PPC Testing Part 4: Google’s AdWords Campaign Experiments Tool
For those of you just getting into paid search, count yourselves lucky. You’re lucky not only to have landed in one of the hottest fields in marketing, but also lucky not to have gone though some of the major workflow frustrations that have now been addressed and fixed in the first ten years of paid […]
For those of you just getting into paid search, count yourselves lucky. You’re lucky not only to have landed in one of the hottest fields in marketing, but also lucky not to have gone though some of the major workflow frustrations that have now been addressed and fixed in the first ten years of paid search. One of the major advances that is now a luxury for search engine marketers is the wide availability of desktop editors such as Google AdWords editor, as well as Microsoft and Yahoo versions. “In my day…” (said in dusty old grandpa voice), we had to use the engine bulksheet uploaders which took almost twenty times to validate. You uploaded, downloaded your sheet back with all of the errors fixed them and repeater the process over and over until it was successful. I don’t have any hard data but I think Google AdWords Editor alone would have shaved nearly four or five hours off of my week back then.
Why do I bring this up? Because “you kids today” now have a new experiment tools from Google called AdWords Campaign Experiments (ACE) that now makes testing—one of the most crucial parts of successful PPC management—a bit easier to implement. It’s not a world changer (yet), as it’s still in beta and needs some work to become a fully fledged testing super-machine. However, it’s definitely a giant leap forward for those of us that have had to do the same thing manually for years.
So, first, let’s talk about how testing has been done up until now. Basically, it’s hard to run a side-by-side test on keyword bids, match type and so forth, because your search account is an ecosystem where everything is interrelated. For example, it’s very hard to get accurate results on keyword changes because if you put the same keyword in different ad groups, Google’s system decides which one will show. Chances are, if you are testing the same keyword with a $1 bid or a $2 bid, Google will start running the instance of that keyword with the highest bid. So, side-by-side testing is just very difficult.
What’s the alternative? Well, in month A you could run the keyword with a $1 bid and then month B you could run it with a $2 bid and then compare the results. But the results will inherently be inaccurate because the tests weren’t running at the same time. Seasonality, changes in the economy, and zillions of other variables could be the cause for the positive or negative data when you don’t run side by side ads in testing campaigns.
Another major problem that’s caused by testing is that you may disrupt normal business flow during your testing periods. Let’s say that your account is performing rather well with decent sales and a fairly low cost-per-click. You want to test how some bid changes might affect your keyword set, so you make the changes and watch the results. On occasion, your test hypothesis might be completely off base, and suddenly your sales begin to fall. Do you let the test run its course or do you pull the plug and revert back to your campaign settings that were in use before the changes?
The AdWords Campaign Experiments tool attempts to solve all of these problems by allowing you to create separate versions of your campaigns which you then run side by side at whatever percentage of volume you wish to run. Basically, you can choose to send as little as 10% of your traffic to the experimental version of a campaign so that you can gather some good intelligence without messing with your current success. And the good news is that once you’re done with the test, you can click one button and apply all of the experimental changes to the existing, live campaign.
How do you set it up? Here is the walk through from the Google AdWords Help Center that can walk you step by step through the process:
- Step 1: Setting your experiment parameters
- Step 2: Defining your experimental changes
- Step 3: Monitoring your experiment
- Step 4: Applying or deleting your experimental changes
Good luck with your experiments and remember to always leverage the technology in your favor.
PPC Academy is a comprehensive, one-year search advertising course from beginning to end. Starting with the basics, PPC Academy progressively explores all of the varied facets of paid search, and the tactics needed to succeed and become an advanced paid search marketer.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.