Before and After: An A/B Testing Case Study
February 3, 2013
Businesses can only accrue sales when they know the right items to put on shelf the kind of consumers that would be interested in them. It is therefore imperative for entrepreneurs to vehemently study their niche before they make investments. This ensures that the time, money, and effort they wagered would produce ideal results.
Needless to say, the ingredients to success include affable products that can click to the target buyers.
The same holds true even in online marketing. Internet-based merchants actually have two things to sell: the goods and their website. It is useless to have the best merchandises in the world if they fail to create a virtual store that can effectively ensnare buyers as superfluously as possible.
This is where A/B testing comes into play.
A/B Testing, also dubbed as split testing, is a method integrated in the marketing strategy where consumers inadvertently participate in a “survey.” Two versions of a webpage will be crafted to contain varying buttons, layout, content, design, colors, etc. These variants will be showcased to website audiences in a monitored environment – meaning, web masters will be able to trace the navigation behaviour of the visitors.
At the end of the experiment, businessmen would need to make an analysis on which variant worked best in terms of traffic and conversion rates. Several criteria for performance measurement will be laid out, such as click out rates, purchases, number of views, etc. These will all be the basis for the evaluation.
Truth is, split testing is not something unknown. In conventional procedures, this can be done in the guise of questionnaires or surveys among customers. The only difference of this scheme is that website variants are being voted for, and the contribution of the respondents is covertly solicited.
A/B Testing is needed to acquire influx of customers and profits. Given that entrepreneur will have a clearer idea on what their niche demands, they can present the products that can effectively motivate visitors into making a purchase.
The test conducted by Sim Only was a living epitome of how split testing works.
Sim Only: Before the A/B Test
Sim Only operates through its website, Sim Only. Prior to the split test, the site is endowed with a modest 18,000 unique hits per month. The page views reach at least 54,000 monthly.
Conversion rate, on the other hand was at 13.35% with a plus/minus 1.67 margin of error. This was taken from a 1,596 default generation with a 213 count.
Sim Only was then curious on how these numbers can possibly be raised. As a result, the company partnered with Maxymiser, a resounding name in split testing.
Note that the company makes profits by selling both SIM cards and phone units such as Blackberry and iPhone. The goal of Sim Only was centered on increasing their click out rate towards the website of the external partner domains within six weeks. With the help of Maximser, the company opted to test its homepage.
The test was launched after the developers whipped out engaging designs for the variants.
Sim Only was aimed at studying how changes on the designs of the homepage can influence their click out rate. Each of the variants’ performance will be measured will be the number of click outs made towards the site of every external partner. The secondary metric is set at the number of homepage clicks through rates.
Gadgets banner is tweaked for this particular test. One version contains a vertical layout of the products on the right hand side of the page according to popularity from top to bottom. The other is designed to have a horizontal presentation of the products across the page according to popularity from right to left.
Both variants come with checkboxes that allowed users to select two or more products at the same time. The images of phone were used as icons.
Sim Only: After the A/B Testing
The findings from the test were summarized in this table:
It is clear from this table that the customers have a sharp inkling towards the vertical layout over the horizontal design. This means that when the company implements the variant design that garnered the highest conversion rates and confidence index, it can attain greater conversion rates. Vertical phone can potentially give Sim Only its most sought after results.
If the data is to be analyzed according to the number of generation, counts, conversion rates, lowest rate error and confidence index (along with uplift percentage), it can be inferred that vertical phone design increased the company’ click out rates by 20%.
This shows how A/B Testing can give businesses a competitive edge that may not be obtained elsewhere.
Author bio: Ruben Corbo is a freelance writer and writes for a number of online marketing websites including those that help online businesses improve A/B testing techniques to increase sales conversions (which you can learn a deal about in Maxymiser). When Ruben is not writing, he’s producing or composing music for short films or other visual arts.