What is A/B Testing?
So what is A/B Testing and why is it so effective you might ask? A/B testing is a method of pinning two versions of an app or website against each other and seeing which one out performs the other. The two different modifications of the page are given out at random to visitors and statistical analysis is used to decide which of them efficiently leads to more conversion goals. This technique is successful because it allows companies to compare experiences side by side in real time and collect data to actually prove which change(s) to a site page is more effective.
Marketers should focus on A/B testing specifically, because it allows them to make the most out of existing traffic. For example, the cost of 5-10 Google Adwords clicks is roughly the same price as a Small Business Plan of Visual Website Optimizer at $49 (VWO). Or in other words, the cost of securing paid traffic can be huge, but the cost of increasing conversions is minimal. Even fairly small changes on a landing page have resulted in huge bumps in sales and revenue when looking at companies who have implemented the testing program.
There are a handful of different ways to effectively run the A/B testing process but the most practical one I came across was the following scientific process:
1. Study your Website Data: Use tools like Google Analytics to find problem areas in the conversion funnel.
2. Observe User Behavior: Find what is stopping users from converting.
3. Construct a Hypothesis: Build an If, then statement aimed at increasing conversions.
4. Test Hypothesis: Test your original page to your variation page per your hypothesis.
5. Analyze Test Data and Draw Conclusions: See which variation produced highest conversions.
6. Report Findings: Report findings to all concerned.
There are also a few practices worth noting to avoid trouble with Search Engines. There is no Cloaking, or showing Search Engines different content that a regular visitor would see. Use the rel=”canonical” attribute to help prevent Googlebot from getting confused by multiple versions of the same site. Also remember to use 302 redirects and not 301s, which will tell SEs like Google that the redirect is temporary. Finally, only run your experiments as long as necessary to avoid deceiving SEs.
I was able to find many companies that have put A/B testing to outstanding use. On the larger end of the company scale, Sony was able to test “variations of its homepage checkout page layout to increase purchases by 20%”(Optimizely). 20% for an international multibillion dollar company like Sony has got to mean tens, if not hundreds of millions of dollars in revenue gain. Impressive. On the other end of the scale but by no means short of remarkable statistics, Majestic Wine was able to redesign the category webpage to increase online enquires for Wedding services by 201%. The list goes on and on with real life companies benefiting from A/B testing, which now seems like more a staple for growth in todays online world.