Ad Testing

Paid media in eCommerce is very much a matter of trial-and-error. Once you've devised an advertising strategy, the majority of your time will likely be spent testing, iterating, and testing again. This process is known as ad testing. 

Testing a paid advertisement generally involves swapping minor elements between two ad test pages — a visual component, a font, or a piece of copy — then assessing both ads to see which performs the best; rinse and repeat ad nauseam. 

Why is Ad Testing Important? 

Caption: It's crucial to monitor your ad spend at every stage of a campaign

‍

Ad testing helps ensure you make the most of your advertising budget. It takes a data-driven, iterative approach to paid media, and in the process helps you develop a better understanding of both your audience and your brand. In the long-term, it also helps you create and deploy better-performing ads. 

All other things being equal, an eCommerce brand with a comprehensive ad testing process will always enjoy greater reach, impact, and ROI on its advertisements than one that doesn't test its ads at all. 

Understanding the Basics of Ad Testing

Depending on your platform, there are many different ad testing tools and techniques available to you. However, they can all be boiled down to the following basic process: 

1. Ensure you have the necessary budget — typically, you'll want to spend no more than 15 percent of your total advertising budget on testing. 
2. Start by defining your objective. What was the original goal of the advertising campaign? What do you want to test about your advertisement, and how will this help you pursue said goal? 
3. Create several variations of the same ad, each with minor differences. Do not change more than one characteristic or component for each advertisement — too many variations can make it functionally impossible to discern the precise reason one ad outperforms the others. 
4. Lay the groundwork for performance monitoring. You'll want a high-level view of each ad for the duration of your test. Make sure you have a means of monitoring metrics such as conversion rate, total impressions, click-through rate, and return on ad spend. 
5. Run your test ads over a set timeframe. We recommend between two to four weeks, or however long it takes you to get to several thousand impressions. 
6. Assess which variation was most successful. Refine your strategy and start testing a new set of variations. 

‍

How Does Ad Testing Vary By Platform? 

Caption: Varos can show you where you're directing the most paid media spend — and by association, where you should direct the bulk of your ad testing efforts. 

Facebook Ad Testing

Although TikTok is gradually eclipsing its user base, Facebook is still one of the largest social networks in the world. With that said, paid advertising on the site is more challenging than ever. Facebook lead ad testing is therefore non-negotiable if you want your advertisements to have any chance of success. 

Your options include: 

• Creative. This involves using different types of creative media, including static images, user-generated content, and video.  
• Ad copy. Testing different combinations of copy and creative. 
• Audience. Tweaking ad targeting based on interests, behaviors, and demographic information. 
• CTA. Modifying the call to action button. 
• Landing Pages. Changing the location to which your ads link. 

Google Ad Testing

Given that it remains the world's largest search engine, Google is still the reigning king of paid media, particularly since its parent company Alphabet also owns YouTube. It follows that in addition to a powerful suite of search engine optimization tools, Google also provides a multitude of options for testing ads within its Experiments tool. 

Accessible within Google's Ads platform, Experiments automatically divvies up your ad spend between variations based on:

• Copy
• Creative
• Keywords
• Landing pages
• Adwords bids

Other Ad Testing Tools

Most social networks and search engines provide their own tools for testing and optimizing your advertisements. Provided you've the right data available to you, this is also something you can do manually. If you'd rather not, some potential options for ad testing — and market research — include: 

• Kantar Marketplace
• Zappi
• quantilope
• Ameritest
• Response:AI