What's the easiest way to do this? I want to compare CTR, clicks, impressions, conversions etc, but it appears that it's all lumped together.
Any advice appreciated.
Separating Google Search stats from Search Network stats has been been one of the most requested features from Adwords users for years.
Google is, IMO, guilty of distribution fraud = selling search network as 'search' traffic, when in fact a large part of it is contextual traffic or domain traffic that converts - surprise - at a much, much, lower rate than search traffic.
[edited by: RhinoFish at 1:16 pm (utc) on Mar. 25, 2008]
This is where Google loses all credibility re. the quality issue.
All I know to recommend is to run with and without search network for two equal test periods and compare the results to try to measure the impact of the marginal cost, sales, etc when the Search Network is in play.
so producing 2 campaigns (1 with Google only + 1 with Search Network only) could be an idea?
Unfortunately you can't do this. You can turn the search network off to run a Google only campaign but you can't turn the Google network off to run a Search Network only campaign.
I think what he was suggesting was to run one campaign with just Google enabled and track the results, then enable the search network in the same campaign and run it for the same amount of time and compare the results to see if enabling the search network hurt or helped.
I see Google doing some (not all) stuff in the way where as long as people make a penny through them, Google will continue (mis)using that fact.
We all cry but we all still advertise.
You could create 2 campaigns and in "Google only" set your bid to what you want it to be. In “Google + Search” set you bid for 1/4 or more lower then in the first one.
I haven't tested this setup yet but what might work is to have two identical campaigns, one for Google only (Campaign A) and the other for Google + Search Network (Campaign B). Set the bids for ad groups in Campaign A $0.01 higher than the corresponding bids for Campaign B. Then, on Google Campaign A will be shown because it has the higher bids, and on the Search Network Campaign B will be shown. That should allow you to better analyze the performance of the Search Network.
Doesn't CTR come into play?
Remember this is all work around and not a dedicated solution. There is no perfect thing here. Even with stuff that Google has made for us to use, we still get “so so” results. Look at the “search query report” example. It's not perfect. Yet, it is still better to have it as it is, than not to have it at all.
Anyway, you can do the setup as suggested, and use some tools in order to figure which campaign is showing on Google. I can think of two so far:
- ad diagnostic tool
- AdWords preview tool (here you would need to make your ad text different)
Correct. So the best you can do is to run equal time test periods under each, assume the "Google" performs at same ratios during both your "Google" and "Google + Search Network" runs, back out the "Google" results from the latter, leaving you with the results for "Search Network".
It's not easy or clean and it introduces test variability, but sometimes you'll want to know what the Search Network is doing and a relativistic inspection is about all you can do.
As time passes, G gives us more and more info and more and more control, so I'd expect that we'll eventually have split-able reporting or at least the ability to run Search Network stand alone (without being forced to run it in combo with G itself). I look forward to then. :-)
