Crash course: A/B testing for your PPC campaign
By: Sarah Rosenquist
August 22, 2017 | Reading Time: 3 mins
There are a lot of different variables that go into a PPC campaign – bidding, targeting, budgeting, copy, and imagery. How do you know that what you’ve launched is the most effective way of finding and talking to your audience? You really don’t until you do some A/B testing.
Make note of your assumptions
Assumptions are an unavoidable part of marketing. Having a firm grip on what you think you know is the first step. These are the theories you can A/B test.
Some common examples of assumptions in a PPC campaign could be:
- This targeting strategy is the best way to find my target audience online
- This copy and image will resonate with my audience
- This bidding strategy is the most effective way to be spending my budget
We find the most crucial assumptions to test are targeting strategies and creative. Testing your bidding and daily budget is something we usually do once the first two are locked in.
Determine what your variables and controls are
To get helpful data out of your A/B test, you need to isolate variables. That means you create two campaigns that are the exact same in all aspects except one. This way, any differences between the campaigns’ performance can be attributed to that one alteration.
So the only thing we’ve changed in this example is the interest-based targeting. The same ads are running in each group. What way is better at honing in on my target audience? That’s what this test would help you answer.
Let the campaign run its course. The only time you’d want to make adjustments during the test is if you’re not coming close to your daily budget. If that’s the case, adjust your bid or expand the demographics equally on both campaigns.
Listen to your data
So the A/B test campaign has run its course. What won? The answer can be found by comparing the stats of your two campaigns.
Hmm, it looks like campaign A won in some ways but campaign B won in others. How do you determine the true winner?
That answer is based on the objective of your campaign. Here are some ways you can choose a stat that aligns with your goal:
- Brand awareness? Look at impressions.
- Driving lots of visitors to your website? Look at clicks and CPC.
- Having a message that resonates with your audience? Look at CTR.
- Generating leads? Look at conversions and CPL.
In almost every case you should let conversions and cost per lead (CPL) determine your winner. This is because it’s a stat that measures the success of your entire campaign. It takes into account who you’re reaching, how effective the ad you’re serving them is, and how convincing your landing page is.
Let the winner run and move to the next A/B test
Now you have confidence in your campaign. Put the remainder of your budget behind the winning test and keep checking in to make sure it’s keeping up with its performance.
With that cruising along nicely, you can now start planning your next A/B test. Never stop iterating and optimizing. After all, more data to inform decisions never hurts!