Facebook ads are an invaluable tool in digital marketing for auto dealers. They help you get your message in front of the people who are most likely to care about it.

A/B testing, or split testing, can help you hone in on that message even more.

Split testing helps you narrow down your audience or your ad creative until you find the sweet spot of audience and message intersection. That helps you make better ads that your customers actually interact with, which is a win/win situation for both you and your audience! What could be better?

. . . How about a comprehensive article on how to make awesome split tests on Facebook?

Read on. And if at any point in this post you get confused about a term or two, here’s a helpful and quick guide to some Facebook ad jargon!

1. Start with a Goal

Remember in science class when your teacher told you that the only way to do a good experiment was to start with a good hypothesis? Turns out that applies to split testing Facebook ads, too.

In other words, what do think you’ll accomplish with this test?

If you can’t finish the sentence “At the end of this split test, I want to know _____,” then spend a little more time thinking before you set up a Facebook campaign in the updated Ads Manager.

(Psst: not sure how to set up a Facebook campaign? Check out our full guide to setting up a pro-level Facebook account).

Ultimately, the goal of every split test is to increase your relevancy metrics and make your work a little easier in the future. It’s always easier to make ads that perform well according to the Facebook metrics that matter when you’ve gotten some answers ahead of time!

For example, we used a split test to figure out that photos of real vehicles outperform stock images. Other times, we’ve tested whether one call to action (CTA) performs better than another.

We use those findings to build better Facebook ads in the future.

2. Test Just One Thing at a Time

The key to a successful Facebook split test is to test just one thing. That’s called the variable, and it should line up directly with your goal for the ad. 

So, for example, if you want to know if your audience prefers a clearly defined CTA button or no CTA at all, test one ad set with a CTA and one ad set without a CTA. Your results will be far more conclusive if the only element that changes is the presence of a CTA button.

If your CTA ad also has a different headline than your “no CTA” ad, you can’t be 100% certain whether it was your clever writing or the button that made the difference. 

Keep your variables limited, and your answers will be clearer. 

3. Find the Right Folks

Targeting the right people is super important with Facebook ads. If you feel like your target audience is almost perfect, but not quite, then consider a split test comparing one audience variable — like a specific age group or interest set — against another.

Again, keep your variables limited.

If you’re trying to narrow your audience down, keep the creative the same. If you’re testing the creative, keep the audience the same.

Make sense? Cool!

4. Set the Timer + Take Note of Your Findings

Facebook recommends running a split test anywhere from three to 14 days, but it also says that four days is the sweet spot if you don’t quite know where to begin.

If your results aren’t conclusive after four days, you could try running the test again for a longer amount of time or testing a different variable.

Sometimes, our A/B tests are incredibly conclusive. Sometimes, even with the right variables and amount of time and everything . . . not so much.

Either way, it’s worthwhile to document your findings so that you can make shorter work of testing out the next set of variables. Make sure that you’ve got a master list that holds your split test findings all together in one place.

Where to Go from Here

Now that you’ve got your ads (or ad sets) set up, let them run their course, and then analyze their results. The kinds of thinks you’re looking for is which one had a higher click-through rate, which one had a lower cost per click, and so on. 

However, take note if one ad had a significantly different reach. Unless your ads are distributed to about the same number of people, your A/B test results will be somewhat contaminated.

Getting your Facebook ads up and running is awesome, and split testing is fantastic — but it’s all even more effective if you set up a kickass Facebook account for your business, take advantage of the Facebook pixel, and understand the recent updates to Facebook privacy laws.

Once that’s all in place, you’re off to the races.

Get even further with some help from our Facebook field guide for auto dealers (order your copy right here). Or soak up some free Facebook know-how from our 9 Clouds Live video series by clicking below!

Happy testing!