Site icon Search Engine People Blog

Quick, Affordable A/B Testing With PPC

ppc-ab-testing

Making decisions is difficult. The more people involved in that decisions and you have more and more opinions that potentially cloud the decision. Add in that the highest-paid persons opinion (HIPPO for short) carries significantly more weight and you could end up making a poor decision.

But what if data could inform your decision? What if you could put the choices in front of real customers and see which one they respond too more positively? You can! And it doesnt have to break the bank or take months.

A/B Testing With PPC

As mentioned above, the idea is to let data do the talking. Pay-Per-Click (PPC) is simply the strategy by which you obtain the data. You can use Google AdWords, BingAds, Twitter Ads, LinkedIn Ads, Facebook Ads, etc. as long as you set up the test correctly and measure your results accurately.

Setting Up Your Tests

There are a couple of things you need to know when setting up your tests. First, you need to make sure that the options (the A and the B of A/B testing) are being rotated randomly and evenly. If these conditions arent met you could easily introduce a bias that will throw off your results.

Second, you need to compare apples to apples. This means that if you want to test a phrase in the headline of an ad, the other components of the ad must be identical so that the only variation would be due to the headline. You want both variations running during the same time period instead of one after the other. You control as many variables as possible so that the difference is a result of the tested component.

Lastly, you need to achieve statistical significance. This is a fancy way of saying that you need enough results to make sure the difference wasnt just chance. For example, if I flipped a coin twice and it was heads both times, it would be ridiculous for me to say that this coin only lands on heads. Same with your data. I dont have the time to cover it here, so Ill point you to a great tool here.

What Can You Test?

Most people readily think of testing different headlines or copy via PPC. Thats easy, but you can also test visual elements as well by using display ad units in your remarketing or display campaigns. Just use the criteria above to set up the test and a world of possibilities opens up to you.

Allow me to share a few examples of how you would put this into action.

You can test virtually any text or image-based element through PPC very quickly, but how much money is it going to cost? Doesnt have to be a lot actually.

Testing On A Budget

To keep the costs down on these tests, there are a few tricks you can employ. Consider these:

Conclusion

In short, PPC is a great way to A/B test any component of your marketing efforts. It can be done very quickly and provides you with real data on how real customers respond to the different messages/images. Then its not about whose opinion is right or wrong, but which component elicited a better response from customers. It can also be done quite affordably if you put your tests in the right places, so get out there and start testing instead of just guessing!

* Leader image by futureshape