Wouldn’t be nice if there would be a way you can find out which webpage design your visitors like the most if we provide design alternatives to them?
Isn’t it amazing if we can directly find out from our visitors which look-&-feel they prefer and we can eliminate all guess work?
Absolutely, now you can test these variations by performing A/B testing.

What is A/B testing?
It is exactly what it sounds like. You have two versions of one element, A and B. To determine which element is better, you test them simultaneously and split your traffic between these two versions. In the end, you measure which version got more click; select the version which performs the best and use that version in your webpage.

How does A/B testing works?
It is very simple. First, you need to decide what you want to test. Then, you need to create two or may be three versions of your webpage or element, depending on your requirements. It can be any element or module from your webpage, which is uniquely created and stand alone. It can be just simple call-to-action ‘submit’ button by changing color red vs. green. Next step is to select tool for this job. Set up your conversation goal. Typically, you will get a JavaScript code, which you would implement in your webpage or element. Test these alternatives with your visitors concurrently. Website optimizer will then show these alternatives to your visitors randomly while monitoring which combinations lead to the highest conversion rates. The tool records the variations that were shown to the visitor. Finally, instead of all speculations now you have proven findings statistics in front of you.

Even though, A/B testing concept is very simple to understand, one important thing you need to keep in a mind.
Always test these variations simultaneously. If you are testing version A for one week and version B for another, than you are doing it wrong. It is possible that when you tested worst version B, but it’s just happen to have better sales that week. Obviously, it is hard to control external factors such as search traffic, seasonality, etc. so you cannot be 100% confident of the result. Heard this phrase, “compare Apples to Apples”? Always split the traffic between two versions. Let the tool randomly select the version.

What are the benefits?
• It’s easy.
• It increases website conversation rates. Users’ loyalty and conversion rates are more important than number of click-thru rates.
• It eliminates guess work because you can find outcome directly from your visitor
• It’s an easy way to find what change drives more conversions.
• It increases visitor satisfaction overall
• It might be a cheapest way of doing testing since you will be using same resources and tools.
• It helps increase sales. This is directly tied in to your company ROI and more importantly increased revenue of your company

Many well-known companies such as Amazon, Google, Microsoft and BBC use the designed experiment approach to making marketing decisions. There are several free tool such as Google Website Optimizer and paid tools such as Optimizely, Visual website optimizer, etc. available out there in the market to perform these tests.

On the whole, you can A/B test everything. But, resist the urge to nit-pick every tiny portion of your site and change something every time by performing this test. One player is never more accurate and reliable than other players, so always look for other significant ways to improve your website goals.