How do most marketers decide what will work and what won’t when designing the landing pages for the PPC ad campaigns? Their intuitions. Yes, that’s true. Most online advertisers like to predict what their customers will like, where they will click and how they will convert.

While it’s great to have a sense of the visitor’s behavior and intention, it’s equally risky to base such important business decisions just on feelings and intuitions. When you know that you’re paying for every click on your ad, you would rather go with tried-and-tested tactics than just using your assumptions and guesses.

A/B split testing plays an important role in any PPC campaign. It helps the marketer determine how the target users react, behave, like, etc. by trying out different variations of the ads and the landing pages. Another reason why A/B spilt testing is useful is that it helps you identify the tricks that are unique only to your business. It’s not necessary that what might have worked for other PPC advertisers would surely work for your business too.

So, it’s very important to include A/B split testing in your PPC ad campaigns. However, it’s very complex to run and manage A/B split testing. If you are not being attentive enough, you may end up spending unnecessarily without attaining any important insights about your customers.

That’s why I decided to give you some of my A/B testing secrets that I have discovered over the years while managing client PPC campaigns. Feel free to bookmark this page for your future reference:

When optimizing your landing page for a PPC campaign, there are many factors that you could consider for A/B split testing, such as, buttons, colors, fonts, copy, images, positioning, layout, form When optimizing your landing page for a PPC campaign, there are many factors that you could consider for A/B split testing, such as, buttons, colors, fonts, copy, images, positioning, layout, form fields, offers, etc. But, obviously, it is time consuming and also costlier to do a split test for all the variables. So, you need to identify the one factor that’s the most important to your campaign and focus on it dedicatedly. Observe the minutest of the detail in testing this factor, because remembert the smallest change can potentially bring you thousands worth of revenue. If you’ve got handson experience with PPC over the years, only then you should consider multi-factor split testing.

There are several metrics that you can aim at improving by doing A/B split testing. So, you need to determine which will be your primary objective in terms of metric optimization for the PPC campaign. This will help you differentiate between the two variations of A/B split testing in aconcrete manner. Once you have a clear goal, you can easily see what sort of changes to go after so that they can initiate or influence the desired action or behavioral change among your target users.

Do not try to experiment with both of your test conditions when conducting an A/B split test for your PPC campaign. You must control for one of the two conditions, meaning that in that specific conditions you should not make any changes. If it’s a landing page you are testing, then one copy of the landing page should remain unaltered at all times. And then, you can create different variations of the landing page based on what you already have in the controlled landing page copy. For example, if you are testing for testimonials on your split page, then the control page should not have a testimonial.

Just like testing two different pages of the landing page, you need to experiment with different sets of target users for you’re A/B split testing. Here, you control one set of audiences and only test the other set or sets of audiences. For example, if you are doing an A/B split test for your email campaign, then one email list should continue to receive the same type of emails all the time whereas in the other lists, you can experiment with the layout and the copy of the emails.

Once you have set your testing goal and created several versions of your test, then you must decide how significant you want your results to be when determining the importance of one variable being tested. This means that you need to set your confidence level for your testing. If you are 95% sure of the result of your split test, it means that the more significant would be the result. The more you can sure about your results, based on the higher the significance level. In simpler words, the more radical the change you expect from your A/B split testing, the lesser strict you need to be about the significance of the result. The more specific the change such as color, copy, etc., the more confident you need to be, that’s because the change is less likely to have a significant impact on your overall campaign conversion rate.

Even if you are testing for more than one factor for your PPC campaign, you need to ensure that you only run one test at a time. Let’s say that you are A/B split testing ad copies that bring customers to your landing page and also A/B split testing the landing pages, then how can you be sure what worked in the ad copy and what worked on the landing page? It will be difficult to differentiate between the two.

It’s easier to manage the A/B split tests if you use tools for that. For example, an automation tool like HubSpot will allow you to run your test very smoothly as it offers you numerous features. Even Google Analytics offers you Experiments features, wherein you can A/B split test up to 10 web pages simultaneously and compare their performance to find an optimal combination.

When running an A/B split testing, it is important that you run both or all variations in your test simultaneously. It has to be the exact same time. You cannot run one variation today and the second one after a month to decide which one is more significant. You will need to run both versions at the same time, or else it’s as good as doing a guesswork. However, if the factor you are split testing is timing, then you can run then at separate times. For example, on Facebook, you want to find out
when it the best time to run ads to get the optimum responses. Only in that case, it is okay to run the A/B split test at different times and still find relevant results.

You cannot run a test for a very short duration and then draw conclusions out of it. Of course that won’t be specific at all. You need to run the A/B split testing long enough to get substantial findings. So, you may ask, how long should we run the split-test? Well, there is no right or wrong answer to it. It really depends upon your type of campaign, your products, your niche, your audiences, etc. Sometimes a week is enough whereas sometimes even six months are too short to find significant conclusions. Until and unless you start noticing the difference in your campaign results, you should run your split tests.

Just because you finished one A/B split test, doesn’t mean that you cannot run another test. You might have discovered some important findings from your test, but you still need to continue with different types of tests. The online audiences are getting more and more dynamic and increasingly lesser patient. So, it’s very difficult to keep them pleased and satisfied at all times. As a result, you would need to run more A/B split tests. There’s always room for more optimization in your PPC campaigns. The secret is to always keep an eye out for the next possible opportunity that helps you increase your conversions and lower your costs.

So, there are some tips from my end for running A/B split testing for your PPC campaigns. Of course there are many more tricks and secrets. Do you know some secrets that you’ve learnt through your experience of managing PPC campaigns? Do share them with me in the comments below.