Senior Production Specialist, Lauren Rosner, discusses how you can get started with A/B ad testing and gives real life examples of "wins" in ad copy from Hanapin clients.
4. #thinkppc
Getting Started
•Define your testing groups
–Account-wide
–Ad Groups
–Campaigns
•Verify your groups make sense
–Think about time and audience
•Select your metric
–CTR, IUC, Revenue per Impression
5. #thinkppc
What question do you want to answer?
Client: Drug and Alcohol Rehabilitation Center
Current copy: Do You Need To Go To Rehab? Get The Help You Need
Now. Call Today
Question I wanted answered: What if we talk to a different audience?
What happens to click through rate?
6. #thinkppc
What question do you want to answer?
Client: Drug and Alcohol Rehabilitation Center
Current copy: Do You Need To Go To Rehab? Get The Help You Need
Now. Call Today
Question I wanted answered: What if we talk to a different audience?
What happens to click through rate?
Test copy: Is Someone You Know In Need Of Rehab? Get Them Help.
Call Now.
7. #thinkppc
What question do you want to answer?
Results: Control 1.84% CTR
Test: 2.23% CTR
99% Statistically Significant
Client: Drug and Alcohol Rehabilitation Center
Current copy: Do You Need To Go To Rehab? Get The Help You Need
Now. Call Today
Question I wanted answered: What if we talk to a different audience?
What happens to click through rate?
Test copy: Is Someone You Know In Need Of Rehab? Get Them Help.
Call Now.
8. #thinkppc
Moving through an a/b testing cycle
Description Lines A vs Description
Lines B
Winning Description
Lines with Headline A
vs Winning Description
Lines with Headline B
Title Case vs
Sentence Case
9. #thinkppc
Looking at pre and post testing metrics
Client: Travel agency
Test: Experience-based messaging versus Credibility-based
messaging
CTR Prior to test: 1.28%
CTR at conclusion of test: 1.56%
22% increase in CTR
Client: Supplement E-comm Client
Test: Sentence Case vs. Title Case
CTR Prior to test: 2.84%
CTR at conclusion of test: 3.38%
19% increase in CTR
The other day I was reading Dr. Seuss “The Thinks You Can Think” to my 9-month old son. And my mind wandered to PPC, as it often does at the most random times. I thought to myself “If you can think it, you can test it”. The possibilities become endless. Which is not useful when trying to make actionable decisions for your account.
So where do we start? If you’ve never done a/b testing for ad copy in your account, it can be a bit overwhelming to figure out where to begin. There are so many different ways to segment your testing groups. Are you testing account-wide? A set of ad groups? Your top-performing campaigns? Things to keep in mind: make sure your groups make sense. You don’t want to put a B2B and a B2C campaign in the same testing group. More than likely their messaging would be different. And you want to make sure you are comfortable with the size and scope of testing. If you are testing within low traffic campaigns, you will need to let the test run longer than if you are running the test in a high traffic campaign. Once you have your testing group selected, you will need to find your control. Find the best performing ad currently running. Judge this by however you are going to judge your a/b test results: IUC, CTR, Rev per Impression. If you are just starting out, I recommend starting with CTR. Once you garner some data as you go through your testing, you can make a more informed decision later.
The process of getting control ads selected and other ad copy cleaned up can be tedious. So give yourself time if you have a plethora of old ads running. So once you’ve got things cleaned up, depending on the scope of your account that could have been some heavy duty work, give yourself a high five. Now for the fun part. Look at your testing group. Look at your account. Think about your audience. Then ask yourself, what question do I want to answer? One of our clients, a drug and alcohol rehabilitation center, has continually used ad copy that speaks directly to someone in need of rehab. But thanks to back-end data from the client, I knew that the individuals calling in or chatting with admissions were often times NOT the individual needing rehab services, but rather a family member, friend, or spouse. So what happens when I tweak the ad copy to speak to those individuals instead?
So I switched up the ad copy to speak to the other individuals. This is more of a large, sweeping test, focusing on overall messaging. So what happened?
After running for 30+ days, I found that someone you know does in fact perform better! So this is fantastic. I have information now about my audience and can move forward by tweaking the copy further. And when you go move to the next round of testing, tweak the copy just slightly rather than make huge sweeping message changes. So the a/b testing cycle looks a bit like this
So you will continue to cycle through A/B testing, always starting with a big sweeping change, then moving to a smaller change, then smaller still. Once you move through the third round, cycle back to the top. The continual cycle allows you to make adjustments with seasonality, messaging, branding, etc. and constantly hone in on minute details. As the ultimate goal is to move the needle in your account through constant testing.
So another important thing to keep in mind is how testing affects your performance in whole. In these two examples, you can see improvement in CTR at different points in the a/b testing cycle. So whether you are at the top of the cycle testing big changes or at the bottom of the cycle testing small changes, testing can help move the needle in your account and you can start making actionable choices when it comes to ad copy.
Thank you so much for joining this morning! I hope I have given you a few takeaways in order to get started testing in your accounts!