Web Site Monetization,
|
Split Testing your AdWords Campaigns © Copyright 2006 Ian Feavearyear. All rights reserved. Split testing (also known as A-B or A/B testing) is a vital process to follow in order to continually improve your AdWords advertising campaigns. Without such ongoing testing, your AdWords campaigns will stagnate and will never be optimized to the fullest extent possible. Follow the procedure outlined below to bring a fresh lease of life to your campaigns. What is Split Testing?
Split testing is basically a method whereby you assess the effectiveness of two variables—in this case, your AdWords ads. This procedure can also be used to test the effectiveness of variations in landing pages, keyword sets and any other settings you can think of. In essence, split testing requires that you have 2 variations of only one variable, with everything else remaining constant. For the procedure described below, we will be split testing 2 different ads, with the same set of keywords and all other Adwords settings the same. You may choose to use 2 completely different ads, or you may simply rearrange a couple of words in the ads; it's amazing what a difference such minor tweaks can make. Basic Settings
Before beginning a split test for two ads, you need to ensure your campaign has the following settings:
Procedure
This really is the easy part! Now, all you need to do is to create two variations of you ad. For example, you may have read that it's often a good idea to put a price in your ads. In order to test this theory, create one ad with a price and another ad that is as identical as possible but without the price. Then, all you have to do is wait until you've had a reasonable number of impressions over a reasonable period of time. Of course, what's reasonable is really a matter of statistical analysis, but it's best to wait for at least a few hundred impressions, preferably a thousand or more. Sometimes, however, you just have to go with your gut as to when to stop testing and move on. Once you're satisfied that you've reached the desired number of impressions, then you examine the stats for the ads and scrap the one that was least effective. At the same time, you create another variation of the ads in order to test another theory, wording change, etc. and start another test. You then repeat this process ad infinitum . . . or until you've found an ad that you just can't improve on, of course! Which Ad Do I Scrap?
At first glance, the answer to this question seems obvious: "The one that didn't perform as well." However, things aren't necessarily that simple. For example, many people are obsessed with click-thru rates (CTRs) and automatically scrap the ad with the lowest CTR. But a lower CTR ad isn't always the one to scrap. Let me give you a "for instance:" Supposing your budget is $5 per day and you meet that budget every day. You have no additional money to spend. You run a split test on 2 ads and find the following:
Now, in this instance, all other things being equal, you are better off scrapping Ad1, even though it has a lower CTR because it has a lower CPC. When your budget is always being met, you are better off going with the cheapest ad and, in this instance, Ad2 will give you 25 click-thrus per day, whereas Ad1 will only give you 20, an increase of 25%! Of course, if you have the opposite problem and your budget is never met (i.e., you're willing to spend more than your advertising is costing), you are probably better off going with the higher CTR ad. Beyond the Basics
Of course, in practice things may not necessarily be quite so straightforward in determining which ad you should scrap as there are many other variables to consider. For example, ads with higher CTRs may eventually have a reduced CPC because they will end up with a higher quality score; and one ad may lead to more conversions than the other, and not necessarily the one with the lower CPC or higher CTR. The bottom line is, "which ad will be the most profitable?" As mentioned above, in order to fully optimize your campaigns you need to split test not just your ads but every other aspect of your campaign, such as: landing pages, maximum bid/position preference, delivery method (standard or accelerated), etc. As long as you keep split testing, your ad campaigns will continue to be more and more optimized and, hopefully, you will reap a signifcant monetary reward for the time and effort you've invested! So, keep testing!! Reprint Instructions: This article may be reprinted in opt-in email newsletters only. It must be reproduced unedited and in its entirety. The full copyright notice must be included. In addition, an active hyperlink to http://www.moneyties.net must appear either directly under the copyright notice or on the same line, as per this example: MoneyTies.net. For non-HTML newsletters, the full URL must appear. Any methods that state expressly or mislead readers into believing the author is anyone other than the author stated in the copyright notice above are expressly forbidden. Please email us notification of where the article will be used and a copy of your newsletter. Notifications should be sent to webmaster at adwords-for-profit dot com. Thank you!
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||
