An A/B test is a scientific process. You start with a hypothesis, conduct a controlled experiment to prove or disprove this hypothesis, and end up with hard data you can use for insight into future decisions or further experiments.
With the help of the science behind a good A/B test, you won’t have to rely on educated guesses to know how to improve the effectiveness of your ads and maximize the impact of your ad spend — you’ll have evidence to guide your decision-making. And you can easily conduct A/B tests of almost any facet of your ads with LinkedIn’s Campaign Manager tool.
What is an A/B Test?
A true A/B test is a test that changes one variable at a time to see if it increases your metrics.
To conduct an A/B test, design two ad campaigns that are exactly the same except for one factor – whether that’s an element of audience, creative, timing, spend, etc. – and run them both at the same time.
A/B testing can be used to arrive at data-informed answers to questions like:
- Which type of image will make this ad resonate more with members?
- Which job title will be more drawn to the messaging in this ad?
- Which call-to-action language will be more effective at driving the desired next step?
By monitoring the effect that changing just one variable has on audience response, A/B testing can help you isolate and understand what you should and shouldn’t include in future campaigns.
Creating an A/B Testing Strategy on LinkedIn
No matter which variables you want to test, follow these four steps to run an A/B test for any content you’re using Campaign Manager to publish on LinkedIn:
1. Before you start designing the test itself, decide on a hypothesis you want to explore by slightly altering the campaign. For example, if you’re testing two different images, do you think one image will perform better for the audience than the other? Why?
Keeping track of your hypothesis throughout the experiment will guide what you pay attention to and make your findings considerably more helpful.
2. Create a campaign that you want to test. Customize the creative, bidding, budget, etc. Then, create an exact replica of that campaign as a second campaign.
3. Once you have two campaigns that are exactly alike in every way, change exactly one variable.
For example, if you want to test one display image versus another, run one campaign with “image A” and the other with “image B.” If you want to test the effectiveness of image A with two different audiences, meanwhile, run both campaigns with a single one of the two images, but change one attribute– such as job title, industry, seniority, etc. – of the second campaign’s audience targeting. The better you isolate individual facets in your testing, the easier it will be to pinpoint specific performance drivers.
Let both campaigns run for at least two weeks for the best results.
4. Make sure to also keep track of your A/B tests over time. Use your original hypothesis to help you glean the most relevant information from the data. Was your hypothesis correct or not? Why or why not?
Just as importantly: Does the data reveal any unexpected discrepancies in addition to proving or disproving the hypothesis? For example, maybe your hypothesis was “image B will perform better with younger audiences than image A.” Maybe this hypothesis was proven correct, but only for younger audiences who also had a particular job title.
By conducting another A/B test and applying a hypothesis about why young audiences with that job title responded to image B, you can learn even more about your target audience and how to appeal to them in future campaigns.
Following these four steps will allow you to A/B test LinkedIn content from directly inside the LinkedIn Campaign Manager.
You can also test even more variations at once with LinkedIn ad variations. Just follow these steps as you set up or edit your ongoing campaign:
Creating a new campaign
1. During the Step 2: Set Up Ads section of setting up a new campaign, add multiple creatives to the campaign by clicking Create new ad or Browse existing content.
2. After adding all creatives to the campaign, click the settings icon next to Ads in this campaign at the top of the page to select ad rotation options.
3. There are two types of Ad rotation options:
- Optimize for performance: This option will collect performance data from each variation and then begin serving the highest-performing creative more frequently as the campaign goes on.
- Rotate ads evenly: This option enters each ad variation into the auction evenly, and doesn’t not consider performance even after collecting data.
Editing an existing campaign
1. Click into the campaign you wish to edit.
2. Click the more icon next to the campaign name and select Edit from the dropdown.
3. Scroll to the bottom of the Set Up Campaign page and click Next.
4. On the Ads in this campaign page, add variations to the campaign by clicking Create new ad or Browse existing content.
5. After adding all creatives to the campaign, click the settings icon next to Ads in this campaign at the top of the page to select ad rotation options.
A/B Test or Run with Variations?
A/B testing and running ads with multiple variations are both very helpful tools, but they’re useful for different things.
How to use A/B testing
Use A/B testing when you want to test a hypothesis about why one ad variation will perform over another. Because A/B testing runs both campaigns simultaneously and only alters a single variation, it is a true scientific experiment that will produce a result you can draw comparative insight from.
A/B testing won’t help you optimize the campaign you’re using it for, but it could help you better understand your audience and creative assets to improve all future campaigns.
How to use ads with variations
Use ads with variations when you want to find the best creative for the current campaign (when optimizing for performance) or when you need to test several minute creative changes at once (when rotating ads evenly).
Running a single campaign with ad variations will only allow you to test different variations of the creative in question. If you want to test against multiple audience types or other factors of the campaign, you will need to create two campaigns and A/B test.
Running ads with variations is a great way to optimize campaign-by-campaign, but A/B testing will teach you more about why your individual variations perform the way they do. LinkedIn recommends using both frequently. You could even A/B test along with multiple creative variations by creating two campaigns, making one slight variation to targeting, and then running them with the same set of ad variations and ad variation options.
The most important piece of advice we can offer on creating A/B testing strategies with your LinkedIn ads is to keep experimenting. Document any and all findings you derive from your A/B tests and use them to come up with new hypotheses to test. The more you test, the more you’ll find, and the better your ads will speak to your audience.
For more practical advice on making the most of your creative assets on LinkedIn, keep up with the LinkedIn Ads blog.
Topics: LinkedIn Ads
Related articles