Creating Creative Creative: Best Practices for Creative Testing and Optimization

By Darren Kuhn
Appeared In DM News

With Yahoo now joining Google in using click-through-rate (CTR) to determine paid search ranking, creating compelling ad copy is more important than ever. The process of developing effective creative is an iterative one, in that what works today may not work next month or next quarter. Because of this, we always recommend continuous creative testing to improve and optimize performance.

When implementing a creative test, we typically focus on Google first as it currently has the most robust creative rotation tools. Additionally, especially now that Yahoo's character specs match Google's, the results from Google can be rolled out to other engines.

For current campaigns, when selecting the Ad Group(s) to test within Google, start with Ad Groups that contain a strong return on investment (ROI) and low click through rate. Next, find out the industry average CTR from the engines to determine where your CTR ranks against the competition. Before the test begins, make sure keywords within the ad groups are related to one another and the ads link to the same URL/landing page. At a minimum, each creative should ideally get at least 50,000 impressions or 500 clicks during the test period.

Duration is an important part of the testing methodology. The duration of the test depends on the amount of time it takes each creative to attain the minimum number of clicks or impressions mentioned above. For high volume sites, the test period should run for at least 2 weeks to account for sales cycles and weekly and daily trends.

To begin the actual test phase, create 4-6 new ads with different creative text and different message types (e.g. product guarantees, price points, free shipping, etc). Place unique tracking on each creative so both conversion and CTR can be tracked. Be sure to disable "Google optimizer" so Google will no longer give more impressions to the creative that has the higher CTR. Continue the test for at least 2 weeks or until each creative receives the minimum number of clicks or impressions required.

When analyzing the results of a creative test, always keep in mind that ad success is not determined solely by CTR, but a combination of CTR and conversion rate. Whatever your conversion action may be, the best creative results in the highest number of conversions, not the highest number of clicks. Once you have uncovered your optimal creative (from this particular test), begin running only the winning creative on Google and extend the findings to other engines. If there are no clear winners, create new iterations on the tested themes or develop new themes altogether and begin another round of testing with the new creative.

In the following test for a leading company in the Home and Garden category, 4 ads were created as the test group (against the original/control copy) to determine the most effective selling message. The measurement of success was the highest number of orders, or the optimal combination of CTR and conversion rate.

It is important to note that none of the ads were significantly different. Each were variations of similar themes by changing the order of terms, inserting the searched keyword, etc. Look at the results...

Reviewing the results, it is clear that even though AD1 did not have the best CTR, it is considered was the best performing creative because its conversion was significantly higher than the other ads which offset the slightly lower CTR. While AD3 has the highest CTR, it did not bring in any sales during the test period. This could suggest a relevancy issue, which can and did become a future test.

Another important variable to test is the effectiveness of the Display URL, as in the following example:

Display URL1(AD1): MyDomainName.com
Display URL2(AD2): MyDomainName.com/ProductCategory

As you can see from the above results, using a sub-directory name in the URL of ad2 outperformed AD1 in both CTR and conversion rate. Users were more drawn to the additional use of the specific term in the URL.

As with any testing methodology, there is no "one size fits all" approach. Use these principles directionally and take what you know about your category and your business to prioritize action and interpret results. Remember that, above all, the most important thing to do is to continually test. Once you've pinpointed areas for improvement from a creative test - rinse, lather, and repeat.