This stuck out because it was contrary to what I had heard was a good practice by posters here. I can NOT for the life of me find exactly where in the Learning Center this suggestion was.
Regardless, it seemed that they were saying that Content did not hurt you in any way because your CTR only took into consideration Google properties (Google.com, Google.fr, etc.). Then, in the description of CTR, they seemed to contradict this because they said CTR was dependant upon number of impressions throughout the Google Network (including Content) divided by the number of clicks on Google properties.
I assume that the current "trend" is still to segment Search and Content into different campaigns but I was also wondering if anyone had some thoughts on Google's suggestions regarding this.
-----
IIRC, the "suggestion" was only verbal, not in slide text, and said something like this, "for this reason it is not recommended that you create separate ads for the Content Network". (don't remember what "this reason" she was referring to)
The main reason for doing this strange and labour intensive workaround is to set lower prices per click for content traffic hence making an acceptable cost per conversion.
i have already done this with one of my large campaigns, and i discovered to my surprise quite good ROI on the content network. there were a few adgroups that i paused on content only, and a few that i paused on search only.
if there is some reason not to do this, i would love to know.
For the rest, you must see what's G. target and what's yours!
If you have only one campaign S+C, and for Search you afford paying say 1$/click and get a good ROI, but you will probably pay 0.8$/click for Content with a much lower ROI.
if you split campaigns into content/search, then it seems google cannot use CTR data from the search network to apply to rankings on the content network.
so is it the case, that ctr on content only affects content, and ctr on search only affects search?
any ideas on this?
What happens on Content is a little bit of mystery. The factors are the bid (be careful, the bid for the keyword more appropriate for each click) and probably, "how related" is that Keyword and Ad to the "theme" searched.
i have also thought about doing even further splitting for geotargeting. that way i could drill down and disable the specific adgroups that werent providing good ROI.
We do have campaigns where content is converting fine (a minority but still). Obviously there is no need to split campaigns then.
But I'm wondering abour Smart Pricing and it's relationship to splitting campaigns.
My understanding has always been that by splitting the advertizer could possibly save money.
But wouldn't Smart Pricing manage the difference between the click cost on search and the click cost on content?
I'd appreciate hearing advertizers perspective on that.
take the following example... you have a campaign with 2 ad groups - one aimed at "purchasing widgets" and the other at "repairing widgets"
if you dont split, everything is lumped together, and suppose you have positive ROI overall in both the content and search networks.
but, suppose you split the campaign into two idential campaigns, and after a week or 10 of testing, you learned that both adgroups on the search network had positive ROI, but that one adgroup on content was actually losing you $50 per day, but the other was earning you $200 per day. your overall gain on the content network is +$150... but if you turn off the loser on the content network only, you get $50 per day more profit.
aside from the theoretical example above, i have actual cases when it was beneficial to turn off a couple of ad groups in content only (not search); and believe it or not, there were some cases where it was beneficial to turn off a couple of ad groups on search only (not content).
this doesnt even take into account the ability to adjust max cpc for each ad group by network.
so... unless there is some disadvantage to splitting, it seems to me that splitting is almost always in the advertiser's best interest.
For example:
We allow targeted terms like "big blue wooden widgets" onto the Content network because they are less-likely to be displayed for generalized searches, and less-likely to be abused by AdSense publishers, as the CPC is lower for targeted terms. Our experience has shown that when a PPC ad shows for a targeted term, the click-er is more likely ready to buy, so this improves our ROI for visitors coming from the Content network.
We allow broader terms like "widgets" to be shown on the G network, because we don't mind people using G for research before buying, which is what happens with broad term visits, and repeatedly seeing our domain in the results as the search gets more targeted (more ready to buy) builds some trust ... which makes the sale easier in the end.
So: cheap, targeted terms on the Content network and broad, more expensive terms in the G network. That's our general approach to G.
