AB Split Testing and your Marketing Plan

split testOne of the most important aspects of your marketing plan is the measurement of the advertising campaign's effectiveness. A crucial component of overall testing is the split test, or AB test.

In essence, what this means is that you are adding a variable within one of your marketing or advertising efforts in order to see if one performs better than the other. Sometimes a small adjustment can have a big improvement on conversion rates.

So for example, to test a sales page, the split test there would be to have two versions of the same page except one has a different headline, the content has been changed slightly, or a format/color difference.

The split test is executed by showing one variation of the sales page to one customer, while showing the other variation to another customer. With enough views of your sales page, a difference may become evident that one version works better than the other. You must allow for enough views however because random chance may give a false impression that one version works better than the other. Like flipping a coin, you know it should come out 50% head and 50% tails. The more flips, the closer to 50% you will get. Not enough flips (or views of your variations) and your numbers will be skewed.

Changing one variable at a time allows you to be able to see what has made the difference between the two versions in your split test. Once you have determined that one version works significantly better than the other, you can abandon the version that does not work as well.

At this point, you continue to perform another split or AB test by building off of the better version. Once again you make a small change to this version and serve both to the customers to determine which is the better one.

This does take time, but by constantly split testing your advertising, you will be ever increasing the effectiveness of the ads.

The most important aspect of the split test however is the ability to measure the results. There is no point in making two versions of an ad if you cannot know, or don't make the effort to find out which one has been more effective for you.

Here are more examples of how split testing can be used:

  • Two different ads in two different newspapers
  • Different alternating ads (daily, weekly, etc) in the same newspaper
  • Two different sets of direct marketing postcards
  • PPC ads with different content

Here are some more examples of what can be changed:

  • Different headlines
  • Different content
  • One ad stating price, and one not stating the price
  • Using different images or background colors

Split testing has been great for internet marketing. Measurement of incoming traffic is incredibly easy. But AB testing is not only for internet sales pages, it is for any and every advertising campaign. Although tracking metrics on the internet makes is much easier than with print media for example, tracking where your customers comes from is paramount if you want to improve your advertising effectiveness.

It is not always feasible to setup two different phone numbers for example, to be able to track incoming calls from one ad to the next, sometimes you just have to do things the old fashion way, that is, ask the person exactly how they found you. Regardless of how you do it, split testing by running two similar but distinctive ads at the same time and measuring their effectiveness is the key to improving your marketing efforts,

Popular

Marketing Plan Basics