7 PPC Ad Testing Best Practices to Improve Your Campaigns

AB-Testing1Always Be Testing. The number one rule in ad copy testing according to (probably) everyone in PPC. If you’re not testing, you’re missing out on possible progress that can be made. Right? Right.

But testing just for the sake of testing and without specific strategies can cause you to waste time and effort creating and activating tests that might not yield any useful results. I see this pretty often with accounts we review. Ad testing is either inconsistent, too granular to be effective, or completely nonsensical and ineffective. So today I’d like to share some strategies I’ve used over the years that have worked to improve my ad copy testing strategies and make me much more effective.

Always Have a Purpose

Always be testing is great, but if you don’t have a purpose, you’re just spinning your wheels. Not all ad testing has the same desired outcome. Start with the end in mind. What are your goals? Increase CTR? CVR? Limit low quality clicks, effectively decreasing CTR?

Now, strategize on how you’re going to accomplish those goals. With CVR, you might want to test different offers or calls to action to increase sales. To increasing CTR you might write a couple attractive benefit descriptions to draw some eyes. To improve click quality, you might write some qualifying copy based on the features of what you’re selling and try to weed out anyone not looking for or not able to use those features.

Your goals could also be to determine which type of messaging works best for different products, locations, etc. My main point here is to not test blindly. Make sure you know what you’re trying to accomplish and set up your test to give you the answers.

Only Test One Thing At a Time

Have you ever gone into an ad group to find 5 variations of ad copy that had nothing in common, only to then gone into another ad group in the same campaign and find 3 ad variations with no similarity to the other 5? Headlines, ad text, even display URLs were different. I consider this purposeless testing. You might be able to declare a winner, but you can’t know why it won so building upon that test might be difficult.

There are four parts to ad copy that you can actively influence testing on: Headline, Description Line 1. Description Line 2, and Display URL. For each of your tests, choose one of these that you want to test and leave all the others the same. By keeping all other factors the same, you’ll know at the end of your test that whichever messaging won was due only to itself and not it in combination with something else in the ad.

Use Aggregate Testing Where Possible

Aggregate ad testing has become one of my favorite things recently. With aggregate testing, you choose your level of testing (account level, campaign level, a group of campaigns, ad groups, etc.) and run the same ad copy tests across those areas of the account. See the image below for a campaign-level example:

Screen Shot 2014-06-24 at 9.35.03 AM

Each ad group still retains it’s own keyword focused headline and destination URL, but the description lines are the same across all ad groups in the campaign. This allows for quicker testing of features, benefits, offers, calls to action, etc., for the entire campaign and allows your to make changes more frequently.

Test Different Types of Messaging

I’ve already hit on this a little in the previous tips. By different types of messaging, I mean features, benefits, offers, calls to action, value proposition, etc. Each of these can be very valuable to your customers, but if you don’t test, you might not know which ones. For example, in the B2B world, features messaging might outweigh benefits messaging for project management software because companies already know they need it to help save time, but they’re currently comparing features to determine the best software for their needs. Know your audience, know what’s important to them, and if you’re not sure, test it!

Test in a Cyclical Manner

As PPC managers we can sometimes get stuck in a rut with ad testing. We start big, then start distilling down into more and more specific tests to the point where we’re testing between two very small distinctions. To help me stay away from this pattern, I like to implement a cyclical testing strategy. It’s nothing fancy. Big, medium, small. That’s pretty much it.

Start big. Your initial test should be between the different types of messaging and compare large messaging changes. After your first test is finished, make a modification on your previous test. Use the same feature, benefit, etc., but change how it’s worded or displayed. For the last and smallest test, make a smaller variation on your previous winner, if possible. Give it different punctuation or change up the wording while keeping the same words and meaning. Here, you’re only testing the optimal way to say what you want to say.

Screen Shot 2014-06-24 at 9.44.01 AM

Once you’ve determined your smallest winner, start big again. Find a new feature, benefit, call to action, to test and start the process all over again! This way you’ll be able to continually test large messaging changes then find the optimal presentation for each of those pieces.

Keep Tracking of Winners/Losers Over Time

Keeping track of your ad tests over time can do a couple of things. First, you’ll be able to quickly look back and see which types of messaging you’ve used before so as not to retest them again too soon. Second, you’ll have a bank of ideas to continuously build off of or adjust for future tests. Whether you decide to keep track of these by using labels in AdWords, keeping an Excel sheet, or using your project management software is up to you, but I’ve found it very beneficial over the years.

Don’t Be Afraid to Reuse Copy

One of the flaws of ad copy testing is that it’s not an absolute. Once an ad is determined a winner or loser, that doesn’t mean it will always be a winner or loser. Think of it like fashion. It was popular once, it will most likely be popular again in some way, shape, or form.

If you get stuck in a rut, don’t be afraid to look back at your ad copy document and try and rotate in a previous loser to keep things moving. You’re not looking for the ad copy that you’ll run for the rest of the life of the account, you’re looking for what works best at this moment. I’ve had many previous losers come back and be fantastic performers for a month or two.