A/B Testing

The word ‘test’ always conjures up some level of anxiety. Especially since every test is followed by results. With conclusive outcomes, indicating if what was tested is a success or a failure and if the same needs revisiting. With the advent of various new forms of online marketing, it is imperative that we test the same to ensure optimized results for all our marketing efforts. It only makes sense considering the amount of time we spend researching these new-age methods.


With the evolution of digital marketing and various analytics platforms, every decision was taken for a web page or newsletter, be it color, content or even a CTA (call to action) can be put to the test to gauge audience receptiveness.

The process of quantifying a visitor experience is never foolproof. But certain details like the Click-through Rates help determine the success of a particular feature on a webpage. Simply put a Click-through Rate is the number of times a visitor clicks on a feature or ad divided by the total no of times it was actually made available (impressed upon). CTR = Clicks / Impressions.

In order to determine which feature is eliciting what response, isolate the change and test it live with a potential audience. This is known as A/B Testing.abtesting

Common features put to the test through this process include the following –

Color – Even something as generic as the color on a page or even a CTA button needs testing. The contrast against the background color should be enticing enough for the visitor to want to click on it.

Text v/s Images – Often times an image works better as opposed to plain text and many times stand-alone text will do the trick. What works for your brand and design can only be determined through testing, to gauge conversion rates.


A tone in Text – There is a reason why the Thesaurus is so popular. People respond differently to various forms with the same implication. The use of appropriate words in the right context goes a long way in making a connection. Considering it is your voice in the virtual world, it is definitely worth testing.

Price related variables – The ultimate click is when the visitor hits the Buy Now button. The road to the visitor wanting to hit that button can be a long or short one depending on how the pricing and related information (discounts, reward points, shipping costs, delivery days) is placed. If the visitor is convinced of a good deal and a quality product then the transaction is almost certain. Almost since nothing is certain till it is tried and tested.

Layout / Background / Landing Page Creator The importance of each of these components cannot be stressed enough. Thus, the need to test what you may think may be the next best thing. Often times visitors are intrigued enough to visit a website and read through the Home Page. They start to explore it beyond the Home Page and the website feels like it has been patched together. The flow is amiss and the virtual experience is not interesting enough to stay another second. Result. Scoot. Test it and revisit it till you see the results for yourself.


The Basics of A/B Testing to Improve Landing Page Conversion


Simply put, proposed changes to a website, newsletter or any online form of communication and revenue generation are tested to gauge its acceptability and ability to increase the conversion rate.



The tests are conducted with a live audience on the webpage or through a newsletter campaign being run on categories of subscribers.



Time to introduce A & B. A is the current version of the web page or feature (aka the control) and B is the version with the variation which is being put to the test (aka ……….you guessed right, the variation). The two versions are simultaneously made available to visitors. Thus, each visitor or subscriber (in a case of a newsletter) is able to access either of the versions at one point in time. Their response to that particular version helps quantify the numbers of those in favor of each version. Thus indicating which is preferred. The current (A) or the proposed change (B).
Conclusion: After having conducted the test for a significant period of time or no of times indicative of a consistent winner, you either retain version A or update the code based on version B, based on the outcome.

Leave a Reply

Your email address will not be published.Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>