A/B Split Testing Made Easy with 5 Steps

Magnifying Glass Analysis A/B Split Testing Made Easy with 5 Steps

A/B Split Testing, or just A/B Testing, is a form of statistical hypothesis testing, although it is often used in the less mathematical process of comparing outcomes of two variations is web design in order to maximize the desired result (e.g. clicks, sign-ups, purchases). On websites, two variations of design (labeled A and B) are compared in order to determine which is more effective at getting clicks.

Hypothesis Test vs. Optimize Web Design

In statistical hypothesis testing, a difference in results may be considered statistically insignificant so that it cannot be concluded that a true difference exists. For example, the outcome of test A may result in a 7% higher rate than the outcome of test B, but based on the available statistics (e.g. standard deviation) this 7% may be too insignificant to verify that outcome A is truly higher. However, when used by web-designers, intuition plays a large role and almost any difference is viewed as significant.

The ultimate goal of A/B split testing for web-designers is to maximize desired results, so conducting a true hypothesis test is technically not necessary, which is why A/B testing is a successful optimization practice.

Why Should You A/B Split Test?

The reason why A/B Split Testing is important is because you can maximize your desired outcomes with your current traffic. Often webmasters may have a high daily volume of visitors, but poor click-through-rate. A/B testing attempts to optimize your click-through-rate, purchase rate, newsletter sign-ups, etc. to get the most out of your traffic.

Essentially, A/B Split Testing is a way to find out what your visitors find most appealing, rather than the unrealistic practice of individually asking them which design they like best. Visitors collectively tell you which design they like best by how they respond. Repeated A/B testing can significantly improve desired results.

How to Conduct A/B Split Testing

Rather than getting into all the technicalities and statistics, A/B split testing can be made easy following 5 steps.

  1. Determine Desired Outcome (Clicks, Purchases, Sign-ups)
  2. Determine Amount of Time to Compare (Days, Weeks, Months)
  3. Determine Conversion Rate: Clicks per Visitors (ex: 9 clicks per 105 visitors = 9/105 = 8.57%)
  4. Make Design Change (Color, Location, Size, etc.)
  5. Compare Results

All you need to do is make an adjustments to your current web-design and compare the desired results. For example, if you want people to sign-up for your newsletter, then change where you offer your newsletter sign-up form on the page (location), or change the color of the sign-up button and form background.

You can literally test anything and everything, although you may want to test one change at a time in order to isolate which change made the improvement. If you change the location of your sign-up form and simultaneously change the colors, you may not know which change was the most effective.

You will want to determine your sign-up rate before you make the change (ex: 41 sign-ups for 2,642 visitors = 41/2642 = 1.55%), using a set period of time, and then compare the sign-up rate after you made the change for that same amount of time (ex: 67 sign-ups for 2,839 visitors = 67/2839 = 2.36%). Whichever test produces the higher outcome, keep it, and compare it to a new change.

Ideally, you want the time period to result in a large volume of visitors for comparison. The more visitors you compare, the more reliable the conversion rate. For example, 6 sign-ups for 283 visitors is not as reliable of a conversion rate as 67 sign-ups for 2,839 visitors.

Keep testing until you are satisfied with your improved outcomes, but be realistic about what's possible. An overall conversion rate of 4% is a realistic and good outcome rate. If you are able to increase your conversion rate higher than 4% then be proud of your accomplishments. Many websites claim a 100% increase, but they are comparing their old rate to their new one. A change from a 1% conversion rate to a 2% conversion rate is a 100% increase.

Tips to Optimize Website

Colors draw people's attention, so be careful how you use them. Limit your website design to a maximum of 3 or 4 different colors (white, gray, and black don't count). Too many colors can be overwhelming to visitors, and your call to action can get lost.

Your call to action should have a consistent color. For example, if you want someone to click or sign-up, make the link or button green for all links that lead to a desired outcome. You can do A/B split testing with different colors, but just make sure that the color you use is only used for the call to action and not other links.

A page that is mostly white, with dark gray writing, and a small orange button with "sign-up" will get significantly more clicks than a page with lots of colors, because the orange button demands the visitor's attention against the mostly white page. Likewise, if your page is mostly white, gray, and black and then a colorful advertisement appears, it demands the visitor's attention.

Article: Online Business Entrepreneur: Build Your Own Online Business

Marketing Articles

Enjoy this page? Share it!

Get more great stuff like this delivered to your inbox

YOU MAY ALSO LIKE

  1. How to Survive Your First Year in Business [Infographic]

    How to Survive Your First Year in Business [Infographic] Here's the first thing you should know about your first year as an entrepreneur - it's tough. There will be financial and psychological...

    Read More

  2. 52+ Common Interview Questions and Answers

    Most job interviews consist of only 5 to 20 questions, almost all of which are variations of these 52 common interview questions and answers. The following guide provides suggestions on how best...

    Read More

  3. The Millennial Job Dilemma: Why Does Everyone Want Experience?

    Many job-seeking Millennials find themselves asking the same set of questions: Why Does Everyone Want Experience? How am I supposed to get experience if no one gives me a chance? Often Millennials...

    Read More

  4. 5 Expensive Common Mistakes Made by Millennial Entrepreneurs

    As a millennial entrepreneur, handling the funds of your business can be quite tricky. Our friends over at FindMyWorkspace.com discuss the expensive common mistakes made by millennial entrepreneurs...

    Read More

  5. Link Exchange and Backlinks for Business Websites: Link Exchange Directory

    Want to exchange links for your business-related website? The following is a list of ways to get backlinks, as well as a list of websites willing to participate in link exchanges. Submitting your...

    Read More

  6. Forex: How Much Money Do You Need to Trade the Daily Timeframe

    The answer might surprise you how much money you need to trade the daily timeframe in forex. If your forex broker uses standard lots, then in order to trade one micro-lot of 1,000 you only need...

    Read More

  7. How to Trade Forex with $100: Turn $100 to $1 Million in 3 Years Realistically

    Not only am I going to show you how to trade forex with $100, but I am also going to show you how to turn $100 to $1 Million in 3 years realistically in forex. Sound impossible? Let me prove it...

    Read More

  8. Forex Trend Trading is How to Win in Forex

    I’m going to prove to you that forex trend trading is how to win in forex. There are some who claim that it is foolish to buy something that is rising in value, but I think they are foolish for...

    Read More

  9. Advanced Forex Education

    Want my advanced forex knowledge? Learn how I ensure I win with advanced risk-reducing techniques and leverage management. Normally available for $50, you can have access for only...

    Read More

  10. Trading Forex: $1000 to $1 Million in 2 Years Realistically

    Is it possible to turn $1000 into $1 Million in 2 years? Yes, it is possible to turn $1000 into $1 Million in 2 years and it is even realistic with this consistently profitable strategy...

    Read More