
Digital, Strategy, Technology

Mollie Kuramoto
Updated August 12, 2022
The world is digital—and one of the major upsides for marketers is the ability to track user engagement and gather real-time insights. Whether that creeps you out or you’re in the data-is-life corner, that doesn’t change the fact that when launching campaigns or sending emails, you should be following the ABT’s of marketing: always be testing.
Campaign testing is the process of creating a hypothesis and experiment for your marketing campaigns or assets that will return some kind of conclusion—and hopefully lead to future optimization. Let’s start with a simple example—pretend you’re sending an email to promote a clearance sale. You might create a hypothesis that if you include the discount price in the subject line, more people will open the email than if you don’t include it. Here’s what the experiment looks like:
Testing your marketing campaigns is important because it’s an easy way to learn more about your audience and optimize your marketing efforts. While failing to test your marketing efforts doesn’t really hurt your campaign, it’s essentially a missed opportunity to incrementally improve your efforts.
Whether it’s your marketing automation platform or ad platform, most software either has built-in functionality to run tests or will utilize machine learning to optimize campaigns. For example, HubSpot allows you to easily A/B test elements within your email sends, and Facebook can mix and match both ad copy and creative to optimize towards a goal you set in the platform.
In the past, it was more efficient for smart digital marketers to manually control a lot of settings (think: ad combinations and keywords in paid search). Today, most platforms are continuously learning and improving your campaigns towards a specified goal—although each platform has its own set of quirks.
Because of ad network sophistication (Facebook is probably too good at personalizing ads...but that might be changing), a lot of emphasis for marketers testing campaigns is placed on the landing page.
When it comes to campaign testing, there are two main types: A/B testing (also known as split testing) and multivariate testing. I’m going to cover the high-level definitions here—for more in-depth information you can check out Optimizely or HubSpot’s resources.
Made the decision to start testing? Great! But you’ll need a few things.
The first thing you’ll need is a hypothesis or a measurable goal. The point of running tests is to remove subjectivity—you want to gather real data that will help you reach your goals efficiently, not just pick the option you think looks cool. Running a test without a specific, measurable goal is like running a race without setting the distance—it will be pretty much impossible to determine a “winner”.
As mentioned previously, a lot of software (marketing automation and ad platforms specifically) has testing capabilities and functionality built into the platform.
For landing pages, a popular free product that allows you to run various tests on your site is Google Optimize. And, if you’re open to paying for a platform that’s a bit more robust, Optimizely, VWO, and Unbounce are all good options as well.
Finally, you’re going to need traffic in order to test your campaign. If you have low traffic, that means you’ll have to run your tests or experiments for longer periods of time to get to a statistically significant conclusion. When it comes to A/B testing vs. multivariate testing, A/B testing is an easier and better approach to take if traffic is an issue.
When it comes to running campaigns, there are a few really important things to keep in mind.
First, let’s talk about traffic. If your landing page doesn’t get a ton of traffic (we’re talking less than 100 pageviews per week), it’s going to be hard to A/B test and pretty much impossible to try out multivariate testing. So, if the issue is traffic, it’s smart to try and solve that first.
The second piece of advice is to make sure that you understand your goals—this will help create tests and experiments that actually improve your marketing efforts going forward. It’s normal for some tests to come out as a wash and for the results from all variations to remain more or less the same. But if every test you run results in a wash, it might be a sign that you need to push the boundaries more or start testing larger initiatives, like content offers and mediums.
Finally, make sure to document and share results from tests with the team. A lot of times, designers and copywriters aren’t aware that certain headlines or layouts perform better than others—and if you don’t tell them, they’ll likely carry on without that knowledge. Pulling them into the conversation can also help create buy-in, and they’ll be more likely to understand why a landing page might need to be designed in a specific way. They might even be able to contribute new ideas on what to test for upcoming campaigns!
Testing your campaigns can be overwhelming at times, especially when there’s already a lot going on. And while testing won’t make or break your campaign, the potential to try new things and learn more about your audience, what resonates, and what compels them to engage with your offer is lost. Don’t miss the opportunity.
Get the latest marketing strategy, brand insights, and industry perspectives delivered to your inbox.