There can be many reasons for these discrepancies. Unfortunately it's
quite difficult to pin down given your circumstances. One major issue
is that referrers can be stripped in some cases, but not in others.
This is particularly the case with Flash, if not structured properly.
Also, bear in mind that clicks as reported by DoubleClick do not equal
visits as reported by your site analytics tool. In fact, there are
usually major differences in just clicks reported by an ad server vs
clicks reported by a site analytics tool. I've seen as much as 70%
difference in clicks alone. Since you can record multiple clicks per
one visit, this difference become exacerbated when comparing clicks from
an ad server to visits from site analytics.
I realize this situation isn't ideal, but one option you have is to look
at traffic to the landing page before and during the campaign period.
If you set a pre-campaign baseline, then extrapolate that out during the
campaign period, you can get some sense of lift. If you have other
landing pages that were not used during this campaign period, I would
also recommend using those as a control to account for any site-wide
lift that may have occurred during the same timeframe.
Hope that helps!
Vice President, Best Practices Group
Inside Web Analytics - Check the blog today!
550 East Timpanogos Circle
Orem, UT 84097
On Behalf Of richgrog
Sent: Tuesday, October 03, 2006 11:38 AM
Subject: [webanalytics] Identifying traffic from online advertising
Here's my issue:
I work in an ad agency, which does a lot of online creative for our
clients. One of my clients asked us to pull together an analysis of
site traffic during the time an online advertising campaign was
running. We were brought in after the campaign went live so we had
no opportunity to suggest the use of campaign variables for
tracking. Our client uses WebTrends as their web analytic tool. I
was looking at referring sites and noticed that the visits associated
with the DoubleClick domain was significantly lower than the clicks
on OLA units that should have been driving traffic to the site (7,757
visits vs. 206,428 clicks). These are Flash OLA units and I've seen
Flash units with as high as 50% drop-off rates, but this doesn't
account for the loss of all these visits to the site that should be
attributed to the OLA campaign. The media agency didn't tag the
landing page so I can't get a number from DART. Could it be that all
the other visitors are being bucketed in with the Direct/Bookmarked
traffic? Why would some be associated with the DoubleClick domain
and others not?
Thanks for your help in advance,
[Non-text portions of this message have been removed]