Pay-Per-Click As A Testing Aid
Most marketers on the Internet use PPC (pay-per-click) eventually within their marketing mix. Much of the time these businesses hope that the increase in sales will be at least sufficient to cover the advertising andeven bring in additional profit. Other times they look to recover those advertising costs over a longer period by using the PPC campaign as a means of obtaining leads that they use to develop a lasting relationship with new prospects. Still other times, the focus of some Internet marketers during a PPC campaign includes using the data that they collect for research and planning purposes.
It is this last approach to pay-per-click that I want to address in this article. Here are a few very useful tests that you can incorporate into your next PPC campaign. I am already assuming that you have performed superb keyword research.
* Using tracking data, reported by Google Analytics or your own software, identify the exact key phrases used by all of the visitors who come to your landing pages via PPC. Obviously, if you set up your campaign properly, you know which of the phrases that you bid on are bringing the visitors, however, unless you are using only exact match phrases, that does not alert you to the precise search terms entered by your traffic. For example, bidding on a term such as “buy green lamp” set up as a broad match, would get traffic from people who searched for phrases such as “buy a used green lamp in Columbus or Dover,” “buy green lamp,” “buy a green lamp in need of repairs,” “buy expensive tiffany green lamp” and many more. Any traffic you receive would be looking to buy some sort of green lamp. You may want to create pages for the key phrases that are applicable (and which seem to be giving you enough traffic to justify the relatively minimal effort). You can work on your SEO for those pages in order to get organic search engine traffic to those new pages. That can help justify your PPC expense for years to come.
* Test your headings (headlines) on your PPC landing pages. Set up two pages for the same ad group. The pages should be identical in every other way except for the heading. You might have a content management system or software that can alternate those. It’s also very easy to simply change the landing page to the different version within your ad after you have received a sufficient number of clicks to provide your with useful data—at least 100 clicks. Look at the data you gather concerning the results of the two versions according to whatever metric you are using (e.g. sales or leads). If there is a clear winner, keep it in the rotation and set up another test with a different alteration in the heading.
* Conduct the same format test as with headlines, but test a different variable. You may want to test listing benefits followed by features versus having the features list come before the benefits. Alternatively, you might want to test the impact of landing pages which have two different videos of product demonstrations.
Make sure that on each of the content related tests you are only changing one variable. If you alter both the image and the headline at the same time, for example, you will have difficulty determining which variable is responsible for any changes in the results or what the relative impact of each is compared to the other. (Actually, if you have some statistical sophistication, you can set up a test in which you change multiple variables at once across multiple versions of the landing page.)
The bottom line of this article is simply that you should be using your pay-per click campaigns to attempt more than selling more products; you should also be collecting useful data. PPC can be expensive, so stretch those dollars to accomplish as much as possible. Test, analyze and use the data that you wisely collect!
Popularity: 1% [?]

