Optimize Facebook & Google ad campaigns

Posted by on Nov 12, 2012 in Advertising, Search, Social

Last week I posted some basics about launching ad campaigns and setting up ways to track my progress. This week, I want to take a look at my first week’s results and explain what changes I plan to make to optimize my Facebook and Google ad campaigns and why.

So lets see how I did in my first week:

Advertising metrics

Week 1 Dashboard

My dashboard shows some good and bad signs. Most notably, I had three conversions for the week – two on Facebook and one on AdWords. So in three cases, someone saw an ad and signed up at the AdStage homepage. That’s why I’m running these ads, so it’s a positive sign. CPA was about $12 for AdWords and $12.50 for Facebook. I think I can bring these down and analyze what impacts CPA most in coming weeks. This week, there’s no previous state to compare against. Right now my CTR appears to be low – 0.052% on AdWords and 0.03% on Facebook and clearly my campaigns on Bing and LinkedIn were not effective, thus I’ll be making changes for next time.

Here are two good Quora discussions about CTR on both AdWords and Facebook ads. Keep in mind that a campaign’s goal should ultimately be to drive conversions, and it is possible to have a high CTR without driving conversions at a low CPA.

What is a good CTR in an AdWords campaign?
What is a good CTR for a Facebook ad?

In my dashboard, I can also see best and worst performing ads, all of my top performing ads, and other high performing ads of similar companies or companies in a similar space.

Ad performance

Highest and lowest performing ads

Online ads

Other high performing ads & ad suggestions

These insights are really useful, both in making changes to my less performant ads and in incorporating elements of effective similar ads into my own ads.

Social campaign changes

In setting up my first campaigns, I wanted all ad text, headlines, and images as consistent as possible. I discovered a few differences that I adjusted for week 2. This is an issue I ran into partially because updating and adding new ads to each campaign while comparing for consistency across networks can be a pain.

For Facebook, I highly recommend using the power editor once you have more than a few ads and definitely once you have multiple campaigns. You can perform lots of useful actions that would otherwise move very slowly in the typical interface. For instance, you can duplicate ads and whole campaigns, and you can quickly edit ad details like the ad copy and images.

Here’s what the power editor looks like. You can see many options within the interface. With your own ad data in the editor, you’ll see it’s real power.

Facebook power editor

Power Editor

LinkedIn and AdWords have better interfaces and you can easily duplicate and edit your ads.

I should have also mentioned last week that you can import AdWords campaigns into BingAds. This will speed up launching your BingAds campaigns and it will ensure that your Bing and Google campaigns are the same. (I’ll warn you that Bing calls this a beta feature, and the first time I tried it, it did not work. However, the second time I tried, it did work.)

Bing Import from Google AdWords

Bing Import from Google AdWords

My next optimization was in network targeting. As I mentioned above, I had a few inconsistencies between my Facebook and LinkedIn ads. For my cross-channel campaigns, I want as much consistency as possible. Beyond the actual ads themselves, I also had targeting inconsistencies within my Facebook ads. To fix this, I took the targeting my top performing ad and made it the targeting for all of my Facebook ads. This way, I can start comparing apples to apples (ie. i can compare the ad headlines, ad text directly since the targeting is the same).

LinkedIn targeting needed a lot of change. I initially setup to target job titles and groups. This gave me an audience of about 20,000 users related to marketing and advertising. I expected my ads to be highly relevant to the audience and result in a high percentage of conversions. Instead, the audience was so small and I had hardly any impressions at all and didn’t engage any users. In week two, I’ll try targeting job function to gain a broader audience of about 1,000,000 users. Those users will still find my ads relevant, but the larger audience should result in more impressions, clicks, and hopefully more conversions.

Search campaign changes

My search campaigns also needed some optimizations – primarily concerning my keyword bids. On BingAds, I somehow didn’t change my chosen bid from the minimum allowed bid of $0.30. Beginner’s mistake! This was the first thing I did to improve my BingAds campaign. I checked my AdWords campaign’s keyword bids (which I set to auto-bid) and the average CPC was about $2.00. I switched out the old minimum bid on BingAds and matched it to my AdWords average. As I move further along, I will do more analysis on these changes and the predicted improvement I expect to see. For now, there are more obvious adjustments I need to make first.

While the auto-bidding for AdWords worked well since I didn’t know quite what to bid, it also resulted in an overpriced, low value click. Partly at fault for this click was that I included many keywords in my first week, some of which were not highly related to AdStage’s primary audience. Also partly at fault was that I did not set a bid maximum in AdWords. Specifically, I paid $4.14 for a click on an ad that displayed when someone searched for “ppc programs“. The chance of a conversion coming from that search term would seem to be very small.

Here’s a look at my week one keyword activity and my dashboard suggestions for new keywords:

There are three major lessons that come to mind specifically from week one.

  1. Pay attention to your keywords. You may find a good list or come up with your own list, but make sure most of your keywords are directly related to your product and company. Otherwise you risk paying for clicks that will not result in conversions. Google provides a good keyword tool that can help you here. You can do some simple searches to figure out possible good keywords.
  2. Make sure your bids are thought out. While you might not know anything about bids right away, there are some quick ways to learn. Bing offers good information like – whether your bid is likely to place your ad on the first page, the first ad slot, or otherwise. Set a bid that will land you the result you’re looking for. For me, a mainline (an ad placed in any position 1-4 in the top of search results) position and bid is the goal. Though, paying a little extra to guarantee top ad placement may be worth it for certain keywords. It all depends on how you much you value a given keyword. AdWords gives you less information about bidding, but a quick tip is to use auto-bidding and set a bid maximum. That way you’ll get the AdWords intelligence, while not paying more than you’re comfortable with for any given keyword.
  3. Double check your targeting. A good balance of size of audience and relevance of audience is most effective. As I saw, my relevance may have been good, but my audience (at least on LinkedIn) prevented my ads from getting enough exposure to have a successful campaign.

Those were my changes in the first week. Some of these changes could have been addressed before I even launched campaigns in the first place. This is part of the learning process and I’m already learning to take advantage of the tools I put in place specifically to help me optimize my campaigns. Hopefully pointing my mistakes out and showing how and why I’m changing my campaigns will help you build your own successful campaigns.

Thanks for reading and check out AdStage if you would like a great, simple way to compare your campaigns and get insights to improve your campaigns in one, well designed dashboard.

If you have any questions or need any help with your campaigns, please let me know. You can reach me at clark@adstage.io with anything on your mind – questions, comments, critiques, suggestions for future posts, or just to say hi.

Online advertising tools

Clark Gates George