Every marketer feels the pressure to demonstrate value to their respective organization. Drive more revenue, increase conversions, lower the cost-per-lead — these demands are top-of-mind for marketers and motivate our work. But marketing departments can do more than generate revenue for an organization.
Marketing is positioned to serve as a business insight generator. With statistically significant data from experimentation, we can interpret our audience’s behavior and shed light on customer motivations.
Unfortunately, not every experiment is viewed as an opportunity for insights. Experimentation has been reduced to a Magic 8-Ball that tells you what to do based on the result of an A/B test, but does not reveal the “why” behind the “what” of a customer action. This short-sighted experimentation is an inherent flaw in marketing’s approach toward testing, and leaves us thirsty for actionable insights.
Below are three ways to take your marketing experimentation to the next level.
1. Expand your scope of insights
Marketers have a tendency to limit their scope of experiment insights by testing solely for conversion rate. The best (and belabored) example of this is testing which button color is more likely to convert visitors on a landing page. To produce insightful information, marketers must start asking insightful questions. When a scientist conducts an experiment, it is to test a theory. To test a theory, the scientist must start with a question that needs answering. In essence, your testing and experimentation as a marketer is rooted in curiosity. According to a study published in the Neuron journal, curiosity is a “drive state for information.”
Some of the best experiments start with the best questions about a client’s buyer persona. Would persona Cindy be more willing to convert if she knew our client was certified in the service they perform? Or is locality more important to her? What test can we conduct to find this out? If you’re not feeling curious, make a list of client problems that need solving or a list of assumptions to test. Better yet, ask your sales team for questions they have always had with their audience.
2. Narrow your testing variables
If we think of marketers like scientists, then it is right for us to identify a control and variable in our experiments. Like a scientist, we should test one variable at a time in our experiments so that any result or effect from the test can be attributed to one, singular change.
We fail when we test too many variables in an experiment. If we wanted to test which call-to-action increased conversions on a landing page, version B (the variable) must only differ from version A (the control) in its call-to-action. If a second variable on version B changed, such as the client testimonial, the experiment results are invalid because we don’t know if the call-to-action or client testimonial influenced the site visitor. (Note: Use an A/B confidence calculator to measure the statistical significance of your control and variable results.)
Limiting your testing variables is not to say that you cannot conduct multiple experiments at one time. Marketers are free to split test banner ad creative while testing a variable on a landing page. Just be mindful of where your experiments intersect to avoid competing influence on your audience.
3. Identify the right test
The insights we generate from testing transcends the marketing department and can impact all facets of our business, from sales call scripts to traditional media, branding and even market segment. With this in mind, choosing the right test is a critical first step in generating the insights that will increase the value of marketing within the organization.
Many articles offer ideas of what to split test, but only you know what question or problem needs to be tested to reveal the best information for your business. According to Peter Kochley, founder of Upworthy, “we rely on testing to just make better decisions. People are really fascinating and interesting…and weird. It’s really hard to guess their behaviors accurately.” Of course, don’t be afraid of trial-and-error. When we learn to choose the right test, experimentation becomes the impartial mediator between what marketers don’t yet know and challenges our strongest assumptions.
Here are some experiment ideas to get you started.
Business To Business
Headline of white paper or guide
Email subject line
Blog post headline
Features and benefits of a product or service
Incentives or special offers
Email newsletter send time
Audience segmentation
Text ads in Google AdWords
Landing page elements such as imagery, headlines, forms, call-to-action and persuasive copy
Call-to-action on creative
Newsletter subscription copy or call-to-action
Newsletter subscription placement on website
Landing page form
Including certification badges on landing page
Including customer testimonials on landing page
Business To Consumer
Special offers or incentives
Call-to-action on creative
Email newsletter subject line
Landing page form
Ads featuring people v. product
Features/benefits of product or service
Customer testimonials on or off landing page
Before and after pictures
Headlines on landing pages, blog posts or free resources
Email newsletter send time
Audience segmentation and lists
Text ads in Google AdWords
Landing page elements such as imagery, headlines, forms, call-to-action and persuasive copy
Newsletter subscription copy or call-to-action
Newsletter subscription placement on website
Including certification badges on landing page
Including customer testimonials on landing page
Leave a Reply