How to use testing to optimize your PPC ads

Why test ads?

If you’re serious about PPC, then you have to get serious about ad testing.  Sure, you could develop an ad based on your gut feel, and you could get lucky and find yourself with a top-performing ad, but most of us aren’t that lucky.  A better—and more pragmatic—way of going about it is to systematically test ad iterations in order to develop the best possible ad for your business.

A/B Testing Concepts

 The most basic form of ad testing is called “A/B testing”, which means creating ad version A, and ad version B, and then comparing them.  Whichever one gives you the best results becomes the winner.

Here are some key concepts to keep in mind when testing ads:

• You can only test one element of your ad at a time.  If you want to test your ad headline, go for it… but don’t test your headline and another part of your ad at the same time.  If you do that, you won’t know which element is contributing to the success of the winning ad, and you won’t know what path to follow in order to do more tests and further refine your ad.
• Test each ad version for the same number of impressions. You have to compare apples with apples, so make sure that each ad version gets the same number of impressions.
• Make sure that you’re evaluating an ad’s success using a relevant metric.  Don’t just go by click-through-rate.  Use Google analytics to look at your bounce rate, how long people are staying on your site, how many pages they’re visiting, etc, to make sure that you’re evaluating your ad versions using the most relevant metric to your business.

Testing your first ad

 I’m going to use Google Adwords text ad to illustrate basic A/B testing.  Here’s what a Google text ad looks like:

Headline
Description which is always
on two lines.
www.displayurl.com

Now here are the simple steps you need to follow in order for you to develop an optimized ad through testing:

1. Choose what you feel a good add should look like.  Write a headline, description, and choose a display URL.  This is the one and only time you’re going to go by your gut feel.
2. Change one element of the ad.  By element, I mean the headline, the description, or the display URL.  Remember, you can only change one element at the same, or your testing results will be unreadable.
3. Let’s say we change the headline to start.  You’re going to now create at least 2—and preferably 3 or 4—different headlines, always keeping the rest of the ad the same.

Here’s an example:

 Get a new car today.
Reliable used cars for the
best prices in town.
www.newusedcards.com

Here’s what my versions could look like:

Notice that the only thing being changed here is the headline.

4. Now run these ads at the same time, taking care to make sure that all three ads have the same number of impressions before you stop your test.
5. Evaluate which headline gave you the best results using the most relevant metric to your business.  If you’re not sure which metric is best for you, start with the click-through-rate (highest click-through-rate wins).

You’ve just optimized your ad headline.  All you have to do now is repeat this process for the description, and the display URL, and you’ll have an optimized ad.