How do you structure your adwords display campaigns?

M

mpbiz

New member
Apr 29, 2010
2,824
57
0
I'm about to expand over into google display (banner ads) and all I've ever done with adwords is ppc so I'm a little confused on how to structure my campaign in regards to tracking.

Right now I'm using 202. It's my own offer so here's what my setup looks like:

Affiliate network: "My Products"
Campaign: My offer's name
Method of promo: Landing Page
Landing Page: Google Content - SiteI'mBuyingTrafficOn.com - Angle 1
Ad Copy: Google Content - SiteI'mBuyingTrafficOn.com - Angle 1
PPC Network: google content

And then at the end of my links I'm using {placement} instead of {keyword}

In adwords I have created a new adgroup for every site I'm targeting. Do you guys create multiple adgroups for the different spots like ROS vs. Home page leaderboard too or am I over thinking all of this?

Do I even need to be using 202? Or can I just use google's conversion pixel and manage everything in their interface?

I kind of feel like I don't need 202 since the adwords interface is basically an ad server anyways.

Any help is greatly appreciated I need to make sure I do this right the first time since the approval process takes so much longer.
 


mediastar100 said:
I just use g pixels for conversions and retargeting.

Keeps it all in one place which is nice.
Click to expand...

Thanks for the reply looks like I'll be ditching 202.

Hopefully you can offer your 2cents on these as well..

1. I'm guessing you create a new adgroup for each new site and also for the diff. placements available for that site?

For example:

Adgroup 1: fitnessblog.com ROS

Adgroup 2: fitnessblog.com

2. Are you still inserting a token like {placement} at the end of your URL's?

I'm a little confused on how this works. Right now I am targeting a site and there are 2 options available: fitnessblog.com and fitnessblog.com ROS, Multiple locations.

I'm guessing I put a token like {placement} at the end of my URL's and it will display stats for all of the individual placements? Like for example if I'm targeting a forum and they show ads on every section of the forum, every thread, etc.

3. Last but not least, when you want to test a new angle, do you just upload your new ads with the new copy, and then simply let them run and view their stats plain and simple? Or is there any other separation via LP or adgroup when it comes to testing new angles?
 
PS. Just in case it changes anything, right now I am trying to do managed placements. No demo targeting or keyword targeting.. not sure if I should be.

I have a very niche product so I already have a list of all of the sites I need to target.
 
just use adwords pixel, p202 is unnecessary.

I also found that if I have all the presale on my product page, dlink converts better.

Also image ad does take forever to get approved, a week or so, but all you have to do is email support after you create your ads, it'll be approved in less than 3 days.

you should do text ad first (cheaper) and catch all the negative, unrelated keywords that might appear in your niche then create an image campaign, port those negative keywords over and set your base.

it'll save you some money. image ads are typically 30% to 40% more expensive than text ad.
 
  • Like
Reactions: Elvis and mpbiz
Jardini said:
just use adwords pixel, p202 is unnecessary.

I also found that if I have all the presale on my product page, dlink converts better.

Also image ad does take forever to get approved, a week or so, but all you have to do is email support after you create your ads, it'll be approved in less than 3 days.

you should do text ad first (cheaper) and catch all the negative, unrelated keywords that might appear in your niche then create an image campaign, port those negative keywords over and set your base.

it'll save you some money. image ads are typically 30% to 40% more expensive than text ad.
Click to expand...

+rep thanks for the help.
 
I'm guessing you create a new adgroup for each new site and also for the diff. placements available for that site?
Click to expand...
Yes that is correct. As with most ppc, the more granular the better as long as it's still manageable. Make the setup as granular as possible.

They changed adwords remarketing, you get one single tag you add to every page now, even your conversion page. Then you can build remarketing lists without using the normal adwords conversion pixel. You'll understand once you add the code to every page.

Are you still inserting a token like {placement} at the end of your URL's?
Click to expand...
Usually use {KeyWord:Wicked Fire} in the headlines to insert keywords dynamically. I think that can go on the end of urls too.

Do you use any display spy tools?
 
Why would you site target?

Google's own pixels work fine, and it helps with retargeting.

Build a small keyword list for the site you want to target. Set the spend at $25/day. Then block everything that isn't converting for you, or if you really only want that one specific site block everything else. Within a week you'll be on the site your targeting with a trickle of traffic from other sources in the network.

More importantly, you'll be paying less than the other affiliates targeting that site specifically through site targeting.

Oh, and the contextual targeting tool is your new best friend. Personally, I still like to edit my keywords manually and ensure everything is ship-shape. But, it does lay a nice foundation to build off of.
 
AYGJoe said:
Build a small keyword list for the site you want to target. Set the spend at $25/day. Then block everything that isn't converting for you, or if you really only want that one specific site block everything else.
Click to expand...

I want to make sure I understand what you're saying.

Are you telling me that instead of just targeting SiteXYZ.com, that I should be using flexible reach and using site placement combined with keyword targeting? This:

"Contextual and Placement Targeting - Want to pinpoint topics on certain websites?

Narrow your scope to reach specific types of customers on certain websites by using both contextual and placement targeting at once. If you sell tulip bulbs, for instance, you might choose a website about gardening (placement targeting) and then let Google technology find pages on that website that discuss tulips (contextual targeting)."

^^ Isn't the above overkill though? I really wanted to avoid keywords all together. It's so much easier to throw banners up on a site and call it a day. Am I being lazy, do I just not understand g display, or are you just really particular about how you manage your campaigns?

I've never done g display before, but the thing I loved about direct buys in the past is you simply rotate banners and weed out the losers. Managing keywords is a pain in the ass, so please enlighten me if I'm not getting something here.
 
Wong Shun Leung said:
Make the setup as granular as possible.
Click to expand...

I'm still confused on how to achieve this. For example for this one specific site I want to target, when I go to add a managed placement in the adgroup, it has options for:

SiteXYZ.com and then SiteXYZ.com::ROS, Multiple Locations

Whats the difference between the two? And how am I able to view specific URL data?

In my mind, from what I'm seeing right now in the adwords interface, I basically have to target those 2 options above and just blast the entire site.

Am I being a dumbass? Are those just the 2 general options? Here's somethign I just found:

"Tip

If you'd like to add multiple placements at once to your ad group, and you already have a list of websites in mind, click the Add multiple placements at once link at the bottom of the box. Type in there the placement URLs using this format: example.com or example.com/topic. "



So if I'm targeting a site thats a forum, is my only option to type URL's for the specific subsections of a forum and track that way? Or is there a way that I can blast the entire site and then google will show me the specific URL's within the site that are converting? For example if I find out the classified ads section of the forum sucks then I can specifically block it.
 
And just so you don't think I'm one of those morons who asks questions just to argue and not listen. Here are the reasons why I am so intent on targeting specific sites with g display:

1. By targeting a specific site, I can match the creative's colors with the site's and make it look like content to increase my CTR.

2. g display makes it easy for me to get my banners up and running as fast as possible and then once I find the winners I can go direct.

3. My offer is very niche. I am not going to be able to run this on general news sites, or even blogs that might have some of my keywords in a post. My offer matches perfectly with the sites I want to target.
 
SiteXYZ.com and then SiteXYZ.com::ROS, Multiple Locations
Click to expand...
Not quite sure but it seems they like you to show ads on the entire site, probably because it makes google more money. Although I think you can find single page placements in there. If there is a single page placement it should clearly say that it is only that page.

Have a campaign targeted to the specifc reach of the display network only, with one single page placement per adgroup only, and then create a specific ad or ads, for that specific page/adgroup. Then you should only accrue impressions on that page, for that ad in that adgroup. Use the exact URL/specific page/placement, as the one and only targeting option for the adgroup, and only include the ad(s) in that adgroup you want shown on that exact URL.

Probably worth going over the adwords learning centre to recap on the display network part.
http://support.google.com/adwords/certification/bin/static.py?hl=en&page=examstudy.cs


how am I able to view specific URL data
Click to expand...
You should find an option to run a GDN report and hopefully that will document the exact pages and where the conversions came from.
 
  • Like
Reactions: mpbiz
It's not that I'm particular about my campaigns its about profitability. You'll get more traffic/traction on display if you go the keyword route instead of the generalized site target.

For me GDN isn't just a display network that I media buy traffic off a few sites from. When I find veins that convert and are profitable I want to expand them as much as possible.

It's all a matter of what you want out of this.

After your campaigns have ran for awhile and have conversion data, Google's new DCO function is autopilot gold. I'd do some research into it if I were you.

Another option is doing image search campaigns on top of your existing display.

I understand that it is an extremely niche market, that's the reason you start your budget so low per campaign you build. So you can immediately start weeding out the sites that aren't working.

As you weed sites out, you should be increasing your budget. Personally if a campaign is starting to show promise after the first 4-5 days of testing I start doubling the budget daily until I find where I'm going to cap out on volume.

Day 4 $25 -> $50
Day 5 $50 -> $100
Day 6 $100 -> $200

Of course a lot of this is going to depend on what kind of budget you have as well.

Matching banners on a per site basis is a decent idea. I've done it previously in the gaming niche and it has paid off in double sometimes triple my normal banners CTR. However, you still want to run remarketing and a general display campaign as well.

I use all of the tools available to me, why don't you want to?

If you'd like give me a campaign idea and I'll build something out so you can copy the structure for your own campaigns. :)
 
  • Like
Reactions: mpbiz
AYGJoe said:
It's not that I'm particular about my campaigns its about profitability. You'll get more traffic/traction on display if you go the keyword route instead of the generalized site target.

For me GDN isn't just a display network that I media buy traffic off a few sites from. When I find veins that convert and are profitable I want to expand them as much as possible.

It's all a matter of what you want out of this.

After your campaigns have ran for awhile and have conversion data, Google's new DCO function is autopilot gold. I'd do some research into it if I were you.

Another option is doing image search campaigns on top of your existing display.

I understand that it is an extremely niche market, that's the reason you start your budget so low per campaign you build. So you can immediately start weeding out the sites that aren't working.

As you weed sites out, you should be increasing your budget. Personally if a campaign is starting to show promise after the first 4-5 days of testing I start doubling the budget daily until I find where I'm going to cap out on volume.

Day 4 $25 -> $50
Day 5 $50 -> $100
Day 6 $100 -> $200

Of course a lot of this is going to depend on what kind of budget you have as well.

Matching banners on a per site basis is a decent idea. I've done it previously in the gaming niche and it has paid off in double sometimes triple my normal banners CTR. However, you still want to run remarketing and a general display campaign as well.

I use all of the tools available to me, why don't you want to?

If you'd like give me a campaign idea and I'll build something out so you can copy the structure for your own campaigns. :)
Click to expand...

Just sent you a PM.
 
AYGJoe said:
Build a small keyword list for the site you want to target. Set the spend at $25/day.
Click to expand...

Man, I don't know what niche you are in, at $25 dollars a day you can barely get any accurate data.

Conversion varies throughout the day, you want a budget big enough to at least have the ability to spread the ad display through out the day so you have the correct data to optimize.

Even if you are paying 20 cents a click, $25 is still not enough. Try $250 a day budget to start off at least......
 
Bump.

I finally figured out how to view specific URL data so you can see if certain areas of a site are not converting for you. I was looking at and thinking about g analytics and adwords as seperate entities and you should be looking at them as one.

Here's the solution:

In analytics you go to Traffic Sources > Referrals. And then it will just show you the general domain like SiteXYZ.com

Then you click on SiteXYZ.com and it will break it down and show you the specific referral path URL's. On the same page you go up by the graph and it will have an "Explorer" tab. Underneath the explorer tab it will list "site usage", "ecommerce", and then all of your goals you have created.

You just click on the goal that you want to segment by and then it will show goal conversion data for the specific referral paths. Then, if you notice that a particular page, subforum area, etc. of a site is sending you crap traffic, you should be able to go in and exclude that specific URL in your campaign settings.
 
Jardini said:
Man, I don't know what niche you are in, at $25 dollars a day you can barely get any accurate data.

Conversion varies throughout the day, you want a budget big enough to at least have the ability to spread the ad display through out the day so you have the correct data to optimize.

Even if you are paying 20 cents a click, $25 is still not enough. Try $250 a day budget to start off at least......
Click to expand...

$25 per campaign will net me over 300 clicks on GDN in one niche and 150 in the other? I used it as a starting point.

I should also mention, I normally launch 10 campaigns at once. I've found that drilling into specific bid types on GDN yields the cheapest results with volume for me.
 
Jardini said:
Man, I don't know what niche you are in, at $25 dollars a day you can barely get any accurate data.

Conversion varies throughout the day, you want a budget big enough to at least have the ability to spread the ad display through out the day so you have the correct data to optimize.

Even if you are paying 20 cents a click, $25 is still not enough. Try $250 a day budget to start off at least......
Click to expand...

I agree, $25 a day isn't enough data to know anything except maybe for 2-5 keywords.
 
You must log in or register to reply here.
Top Bottom