The smartest approach to A/B testing your advertisements

Cut costs by performing split tests on your ad concepts before you launch your campaign

It’s hard to believe that A/B testing didn’t really exist before the turn of the 21st century. People would make their best guess and launch ads, websites, and other products without testing actual designs. Now you can use surveys and other A/B testing methodologies to create the most effective advertising before you start spending your precious media budget.

A/B testing – also know as split testing – is a way to test two ideas head-to-head to see what yields the best result. It’s a way to put some analytical rigor behind testing hypotheses to determine which ad is the “winner” for your business.

In a nutshell, split testing allows you to show two identical audiences different versions of something at the same time. Imagine that you’re running a print ad and you don’t know whether it would be better to show an image of your product by itself or an image of someone using the product. You can use a split test to determine which version of the ad could driver better results.

Don't guess. Know exactly which concept will win with SurveyMonkey Audience.

Let’s continue to use the print ad example. Using a survey to A/B test will enable you to weed out versions of the ad that don’t deliver against your KPIs (key performance indicators). As you create different versions of the ad to test, you’ll want to think about the following:

  • Define objectives: What are you trying to achieve with your ad? Once you know what you’re trying to do – drive awareness, trial, sales, etc – you’ll be able to layout your test plan.
  • State your hypotheses: Now that you have goals, what do you THINK will perform best? And more importantly, will you be able to take action based on the results?
  • Create a list of what you want to test: What do you want to test through a survey? Potential things to test include how the ad looks (color, photos, graphics), what the ad says (headlines, ad copy, call to action) and the overall design (ad concept, layout, number of visual elements).

Want to learn how to test out a few concepts before you go live? There are a few important things to do to ensure accurate A/B test results. (Also, before you launch your survey, think about whether you should be running a survey or a focus group.) You can really win at A/B testing when you:

  • Test one variable at a time. Let’s say version A has an image of your product and version B has an image of someone using your product in context. When you are split testing an ad, make sure you are only measuring the impact of one change at a time, in this case the main image. If version A has a different image AND a different headline than version B, you won’t know what if the headline or the image is causing the result.
  • Show the ad to similar audiences. Show version A and version B to the same type of people to get accurate results. If you show version A to existing customers and version B to new customers, you may get skewed results. Ideally you would take your target market and split it evenly to view and rate the ad.
  • Be patient. Advertising media costs can be a large percentage of your marketing budget. Take the time to test all of the variables you believe are important before you launch your ad.

Let’s continue to think about A/B testing the main image in your ad. Here are some of the types of questions you can ask your split audiences about the ad.

  • How appealing is this ad?
  • How likely are you to take action because of this ad (e.g., learn more, visit your business)?
  • How likely are you to buy this product?
  • How appealing does the brand make this product?
  • How is easy is it to understand the ad?

Remember, you can be showing different visual designs, copy, and layouts, as long as you test one thing at a time.

If you’re running online advertising, you can also do A/B testing in the wild with live ads. The great thing about a survey, versus live A/B testing, is you can also ask WHY someone chose an ad. You can also get ideas through free response feedback you didn’t originally consider.

Ready to get started? Once you start testing and seeing the impact effective advertising has on your business, you’ll create an insatiable appetite to learn more and continually make improvements for your customers.

ブランド マーケティング マネージャー

ブランド マーケティング マネージャー向けのこのツールキットは、ターゲット層の理解、ブランドの成長、ROIの測定をサポートします。



マーケティングトレンド: 2024年にマーケターが優先すべき3つの戦略


選択回答形式の質問: 例と使い方


Don't guess. Know exactly which concept will win with SurveyMonkey Audience.