Web statistics problems

[Martyn]

Adventure Trailers did a trade last year for some banner ads on a web site. The web site in question is very busy with 40,000 visitors a day.

We both became part of a new endeavor and the web site in question offered to keep running our banner ad. I wrote to the owner of the company and told him that we had only had a trickle of referrals from his site this year and could we get better prominence to increase it.

From our web site statistic we have had 270 referrals to our site from their site since January 1st this year.

He e mailed me back, after running their Banner Ad tracking program, and told me the AT banner ad had generated over 90,000 views and 1938 clicks to the Adventure Trailers site ???

So what happened to the 1668 people who clicked on the banner ad and never made it to our site? I know there is a % who would have been dropped in the transfer but 86% ????

I don’t think the owner of the company is lying, but how do you explain the huge differences in statistics ?

 

[Ursidae69]

Maybe the BOT programs that are crawling all over the net?

 

[articulate]

I've noticed that web stat programs are very inconsistent from brand to brand. They all calculate differently, particularly referrers vs. clicks. So, the two of you could be using different tracking software and looking at the results from different formulas.

Have you considered using a landing page for click-throughs from the site in question? It would be a page that perhaps looks identical to your home page but maybe it has a welcome message, "thanks for clicking over!" kind of thing - or put a redirect code in the head tags to your home page. Your statistics package ought to show you page views for individual pages, so if you have a dedicated page for these clickers, you'd know how many times that page has been loaded.

Also, if your budget allows, you can outsource the tracking through a "click tracking service." I have not used these before, but many "large" companies do. I suspect this is a worthy endeavor and I keep meaning to research the costs, benefits, and consequences.

Another note, a free and very cool service is through www.quantcast.com for thrid-party web ranking and traffic analysis. You have to insert a little script on your home page, but it's totally invisible to your website users.

 

[articulate]

Alright, just for clicks and giggles: http://www.quantcast.com/adventuretrailers.com

 

[Martyn]

Alright, just for clicks and giggles: http://www.quantcast.com/adventuretrailers.com

Interesting results from Quantcast. Their estimate of just under 5000 unique visitors is slightly off, we have an average of 16000 visitors per month. But they did state " We have sparse data about adventuretrailers.com, so these are rough estimates.

I did like the "The site appeals to a primarily older, rather male crowd." sounds like the Portal People !!

Articulate I think you could be right that the software is tracking referrers vs. clicks differently. It would explain why we feel we are getting poor results and they feel they are doing a great job referring people over to us.

And I also like the idea of landing pages.

 

[shane4x4]

I HIGHLY reccommend Google Analytics (for you & them). You can setup several different marketing campaigns, goals, etc. It tracks everything you thought you may want to know & then some. It's free too

 

[Dave]

I second Google Analytics. We just started using them a few weeks ago and I am impressed with the amount of data their service provides for free. I haven't compared their results to any other tracking software, but they did come highly recommended.

 

[IntrepidXJ]

Didn't know Google Analytics existed.....thanks for the heads up....just added it to my site to see how it works

 

[mountainpete]

Martyn,

Reporting programs are so varied in the way they track or interpret the server logs. It's usually just that many are configured in a way that inflates hit counts or doesn't remove crawlers. My team is actually building a reporting system right now that parses XSL logs from our portal to Oracle - messy business.

Mark's recommendation of a unique director page is also a solid and simple idea. Just setup a blank HTML page with a with a 0 second meta refresh back to your homepage. Then have that banner target that page. You can then very easily and accurately track the traffic from that banner in your reports. It gets time consuming if you have a lot of banners hitting your site, but if it's just one or two it's easy work.

Pete

 

[Martyn]

0 second meta what ????

OK, I just finished putting the code for Google Analytics into our ATreport.com web pages. I'll let you know in 24 hours what type of results I get.

Pete you said "Just setup a blank HTML page with a with a 0 second meta refresh back to your homepage." Can you explain how that's set up or do I need to crawl back to my "Dreamweaver for Dummies" book ?

It's the "0 second meta refresh back to your homepage." that scares me.

 

[Martyn]

Our main site Adventuretrailers.com gets all it's web static’s from Urchin.com which surprisingly is Google Analytics

We have always been very pleased with the in depth static’s you can call up, so now we have the same thing available on both of our sites!

 

[flyingwil]

You should be able to use Cpanel and use awstats... you can also use the Latest Visitors Stats to see where the links in are coming from. Along with Google Analytics you should be able to get a decent count as to links in.