The A/BCs of Testing Web Pages

Avatar Act-On
Content Marketing

AB testing diagramOut of all the headlines you could think up, all the offers you could make, and all the calls-to-action you could choose, how do you know which one will work the best? Even when you know your customers well and understand their buying habits, your gut feel can be wrong. You have to test to be sure.

The point of testing web page content such as button labels, link text, headings, benefit statements, and prices is to compare how your customers actively respond to one web page against how they respond to a variation of that same page, so you know quantitatively which of the two alternatives delivers the most conversions. This gets you away from depending on intuition and hope, and lets you maximize results with proven content.

The goal of A/B testing is to determine as quickly as possible whether the champion (your existing page) or the challenger (the new page) is best, and move traffic to the better performer immediately.

What to test

Testing the look-and-feel of your landing page is useful, but you’ll get the best results from testing the elements that affect the actions your visitors take.


If you don’t test anything else, test the button people will click to take action.

  • Language on buttons. Focus more on what people will get, less what they have to do. “Order Information” will probably not convert as well as “Get Information.”
  • Button shapes. Draw inspiration from others. Rounded corners usually work better, for all kinds of interesting psychological and physical reasons.
  • Button colors. Try colors that contrast with the other colors on your site. Use a color wheel.
  • Button placement. Above the fold is usually considered to be the best place, but it really depends on your visitor and your offer.If the visitor is going to show up ready to buy, make that easy. If you’re introducing complex ideas for consideration, give the visitor at least a little chance to think about them before presenting a button. If it’s a long page, try using several buttons, but make sure they make the same offer.
  • Keep your button uncluttered, and surround it with plenty of white space.
  • Try directional cues, such as an arrow that points to the button.


Different headlines can increase your bottom line, particularly if you’re using pay-per-click ads and including the keyword. You might consider testing your headline and then using its keyword in your PPC campaign. Things to test include:

  • Statements vs. questions: “Marketing Automation is Affordable” vs. “Looking for Affordable Marketing Automation?”
  • Numbers rather than words: “Five Ways to Increase Your Conversions” vs. “5 Ways to Increase Your Conversions.”
  • Tone and language: “Do You Have Milk?” vs. “Got Milk?”
  • Value propositions: “Save Time” vs. “Save Money.”
  • Length. Shorter is usually better, but not always.
  • Individual words that convey emotion or action can affect response.

Lead forms

Use A/B testing on your forms to improve the information you get and the number of conversions.

  • The fewer fields you ask people to fill out, the more form completions you’ll get. Conversely, if it’s important to get information you can use for targeting, you must ask segmentation questions. Test to see where the balance lies.
  • Test locations. Traditionally, forms in the upper right-hand side perform the best, but that could be different for a specific form, on a specific page.


  • Drawings vs. photographs.
  • Color, black and white, effects or no effects.
  • Living things: groups or singles, men or women, children, pets.
  • Products: different views, alone or with people using the product.

Other things to test

Depending on your unique circumstances, you could consider testing these:

  • Subhead
  • Paragraph text
  • Font color and size
  • Testimonials
  • Links

Try A/B testing – free!

Want to try A/B testing? With no risk? Register now for 5 FREE A/B TESTS
and learn the A/B Sees at Act-On’s online service

Get 5 Free A/B Tests