Jason Puckett

July 10, 2016

How to Build Your Experiment Plan

Planning is a good thing. Clients, customers, or circumstances can occasionally put you in situations that are unplanned, but 90% of marketing growth can be tactical and thought out. Ad testing and conversion rate optimization is no different. Here are some tactics that you can use to approach your experiment plan for Q3.

Create a Graphic Representation - Experiment Tree

We have gone ahead and created a really solid one for you. As you can see below, this experiment plan resembles the roots of a tree. As the plan becomes more and more advanced over time, the number of “roots” splitting out represent individual tests. As you can see from the green dots, it is possible to run a test in multiple areas of your account simultaneously in a well calculated sequence.

Experiment Tree

Brand and Non Brand Segmentation: This should be your highest level of segmentation and optimization. The people searching for your brand terms know who you are and will thus behave differently than those who are searching for Non-brand terms. As a result, we should build these tests individually in order to study behavior differences.

Hint: If you’re creating branded experiments, don’t forget your AdWords fundamentals. Using your brand name at least twice in your ad copy (we like Headline + Display URL).

Evergreen Ads by “Ad Group Set”: Every AdWords account is different, therefore you’re going to have to start using your knowledge-base of your own account. In some AdWords accounts, an “Ad Group Set” may just be a campaign… in others, it may be a “Product Vertical” or “Region”.

Look at your account structure to determine the structure of your testing plan. If you need help doing this, always feel free to check out our consulting services: www.AdBasis.com. The question you should ask yourself, “Which scenarios do I want to be optimal for?”. And go from there.

Promotional Ads by “Ad Group Set”: Once you’ve defined the “sets” you’d like to optimize for you will need to switch the ad copy being used based on your marketing calendar.

Marketing Calendar

Maybe one of your channel partners is releasing new features and you’d like to create a promo to surround that. Let’s say your “Non-brand Product X” ad groups are the ones to reflect this change. What you’ll probably want to do is keep one of your “Evergreen” ads running, and upload 1-3 new variations surrounding this promo.

Let these promo ads run for a fraction of your promo, then optimize. Once promo has ended, pause the ads and restart your “Evergreens”.

Boom. You’re the best optimizer on the planet.

Interested in obtaining this type of data for your own ads? Either contact us or sign up for AdBasis today!

Have a question for us? We'd love to hear from you!

comments powered by Disqus


Popular Articles

Understanding Conversion Layers and Executing Ad Experiments

August, 14th 2014 - Optimization

3 Steps for Building a Paid Acquisition A/B Testing Plan

October, 15th 2015 - Optimization

All Things Tested - 3Q Digital's Sean McEntee

August, 31st 2015 - "All Things Tested"