Lorna Li | Green Marketing 2.0

Congratulations - you’ve set up your Google Adwords account and are ready to launch your first PPC campaigns!

Somewhere very quickly down the road you are going to want to test your ad copy and perform landing page tests in order to continually identify winning creative and improve your conversion rate. Here is an overview of how you might set up and perform basic A/B split tests in Google Adwords. These insights can be applied to other paid search programs.

A/B testing is a method of advertising testing by which a baseline control sample is compared to a variety of single-variable test samples. While a classic direct mail tactic, A/B split testing has been recently adopted within the interactive space to test PPC ads, banner ads, emails and landing pages.

To perform A/B testing correctly, you will need to deploy a control - usually the strongest, most consistently performing version of your ad - alongside at least one or more variations of this ad. Bear in mind, in order to correctly identify which variable is responsible for changes in response rate or desired outcome, the ads must essentially be identical except for one variable difference.

Confused?

If you are simply testing ad creative, you have 3 lines to work with, or 3 variables. Some might include the destination url as the 4th variable, but for simplicity’s sake, I will focus on 3.

1. Title
2. Description 1
3. Description 2

You have limited space, so understand that each line must carry its own weight, yet at the same time, reinforce the whole message you are trying to convey through your ad.

PPC Split Testing – Step by Step Instructions on Testing Ad Copy

1. Make sure ‘ad optimization’ is “off” the campaign settings in order to ensure all ads are served equally.

2. Determine how many ads you can successfully test based on your daily volume of impressions. If you have a large volume of impressions, such as 10,000 daily impressions, you can serve more ads than if you had 100 impressions a day. However more ads means more data to crunch. Also, bear in mind that more ads you have, the less proportion of time each one is served, and the longer it will take to gather statistically significant data.

3. Create 2-4 titles (T below), description line 1s (D1 below), description line 2s (D2 below), and then mix and match them. Remember, these go up exponentially, testing 3 lines is 27 ads (i.e. #t x #d1 x #d2).

Graphically, your ads will plot out to look something like this:

Ad# T D1 D2
1. A A A
2. A A B
3. A B A
4. A B B
5. B A A
6. B A B
7. B B A
8. B B B

As you can see, with just 2 variations of each line, we now have 8 ads, thus the more copy you test, the more your data analysis work will expand.

4. Determine your time frame for testing. Give yourself at least a few days, though 2-3 weeks should yield solid data. Also factor in any holidays, and allow more time to account for any deviations in user behavior during this period.

5. Once you’ve reached your time frame for testing, it’s time to analyze the results. Export your data into an excel spreadsheet. Google’s robust reporting tools will allow you to export practically every data point imaginable in a myriad of ways. In this case, you will want to download a report on Ad Performance, which will give you stats on the performance of each ad by Impressions, Clicks, CTR, Cost, CPC, and Avg Position.

However, in order to get a read on conversions, you will need to enable Google Conversion Tracking. With Google Conversion Tracking enabled, you will have additional columns of conversion related data in your reports that will help you better analyze your PPC split test results.

Technorati Tags: Pay-Per-Click, PPC, A/B testing, split testing PPC, PPC ad testing, SEM, search marketing, Google Adwords, lorna li

This entry was posted on Thursday, May 31st, 2007 at 7:23 am and is filed under PPC, Search Engine Marketing. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.