How to Create Display Network Experiments in AdWords

“Data do not give up their secrets easily. They must be tortured to confess.”

Creating and running PPC ads without experimentation is a waste of time and money. Fortunately, Google has an easy way to create experiments right in Adwords and Analytics! Today, we are going to focus on creating experiments in Adwords, specifically in the Display network, along with some best practices and tips to ensure your experiment is up to scientific standards.

How to Create a Google Display Experiment

1. First you must have an active campaign
2. On the right hand side, by the date selection, you will see drafts
3. Enter a name for your draft – I’d recommend naming it what you are testing like bid, landing pages, etc.
4. After you are done, a new draft will be created – make the appropriate adjustments that you want to test. You will also get to decide how you wish to split the traffic once you make the experiment live. Below are some options on what to test.
   1. Ad copy
   2. Image
   3. Destination URLs
   4. Bids
   5. Display network details
   6. Ad extensions
5. Google will keep tabs on how your experiment is doing in the All Experiments tab. This is where you’ll see the stats on what group is outperforming the other.

Want to create another experiment? Follow steps 1-5 again!

Best Practices for A/B Testing

A/B testing in its heart is scientific. You want to limit the number of variables you have in order to truly test what is working and, at the same time, your control group needs to be significant enough to properly asses your experiment.

First, you want to think about what you are trying to accomplish with your ad . Do you want to drive users to your website or do you want them to complete a certain goal? Then, you want to make sure you are tracking your goal to that specific metric, for example don’t change your ad copy with the goal of increasing your conversion rate then focus on time spent on site. Those are two completely different metrics to track.

For AdWords, test one variable at most, change the image ad, change the CTA, change the header, change whatever you like, just only do one change. Once you’ve decided what to test in your experiment, try it out for two weeks. It’s also recommended, once your test is complete, try the same test in different seasons.  Some users are more motivated to buy in the winter season versus the summer season. Then pick your winning campaign!

Remember, you can run multiple experiments in Adwords and aren’t just limited to just one at a time. Just be sure to name each experiment correctly to avoid confusion!