A/B testing is something you should be doing for all of your campaigns.

Here’s why: At the start of your campaign, you don’t know what works and what doesn’t. An ad with a totally bizarre image may be just the attention-grabber your campaign needs to stand out. An audience you didn’t expect to drive sales could start producing your best ROI.

Accurate A/B tests can make a huge difference to your bottom line.

Start with your creatives. Do you have multiple lifestyle photos or videos? Test which ones are most relevant to your audience. How about different headlines and post text? See what message has the greatest impact on your potential customers.

Facebook will tell you which ads perform best. After 500 people view your ad, Facebook will assign it a relevance score. The higher the score, the better the ad.

Next is targeting. Have you considered testing different types of audiences?

Try dividing interests into different categories. You could have a retail audience with few interests, such as Urban Outfitters and Forever XI, and have a fashion audience consisting of a few fashion celebrities or events in it.

Also, consider utilizing a Lookalike audience. If you have a good source, this can produce major results. Try creating one based off users that have visited your website or liked your page on Facebook.

There's a lot you can test. And you can do it without spending a fortune

As rule of thumb, I recommend testing $5–10 per target audience. That’s it.

When you use ToneDen, if your campaign's not seeing results from those initial dollars, our optimization software will automatically re-allocate your budget to a different target audience--one that is delivering positive results.

After that, ToneDen's smart performance reporting gives you real-time updates so you can easily scale up your campaign when it's successful. 

This lets you get the most out of your Facebook ad spend and focus on more important work.

What's Next?

Learn how to leverage ToneDen's Audience Creator to A/B test the best audiences.

Did this answer your question?