Even the strongest content marketing program on the planet could be improved upon. Chances are good that the person running the program understands this and puts it to work for her every day.
The magic word is “testing.” With today’s access to data, marketers have no excuse for stumbling blindly through their campaigns.
A strong analytics program doesn’t always mean you’re doing it right either. Marketers too often toss content out to their audiences with only a vague idea of where to set the bar for success. Testing requires carefully crafted benchmarks and goals to help you define your success.
Your content marketing represents a huge opportunity for testing. But first, it’s important to understand why testing is an essential function of your job.
Relying on instinct doesn’t count
Instinct, intuition, feelings—these are proof indicators that have no place in marketing. Yet we rely on them often.
Marketing programs with strong analytical backbones sometimes miss the point. Several elements work in tangent to make your content successful or unsuccessful. Testing them helps you refine your formula for content marketing, improving the rate at which content converts your readers into customers.
Content marketing programs everywhere could benefit from asking testable questions like:
- Which element(s)—headline, pictures, suggested content, etc.—of this post make it successful? Which could be improved?
- Does this content appeal to the audience for which it was intended? Or does it appeal most to an unintended audience?
- Where in the sales funnel does this piece of content have the biggest impact? To put it another way: At what point in the buyer’s journey do they seem most receptive to this content?
Your testing can get as granular as you want to make it. Finding the elements that work together to create successful content is the ultimate goal.
Experimenting with your content must follow a proven method. Marketers must approach their own testing and experimentation in much the same way scientists see evolution as an evolving theory itself.
Testing based on the scientific method
Testing and science go hand in hand. It should come as no surprise that your marketing experiments should be rooted in the scientific method.
If you’re a little rusty on the scientific method, here’s a refresher.
- Question: Formulate a question based on an observation. Ex: Why is that blog post converting at half the rate of similar posts?
- Hypothesis: Here’s where you identify an area that may have an effect on the outcome. Ex: The call to action is the main driver of conversion behavior on this page.
- Prediction: In this step of the method, you’ll guess the outcome of the experiment. Ex: If I make the call to action at the bottom of the page more specific, the conversion rate will double.
- Testing: Next, you run the test and collect the data.
- Analysis: Weigh the data against your benchmarks to find out what kind of impact the change had on your conversion rate.
This is a very basic example of the scientific process in action for content marketing. Marketers can (and should) apply it to pretty much any situation you can imagine.
Typical methods marketers use to test
Marketing experiments can run the gamut from very specific to open-ended. Testing methods can differ radically depending on what you’re trying to accomplish. Here are three popular ways to design your experiment.
- A/B testing: One of the most common methods used in design, A/B testing pits a control page against one with a single variation. For example, you might change the headline on a popular blog post with consistent traffic. In this case, you’re isolating the headline to find whether or not a different headline could work better.
- A/B/N testing: This method of testing is similar to A/B testing, only you test multiple (as many as four) versions of a headline against the original in order to find out which works the best.
- Multivariate testing: For a multivariate test, you might change the headline and the featured picture to see which version of the page works best. Here, you have the ability to test radically different versions of a page against the original rather than just individual elements.
Testing requires the right software. A strong CMS or marketing automation platform is usually a good place to find testing technology. If you don’t have access to one, check out inexpensive (or free) options like Google Content Experiments, a part of Google Analytics dedicated to A/B/N style testing.
How do you test content?
Do you use inexpensive tools to test your content? What approach works best for your brand? Share your experiences with us in the comments.
About the Author
Mark Sherbin is a freelance writer specializing in technology and content marketing. He shares occasionally insightful information at Copywriting Is Dead, where he promotes authentic communication between organizations and their audiences. Say hello on Twitter: @MarkSherbin.