| 7:47 am on Apr 28, 2006 (gmt 0)|
It's becoming more and more difficult for me to be a full and ethusiastic proponent of the search network (that I've been in the past). You really must do your best to be there as the vast majority of traffic is of high-quality... driven through sites which include: AOL, Amazon, Comcast and more.
But, more and more absolutely worthless sites, including "parked domains" and arbitrage shopping sites, are jumping in the mix... Relatively few in number, but they deliver a substantial number of clicks and they are growing... quickly. A shame, too, because the few are reducing CPCs and, therefore, revenues for the many quality partners.
In summary: Search syndication was great, is good, is increasingly becoming average to poor.
| 5:12 pm on Apr 28, 2006 (gmt 0)|
I'm doing tests on this right now.
I've created a new camapign for the search network (bids were lowered for this campaign so they would not run for Google traffic) and have Google traffic only running on my old camapaign.
This is only the third day, but so far results are interesting. Overall Google traffic delivers far more sales and loads more traffic. The Search network has higher conversions for some (not all) of my main keywords but total conversions and traffic are much lower. ABout 5x more sales through Google only.
If I made a decision based on this preliminary data I would shutoff about 90% of my search network traffic and about 20% of my Google traffic. All in all it could save me at least $300/day if not more in expenses with only marginal loss in sales.
Losing sales stink but to potentially save $9000+ in advertising costs is far more than the lost sales. Granted there is not near enough data to make a decision, after another week I will have more accurate data to work with.
I will say from the looks of things very specific or "niche" keywords get very little traffic on the search partners (for my market). It seems to be the higher volume, higher priced, more competitive keywords that get the majority of traffic.
So to run a truly accurate test on my more specific/targeted kewyords could really take months to get enough click volume to make a real guess. Then you have the decision on whether to start testing entirely different ads and landing pages for just the search network :-)
All this testing takes time but for a while I've had a separate campaign for the content network and a campaign for Google and the Search Network. What I've discovered on that front is different ad creative and landing pages have helped double content network coversions.
Well, that's my results so far.
No doubt, like any technique or strategy it's a little different for every market but I think it is well worth testing.
[edited by: jkwilson78 at 5:15 pm (utc) on April 28, 2006]
| 5:15 pm on Apr 28, 2006 (gmt 0)|
poster_boy is spot on -
once you get beyond search partners like AOL and Ask the click traffic becomes highly suspect. Over the past year I have noticed that a larger portion of clicks coming from the search network are not passing the URL and not suprisingly these clicks also are not converting.
Unfortunately the good traffic I get from the better search partners is now being outwieghed by #*$!ty traffic from the poor quality sites (who are attempting to hide who they are). Net result is that I need to shut down this channel.
I have been discussing this at great lengths with Google management and have been fighting an uphill battle. My take is the Google does not want to be in a position to throw away revenue from these crappy search partner sites.
My suggestion is that Google allows an Adv to either opt in or opt out of search partner sites. The same way it can be done for Adsense. Until this is done i will not be allocating any of my significant ad budget to this channel.
| 5:20 pm on Apr 28, 2006 (gmt 0)|
jkwilson78 - when you say conversions are you referring to imp to click conversions or click to sale/lead conversions?
I like you have also noticed that my high CPC terms got shown and clicked on the most. Again my feeling is that Google is funneling these terms to partners in hopes of driving up more revenue.
Also keep in mind that if you do broad match on the search network then Google will be very liberal in showing your ad. I highly recommend to only use Exact/Phrase match when using the search network - if at all.
| 5:40 pm on Apr 28, 2006 (gmt 0)|
|jkwilson78 - when you say conversions are you referring to imp to click conversions or click to sale/lead conversions? |
Clicks to sales conversions.
But regarding click to impressions, my CTR is much lower on the search network for my higher traffic keywords and in numerous cases impressions on the search network are higher than Google only traffic.
This tends to back up the speculation that there are a lot of junk/crap sites in the search network. I just don't believe that second and third tier, search engines get more impressions than Google. So there has to be a lot of junk out there.
Now for a perfectionist like me I can't just turn the search network off, I can't help but think if I try some ad copy and landing page testing for the search network I can squeeze more sales :-) Then again, I tried this with numerous 2nd and 3rd tier PPCSE's and after way to much time and money accepted the fact that the traffic is pretty much all junk.
Another example, one adgroup usually got around 3.6% sales conversion. Now that I separated the traffic it's around 6.1% on Google only and ZERO on the search network.
Still only 3 days in to this test, but it's easy to see how important testing something like this should be to every advertiser.
| 9:48 pm on Apr 29, 2006 (gmt 0)|
|...Now that I separated the traffic it's around 6.1% on Google only and ZERO on the search network. |
How exactly do you separate the search network traffic? It is impossible to turn ONLY the search network on.
I can see only this way now - you have Google+Search network in Campaign 1 and Google only in Campaign 2 and you are bidding higher in Campaign 2 than in Compaign 1. Right?
| 11:55 pm on Apr 29, 2006 (gmt 0)|
well, don't blame parked domains,
you can't expect to make a sale with every 2 cents you pay for click.
| 10:54 pm on May 2, 2006 (gmt 0)|
|It is impossible to turn ONLY the search network on. |
I wouldn't say impossible but not very practical to make work :-)
|I can see only this way now - you have Google+Search network in Campaign 1 and Google only in Campaign 2 and you are bidding higher in Campaign 2 than in Compaign 1. Right? |
Exactly. Works like a charm too. To test it I make one very slight change to my ad. Say something like adding or removing a period for my Search Network campaign and as long as my search network bids are lower I have never been able to make those ads show on Google.
A few minutes after upping the bids above my Google Only campaign the ads in my Search Network campaign start to show.
So it isn't impossible just not very practical.
| 11:01 pm on May 2, 2006 (gmt 0)|
|well, don't blame parked domains, |
you can't expect to make a sale with every 2 cents you pay for click.
No you can't but I can attempt to maximize the value of every single click and why should parked domains be included in the Search Network anyway when no active search is taking place? And I don't consider typing in a domain name search traffic although I can see how Google could technically justify that a search or intent for certain information was taking place by typing in the domain.
They should be included as part of the content network in which case I can exclude the sites delivering crap traffic.
That's not to say this isn't how it works but I have a feeling since GoDaddy and other big registrars are so big some might be considered a "network" of traffic and lumped in with search partner traffic.
| 11:31 pm on May 2, 2006 (gmt 0)|
One final post in my string on replies.
I've been messing around with the position preference feature to see if I could find a way to better separate search partner and Google traffic.
Here's what I mean.
I have a keyword that performs much better through the search network than with Google only traffic. This keyword has a campaign setup for Google traffic and Search only traffic.
My test was to see if I could bid to a lower position for the Google only campaign and not have my ads for the search only campaign kick in. My reason for thinking this might work is that my max CPC is a good deal higher for Google only traffic as is my CTR and that in using the Position Preference feature, Google would be taking this higher CPC & CTR into account in order to lower my ad.
Nice thought perhaps but unfortunately it doesn't work. It took almost two days but eventually I started seeing my Search Network ads show in place of my Google only ads.
When I turned Position Preference off it went back to normal. Makes sense Google would think of and block this kind of work around.
So two questions remain:
1. Anybody else try a similar work around with success?
2. Does not allowing advertisers the ability to separate search network traffic really help Google make more money in the long run?
In the end I am going to greatly reduce my spend on this keyword when I could be decreasing spend in one area and increasing significantly in another.
The future success of PPC advertisers is going to significantly favor those that test and track fanatically and get very granualar in their approach and this will hurt Google revenue in the long run as more and more people learn EXACTLY where their sales are coming from.
| 11:46 pm on May 2, 2006 (gmt 0)|
I have done numerous campaign segmentation testing with Google (i.e. IP targeting, search network, content network, etc...).
My recent conclusions have been that the Search Network sucks, while AdSense sucks the big one. In Google's attempt to gain reach they have totally neglected to impose any sort of quality control and fraud detection around their distribution partners. Certain partners like AOL and Ask do generate good results, but the problem is that the bad sites totally outweight any benefit. While Adsense is a complete crap shoot.
Google needs to let the Ad buyer filter in/out specific distribution partners.
|wired in asia|
| 1:59 pm on May 3, 2006 (gmt 0)|
does anyone know what the major sites are in the google Search network (besides AOL..)...?
| 2:46 pm on May 3, 2006 (gmt 0)|
you can begin with this:
|wired in asia|
| 3:47 pm on May 3, 2006 (gmt 0)|
yes, guess should have done my homework;-)
having paused all overture US ads the other day and not felt a dent in inquiries, I getting concerned about the sudden increase in google impressions and clicks, yet revenues remain the same.
in the case of overture US, this turned out to be a clear indication of poor click quality.
having now deactivated the 'search network' for one of our campaigns, I wait and see...
| 11:24 am on May 4, 2006 (gmt 0)|
Ok, let us know the results of the test.
Unfortunately (or not?), the search and content network seem to be different for each country, so a conclusion is relevant only for that country.
|wired in asia|
| 3:39 pm on May 4, 2006 (gmt 0)|
well... close to 24 hrs later and ...
CTR right up and cost per click are down. Too early though to truly judge. Need to watch this another a few days I think.
| 10:27 pm on May 4, 2006 (gmt 0)|
After 8 full days of testing the search network against Google only it's pretty obvious that...there are no definite answers.
From the looks of things, I can easily eliminate half of my Search Network Only campaign without taking a hit in sales.
The majority of the remaining keywords convert just as good or better than the Google only campaign so they will remain.
The problem is with the keywords that convert bad on Google but great on the search network. These keywords cannot simply be deleted or bids lowered from the Google campaign because the ads from the Search Network campaign just take their place.
For these I will lower bids in both campaigns and try to strike a middle ground. Not much else can be done.
After another week I will be able to tell whether the changes hurt or helped overall.
|wired in asia|
| 3:35 am on May 5, 2006 (gmt 0)|
guess it really comes down to what biz you are and in what geographical location your potential targets are located (i.e. US/UK etc..).
let's say that 40% of sales comes from the US, AOL may add to overall sales. then again does AOL provide quality sales leads?
what I have done is that I simply deactivated the 'search network' and compare now to past stats. mind you, monies saved by de-activating the 'search network' could be used to obtain higher ranks under the google search alone. this in turn could lead to more leads/sales.
will keep on testing this.