Plenty of businesses advertise, but many aren’t taking advantage of their marketing’s valuable byproduct: Data.
Unfortunately, many companies are content to send emails, run ads and post on social media without a strategy or goals. Even worse is when these companies don’t use the gold mine of data generated by their marketing.
You can learn a lot about your customers from ads themselves. That’s why we encourage clients to think of it like a free market research tool at their disposal. This email data can be used to improve future emails, and if you’re lucky, the same data can improve other advertising.
How A/B Testing Works
A/B testing (or ‘split testing’) is a data-driven tactic we use to identify high-performing content, graphics, subject lines and more. In short, it helps you find out what’s working and what’s not, and backs it up with hard evidence. The best part of split testing is how it can be applied to nearly any piece of marketing that’s deployed to a sizeable audience.
Conducting a split test is just like a scientific experiment. You create two slightly different versions of the same piece of marketing (like a single email or digital banner ad) and deploy each version to a portion of your audience. After learning which version performs better, you can then apply that variable to other marketing materials knowing that your audience will respond favorably.
Email is the most common marketing channel for split testing because of how easy it is to control your variables and analyze test results. Most email marketing platforms have A/B testing tools built in, too.
A/B Testing Strategies
A/B testing is a simple tool, but it’s not very useful without a plan. Luckily, a little bit of planning goes a long way.
First, set some ground rules for things like test variables, how frequently you’ll test and how you’ll use the results. Like a scientific experiment, you want to eliminate as many variables as possible to make your results more trustworthy.
Testing an email subject line? Determine a standard day of the week and send time for your tests. Analyze your results a week after the email blasts every time. If possible, keep the layout consistent. Keeping these factors consistent will make your test results more valid.
Testing a banner ad’s headline? Make sure both versions of your test ad are deployed for the same time and same target audience. Decide which metrics matter most (like clicks) and measure them against each other to choose a winner.
Having a plan and a carefully chosen set of variables makes it much easier to apply your findings to future advertising.
A/B Test Examples
Any type of ad that’s deployed on a relatively large scale can be tested. Email subject lines are an easy and effective variable to test, but with the right approach, it can be applied to other popular marketing channels.
Our favorite variables to test include:
Variable: Ad Content
Qualifying ad content through testing is a category of its own, which we call an ad grid. This method is a great way to learn about a target audience without prior market research.
Example: Display ads for Unite Private Networks.
Through a fast-paced testing process, we discovered that our target audience of business owners of medium and large companies responded better to the ad on the right.
Variable: Graphic Styles
Graphic Style A
Graphic Style B
Different audiences respond to differently certain kinds of imagery. Testing this variable can provide an excellent guide for the visual side of your marketing.
Example: Facebook ad imagery for TOBA Foods
Through multiple test iterations, we found that grade school foodservice directors were more likely to click on ads with vector illustration graphics than ads with photo graphics.
Advertising legend David Ogilvy famously said “When you have written your headline, you have spent eighty cents out of your dollar.” The headline usually drives the entire concept of an ad, so this variable is a great litmus test for the overall quality of an idea.
Example: A recent subject line split test for a financial client saw a 13% difference in click-through rates.
Email with headline A received a 10.5% click rate.
Email with headline B received a 23.7% click rate.
Putting A/B Testing to Work
These examples are just a small sample of the variables we can test to improve ads in real time. They also teach you a lot about your target audience, which makes them worth their weight in gold.
Need a marketing partner with proven strategies for successful A/B testing?
Send us an email at firstname.lastname@example.org or get in touch with us by filling out the form below. We’d love to talk with you about how we can help you reach your marketing goals.