Why do ad manager add a lot of parameters to URLs

Question

It seems that each time I see an ad that is served by an ad manager application there is always a bunch of parameters added to the URL of the product.

Say for instance one random stackoverflow ad : http://ads.stackoverflow.com/a.aspx?Task=Click&ZoneID=4&CampaignID=474&AdvertiserID=5&BannerID=408&SiteID=1&RandomNumber=464183249&Keywords=

or this one:

http://ads.stackoverflow.com/a.aspx?Task=Click&ZoneID=4&CampaignID=474&AdvertiserID=5&BannerID=408&SiteID=1&RandomNumber=2039490120&Keywords=http-1.1%2ccaching%2ccache%2chttp-header-fields%2cheader%2cx-user-registered

If I go with the logic of the things, when you register a click to a banner, you would normally need a few info : "how many times it has been clicked", "by who" (ip/registered account/...), "when".

Now if we look at the parameters there are a lot more informations to this. OpenX adds a lot more on top of that :

bannerid=244_zoneid=7_cb=1264705683_maxdest=http://www.smartadserver.com/call/cliccommand/3141468/1264705683">http://ox.jeuxonline.info/www/delivery/ck.php?oaparams=2_bannerid=244_zoneid=7_cb=1264705683_maxdest=http://www.smartadserver.com/call/cliccommand/3141468/1264705683

The only reason I can think of this is to save call to the db, other than that I really can't see.

Any hint or ideas ?

Answer 1

Tracking.

The parameters are used to identify you, where you're clicking from, what you're clicking on, which ad campaign your click should go to, etc.

Answer 2

Well they also need to give analytics to their advertisers. You'll see that the URL identifies who the advertiser is, what campaign it's for, etc. All that stuff is used to know who to bill, and how effective the ad is (ie is the picture of the girl more effective then the unicorn).

Answer 3

Usually ad requests contain a size code, sometimes a location code (where on the page the ad appears), and audience segmentation information, which is information the content provider knows (or has guessed) about you and is used by the network to figure out what kinds of ads should be served to you. Most audience segmentation is proprietary, either to the ad network or to the content provider. Some ad networks have special features to support segmentation based on content you've recently searched for (keywords, search_terms, etc.).

Most of the time the name/value pairs will be completely unintelligible to people not familiar with the segmentation rules. You could have something like il=5, which could mean income level is in category 5, which might mean $120,000/yr on some arbitrary network.

The actual image served may repeat some of the information in the original ad request, either for tracking actual ads served or, in some cases, because the image server needs some of the same information to serve the right image.