Back in the old days, knowing how to do ad testing involved adding exclamation marks or using proper case. Those were the days when you also only used one headline. But the times have changed, and so has the sophistication required for ad testing.
So today, let’s talk about how to do ad testing comprehensively.
Advertising is a costly venture and it has high-stakes. A great ad can boost your brand and gather interest in your products, and of cours, generate revenue for you. On the other hand, terrible ads fail to deliver against your investment and can even cause your brand irreparable harm. It’s really easy to get your advertising wrong.
But ad testing can help you get it right. It’s simply the process of vetting your ad concepts with a sample of your market.
One question you need to ask here is whether you should draft a Google experiment, or run a test in the current ad group? After several clients we’ve catered to, we’ve determined that it depends on a person’s preference.
Using a Google experiment is a good way to allocate a portion of your budget to your test, and determine if you’d only like users to see the test alone, the original ad, or if you want to make it random. If you prefer less cluttered ad groups, your experiments are kept in their own experimental campaign.
So then why’d you want to run tests without experiments?
You can see how Google treats your ads through using machine learning.
Start testing where performance is poor, but exactly which metrics should you compare?
Conversion rate and click-through rate must always be two of the metrics that you should pay attention to. They represent how users respond and convert in response to your copy. In addition, the first metric to pay attention to will depend on your goal. So it could either be conversions, cost per conversions, or ROI.
Rachel Poulos, Senior Account Manager at Hanapin Marketing provides a great example.
They’ve used ROAS as their goal, and in that table below, it’s plain to see that ad 4 has the highest ROAS and highest CTR. That means they’re making money, and the ad copy used resonates with their users. But if you want to consider conversion rate and CPT, it’s clear that ad 3 wins.
When you’ve determined your winning ad, make it your control.
Now you need to decide if you’d like to do your test in the headlines or description lines. If you decide to conduct a test in the headlines, make sure that it fits within headlines 1 and 2.
What else can you test?
Add unique value propositions, emit brand authority, or perhaps try dynamic keyword insertions (DKI). There are a lot of possibilities, but be careful not to add too many variables to your test. Because you won’t know if it was the personalization of DKI or your unique value proposition that did it.
Experienced PPC specialists and digital marketing agencies recommend running an ad for a minimum of two weeks before pulling it out to see results. However that depends on the amount of traffic that you’ve actually received and if your ad achieved statistical significance.
There are also other factors that you need to consider. Like, budget changes, previous campaign status, promotions, and seasonality. That’s why doing a PPC audit is important too.
Regardless if you’ve found a winner or not, move on to another test. Either way, you will still have and receive results that will point you to your next step.
Keep on testing.