Google AdWords: Campaign Experiment BETA

A fairly new addition to the current selection of Tools in Google AdWords is the Experiment BETA.

The next time you are editing your campaign settings take a look and see if your account has been selected for the Beta release of “Experiment” – If you expand this feature in your campaign settings, a list of 4 steps will expand which explain how to design a test:

1. Specify experiment settings.
2. Make experimental changes to bids, keywords, and ad groups in your campaign.
3. Start experiment. As traffic accumulates, statistically significant differences may emerge.
4. Evaluate experiment. Apply changes fully or remove changes.

Before we get into how exactly to create an Experiment let’s talk about what it actually does and possible reasons to utilize it.

Campaign experiments allow you to segment traffic to keywords by max bid. For example, you have a keyword in the 4.5 position with a $0.30 bid. Campaign experiments give you the ability to take a % of total traffic and have it run with a different bid. In this example, you can pull 20% of this keywords traffic and modify the max bid by a certain percentage; let’s say 50%. This means that 1 in 5 impressions will be running off of a $0.45 bid as opposed to a $0.30 bid. By using this you discovered that if you bid $0.45 all the time, you can be shown in the 2.6 position. Along with these changes you can see side-by-side performance changes in CTR, CPC, Conversion Rate, etc…

By using this experiment you can effectively test bidding strategies with keywords without worrying about peaks and valleys in daily/weekly/monthly search patterns because they run congruently.

• Perhaps you are in the 8th position and want to know what it would cost to get in the 2nd position. You can segment 10% of the traffic and bump up the bid 150% to see where that gets you and adjust as the data rolls in. This way you can be sure that there are no over-spending surprises.
• Maybe your client/boss wants to be in the #1 position (No ifs, ands, or buts about it!) while the data shows that the 4th position is the most profitable. You can segment the traffic and neatly show the performance changes to with prove or disprove a hypothesis.
• You can also use this to help lower your cost. Say you are doing well in the 1st position – You can set an experiment up to see if you still perform as well paying 20% less than you currently are.

Now… Back to the 4 steps:

Step 1 – Click the [+ Specify experiment settings]. This will open up a new menu. Specify the Name, Control/Experiment split or how much traffic you want going to your original bid vs. your experiment bid, and Start/End dates. Click Save.

Step 2 – Go to your keywords. In the tool bar below the Keyword Tab, you will see Segment. Open this an select Experiment. You will now see a Test Tube icon next to your keyword. Edit your keyword and now you will be able to modify your Experimental Bid.

Step 3 – Once you are done setting up your bids, return to the campaign settings and select “Start Running Experiment”

Step 4 – Based on the results you will see side-by-side on the Keywords tab, you can apply the bid changes or delete them.

Google provides the tools, it’s up to us to figure out how to best utilize them. Has anyone else found a unique way of using this feature to optimize their account? Please post your findings.

As I’m sure you’re aware, there are 4 toll-free prefixes – 800, 888, 877 and 866.

See which “800” prefix yields the highest conversion rate in this search marketing industry study performed by Engine Ready.

Although the “800” prefix has been around for over 40 years, its next oldest sibling, “888” is only 13 years old. The “877” prefix came about 11 years ago, while the youngest “866” has only been in use for 9 years.

Does the longer legacy of the “800” prefix result in higher conversions when tested against the newer toll-free prefixes?

Everywhere you look, there is no shortage of hype about how Web 2.0 is revolutionizing the marketing and promotional world. I’m not a skeptic of the potential of social media, networking, user generated video and blogs, but wonder why there seems to be a disconnect between the Web 2.0 and Web 1.0 worlds.

The real opportunity for capitalizing on the power of this new medium is through the integration of conversation and buzz (Web 2.0) with the traditional role of the web, providing information and marketing prose (Web 1.0).

But when you look at some of the most popular YouTube videos of 2007, some of these companies really blew their chance to take their 2.0 efforts to the next level by not integrating their creativity with conventional 1.0 promotion. And, it appears that the marketing departments of their competitors were also asleep at the wheel.

Here are a couple examples:

1. The Cadbury Gorilla Drummer video has received almost 1 million views. Go toTechnorati and you’ll see many blog entries praising the video. But look over to the right where the Google sponsored links are and you’ll not see a Cadbury ad.Searching for any online press releases related to the Gorilla playing drums again produces no listings. And on the Cadbury site, there’s no mention of the video. Search on “gorilla playing drums” in Google returns the YouTube page, but wouldn’t it be perhaps even more effective if Cadbury had an optimized page that ranked well under that search phrase … or if Cadbury had even bid on the phrase?
   “Buzz” grade: A
   Complete Marketing grade: F
2. The Ray-Ban Catch Sunglasses video has received over 3 million views. Unlike Cadbury, though the Ray-Ban folks carry through the Wow of the video. For instance, at the end of the video, the words “Never hide” are scrawled on a Honda Civic. Go to the Ray-Ban web site, and you”ll see that same slogan prominently displayed. And on their site, you”ll find a series of videos on their site supporting the same theme as the catch sunglasses video. Type in the phrase “catch sunglasses” into Google, though, and you won’t find any PPC ads from Ray-Ban or their dealers. On the other hand, typing in their slogan, “Never hide”, does at least return a link to their corporate site.
   “Buzz” grade: A
   Complete Marketing grade: C
3. The “Blendtec – Will it Blend?” series of videos on YouTube have been particularly popular, with the latest iPhone episode garnering over 3 ½ million views. And kudos to Blendtec – go to their home page, and you’ll see prominently displayed a button that takes the visitor to all of the “Will it Blend?” videos. In addition, they even sell CD’s of the video series on their site. And typing, “Will it blend” into Google produces a sponsored listing to the BlendTec site. They have even purchased the domain willitblend.com, which shows as the number one listing.Finally, they’ve incorporated their name into the title of the video series, so there’s no mistaken who they are. However, why have Blendtec dealers not taken advantage of this great exposure? There are no Blendtec dealers bidding on this keyword phrase.
   “Buzz” grade: A
   Complete Marketing grade: A (if we don’t count the missed opportunities for Blendtec dealers)