Testing is a huge part of paid search and marketing in general. Throughout most of my posts on this site, you’ll most likely end up seeing two common themes in terms of successful campaign management. They are analytics and testing; analytics to make sure you’re tracking everything properly and are looking at the data correctly and testing in order to find out what works best for your particular product, niche, etc.
When testing ads for a campaign, there are some people who say that you should make minor changes or even just one change to your ad and then there are those people who say that you should make each ad extremely different and unique. Each method has it’s advantages and disadvantages.
The advantage to making small or minor changes to your ad is that you’ll have a much better idea of what works and what doesn’t. For example, if you change your call to action from “Order Today and Save!” to Order Today & Get Free Shipping!”, then you’ll know which one appeals more to your audience. However, the disadvantage is that making small changes, especially very minor ones such as just testing punctuation or the replacement of one word, can take a long time to find the most effective ads. I’ve seen one “testing” phase for a creative be split up into several parts that would have taken 4-6 months or more to fully test. It just isn’t logical to test that long because 6 months from now, it might be that the original ad will actually work better than your best performing ad throughout all the tests.
Making two or more, very different ads to test has the advantage that you’ll be able to cast a wide net to find a new creative that works much better than previous ads. While there are times when a minor change could have a huge impact on your metrics, usually, if you have an ad that is very different from the other, one will work considerably better than the other. The drawback to this method is opposite of the advantage to the first: you won’t have as much insight as to why one creative is better than the other. Is it the free shipping that you added or is it because you changed the title? The other disadvantage is that if the creative is bad enough, it could have negative consequences on your overall traffic and conversion volumes to your site.
Which route you take should depend on the keywords you’re testing the ads on. While most of the time people will be testing on higher volume terms, which can bring results quicker, testing for mid to low volume terms can also help your overall account performance. If I’m testing for high volume terms, I’ll most likely use a mix between the two ways. I’ll use 3-5 ads, with at least one being the control, one being a slighlty modified ad and a 3rd being a very different ad. As I continue to test ads, I’ll test more and more sepcific parts of the ads to find which one works best overall. For lower volume terms, I usually will have 2-3 ads, one usually being the same or similar ad to a high performing ad for a similar, high volume ad group and then another one that is slightly different. Many times your low volume terms are also tail terms so they are already pretty specific and the ads should also be pretty specific to the keywords so even slight changes will make them fairly different.
Related posts
