Is your AdWords testing passing the test?

February 18, 2009 by jford
Filed under: Uncategorized

I am talking more and more to prospective clients who are already running or have run an AdWords campaign. This is obviously a positive sign in how many businesses are starting to grasp the effectiveness of paid search management and search engine marketing. The unfortunate part, as you may imagine, is many of these businesses are not feeling that the campaign is working for them. There can be many reasons for this, time, knowledge, resources, etc. But one of the issues that I see that many campaigns that I evaluate lack testing. This is a step that could help many campaigns, and may be easier to tackle than you think.

There can be several, several elements in a search marketing campaign to test. Let’s keep it simple. I would propose that you test 3 areas, keyword variations, ads, and landing pages.

Of the three areas, probably the easiest to test is your ad text. I read a great blog by Dan Theis on a technique for testing using control ads, and weaving in your test ad. I highly recommend you take a look at it. If you feel that as a small business owner that multiple control ads are not in your day to day operations, I understand. BUT, at the very least, it is critical that you implement multiple ads with your campaign. Why? Well, first off, if you do not test different ad variations, then you simply are dealing with the opportunity cost of unknown increased CTR and possible revenue.

Perhaps start with two ads when you begin your AdWords campaign. As we have counseled many businesses, be sure to turn off the Optimize function in your AdWords setting - you want the Rotate option. This will ensure that your test ads all get the same shot at being served.

Next, write two ads that are very different. I say this to make your job easier as you get warmed up to testing. Take one ad, focus it on your pricing. Take another ad, perhaps focus it on a differentiating your product. Utilize them in the same ad group, and be sure that you get over 200 impressions on both ads to get a good sample number. Then, look at the click through rate. From this, you will make evaluations on which ad performed the best.

Next, you may want to move into testing ads that are more similar to each other in nature, with similar texts. Subtle differences may be huge, like changing active language, or adding /keywords to display URLs, – they may not be. At least starting with two very different ads will give you a good feel as you wade in the world of testing. As you can tell, from there is a a limitless amount of testing you can apply. And you have other functions in life as well. So start small, but we in the paid search industry implore you, TEST, TEST, TEST.

Remember, in the portfolio of tactics in your overall marketing strategy, paid search should give you the most accountability. That is the second most important reason you do it (next to the primary reason that you are advertising based on the premise that someone is trying to find you, versus you trying to find the customer) Make sure you are testing your AdWords campaign to ensure that accountability, and make the changes needed to run a maximized campaign.

Share and Enjoy: