Get Better Results from your Ads using A/B Testing. Basics of A/B Testing Process

The traditional advertising methods lack real measuring tools, and advertisers are always in a dilemma whenever they launch a new campaign.  Advertisers are not sure about the real return on each of their investment or ads spending whenever stakeholders come up with questions.  They were always dodged by the question – “how do you know which campaign or Ads were successful? Many a times, they make presentation before sample population, use small electronic devices, and seek feedback form to understand the real effect of their spending to convince clients about their efforts.  Sadly, they did not have any right tool to gauge the performance, and subsequently they never got any real metrics to find out their Return on Investment.


With the introduction of online marketing, Advertises can now test their Ad Copies to measure performance, and pick up the winner using more convincing tools. The option to measure performance through online A/B Testing sometimes called “Split Testing” is to identify which ads work, and which ones are the losers.

The idea behind Split Testing is launching 2/3 different ads over the same time period to find out which one receives maximum exposure in terms of clicks, sign ups, and conversions. The winning Ad is continuously tested further against new Ads or Ideas till the new winner is emerged each time. In reality, it is a continuous process and a tedious job for the ad managers to get bang for the buck.

Before split testing the ads, we need to define the elements for testing – webpage, Ad Copies or an App screen. Once you realize this, create a second version of the page. This variation can be as simple as a single head line, or button, or a complete re-design of the page. In general, almost anything on your website that affects visitor behavior can be A/B tested as follows;

  • Headlines
  • Sub headlines
  • Paragraph Text
  • Testimonials
  • Call to Action text
  • Call to Action Button
  • Links
  • Images
  • Content near the fold/above the fold
  • Social evidence
  • Media references
  • Awards and badges

Using A/B Testing, final decisions are evaluated based on the elements placed on your webpage, Ad Copies. In order to derive the hypothesis, we must get them from the feedback of the users by adding:

  • Live Chats
  • Subscribe Popups
  • User Forum

Alternatively, you could also test the Target Audience (Age, Gender, or Interest Category). Always ensure to keep 3 – 6 Ads in each campaign running simultaneously so that you can test one against the other. If you have more number of ads in your campaign, you will find that some ads surface as top performers, while others receive low delivery in assessment. During such cases, pause those underperforming ads, and create new ones similar to the best one and again adjust them marginally.

There is no disbelief that Split Testing your Ads in one of the effective ways to improve your Campaigns goals, and performance. With testing, you will be able to figure better who your audiences are and what they need the most, and continuously improve increasing CTR, clicks and Conversions   While setting up a new campaign, we need to structure campaigns by targeting exact audience against set of defined goals to make the testing more effective. It’s easy to categorize which of the ads are most successful when your targeted ads are not identical.

How about you? How many times do you optimize and split test your Ads?