmt 0)|
The embedded tags - though it has its own drawbacks, fer sure.
| 7:00 pm on Nov 8, 2004 (gmt 0)|
It may not be WebTrends, but you can get a lot out of AWStats. It's open source, has lots of reports and you get get it right from SourceForge.