Split Testing - What to test and why

Split Testing

There are so many variables that come into play when running Facebook Ads and one mistake that most people make is that they don’t test. I don’t believe there are "failed" ad campaigns. If you run one ad and the results are less than ideal, use the data to find out why, make a change and run them again. The numbers don’t lie and they give you such great data so you know what NOT to do the next time around. But, let me back up a bit. When it comes to testing there are a few elements you’ll want to test and if budget allows test as much as you can.

Audience: I generally teach entrepreneurs to test 1 audience in an ad set at a time. So, for example: Run 1 ad to the fans of Marie Forleo. Then, run another ad to the fans of Digital Marketer. The reason I don’t suggest lumping them together is because you won’t know if one audience worked better than another. Now, when you follow this method, it means that each audience will get it’s own budget and that might be something you can’t afford. If that’s the case, then group the audiences into "like" categories. So, build an audience of ‘Coaches” which could include Marie Forleo, Tony Robbins, Brendon Burchard, etc. and then build another audience of ‘Digital Marketing businesses’ like Digital Marketing, Social Media Examiner, etc. You just want to make sure everyone in that audience aligns. So, don’t group Dog Grooming with Financial Institutions, if you know what I mean.

Ad Image: The image is the first thing people see, but if you’re not targeting the right audience then it doesn’t matter WHAT type of image you have it just won’t resonate. (Which is why I generally test audiences first). Test a few different images (one of which can include a video) against each other to see if there is one that stands out as the winner. You can test an image that features your product (try placeit.net to build mock ups of your product inside a Ipad/Laptop/Ebook cover, etc). You can test an image with your headshot in the photo (I only run this to a warm audience of people who know you though). You can also test a video ad—maybe you have a freebie and you want to talk about what it includes. Be sure to include a strong call to action at the end of the video and also upload the captions. Facebook can do this for you with a click of the button.

Ad Headline: Headlines matter! It’s generally the second thing people see after looking at the image. Test different short phrases, but make sure they are STRONG headlines (Check out this free mini course on 10 Strong Headlines by group member Sarah Cousins , Founder of Ideas Girl. >>>>> www.10strongheadlines.com  

Ad Copy: Short or Long? Test both. You never know what will resonate with your audience. I'm a skimmer and like to get straight to the point so long copy doesn't work for me personally, but it definitley does for others.  Sometimes if you are targeting a cold audience, long copy works better because they need to 'warm up' to you a bit. But, test both and see what works for your audience.

Landing Page: If you use a software program to create your landing pages consider testing that as well. I use Leadpages and they offer A/B split testing so making a variation of your first landing page is easy to do. You can test things like the landing page Headline (that would be my first choice) and also button colors (try red vs yellow)

As always when you’re testing, test one thing at a time. Let the ad run for a few days and review the stats to see what the numbers are telling you.

Does this lengthen the process of running ads? Yep.

Does this mean you might need to spend more money up front. Yes, siree.

But, the rewards include getting back great data. If you put some money into the testing phase when you run another set of ads you’ll know what works - - and what doesn’t!

So test….tweak…repeat.