Marketing folks, especially analysts and content developers, care deeply for the coveted A/B split test. Why? A/B testing provides a benchmark, establishing a stake in the ground as measured by the “control” and the effectiveness (or ineffectiveness) of an additional variable introduced into the mix. Marketers have been using A/B split testing for decades to measure, among other things, the impact that the variable has on open rate, response rate, click through rate, and other key measures.
At times, people will ask me how the Visible Intelligence platform can be used to fuel A/B split testing decisions, or just as importantly, how to measure the effectiveness of the test relative to the population or audience receiving the message. Here are three ways to do it:
1. Assess Share Rates of Owned Content
There’s a function within the Visible platform that enables users to segment by content classification, which can be used to quickly determine how many times owned content from Twitter is shared, mentioned or retweeted. We can use these metrics when examining different techniques for posting on Twitter, that when looked at cumulatively, can yield some extremely interesting findings. For instance, I was recently analyzing about 3 months of Twitter content for a client and discovered that when the client used statistics in the header of their tweets, the content was shared 3x more frequently than content without statistics in the header.
2. Analyze Data Through the Lens of Specific Audiences
Just as two messages can be tested across one specific audience, two (or more) audiences can be tested for how they respond to a specific offer or interact with a campaign. B2B marketers often use SIC codes to pick specific types of businesses to test messaging with, whereas consumer brands will often determine the audience based on their demographic information, or how they identify themselves, i.e. moms vs. sports enthusiasts, vs. parents. There are a few techniques that can be used within the Visible platform to uncover these audiences. One is by creating a custom segmentation of sites that cater to a specific audience group, like moms, and layering that site list over a defined content universe, like people talking about Huggies vs. Pampers. Another technique is to develop a segmentation for audience groups who self-identify. For instance, a parent might say things like “my kids,” or “my son,” or “my daughter,” or “our children.” Users can then layer either of these segmentations over a defined content universe and measure the frequency of discussion about specific topics, like a new product launch, campaign or special offer. Analyzing data from specific audience groups, just as analyzing how content is shared, enables marketers to glean all sorts of insights that can fuel new ideas for testing.
Moms vs. Sports Enthusiasts, RedBull Campaign
3. Topic Discovery
I can’t count how many times Visible’s Topic Discovery tool has incited an “Aha” moment from our clients and prospects. In their own words, they’ve often referred to it as a “strategic differentiator.” Topic Discovery allows users to examine the prominent terms and themes associated with two searches, side-by-side. It shows the terms unique to one search over the other, and the overlap that exists between the two. Many of our clients use this tool to fuel their content strategies because they are quickly able to see the rising keyword trends that can then be analyzed and introduced into new messaging. The implications for those looking for data-driven ideas to enhance their A/B testing is limitless. What’s more, users can easily compare how people are discussing a new offer on Facebook vs. Twitter, or any two content channels, which then empowers marketers to refine their messaging strategies for the specific environments in which they’re deploying content. It’s also a fantastic tool for SEO because users can effectively detect rising keyword trends from target audience groups and introduce these keywords into their organic search programs.
JCrew
It’s not just about measuring the effectiveness of the test, it’s about getting smarter about ways to use social data to incorporate new ideas into your testing methodologies. In addition to using social data to fuel social messaging, the insights gleaned can be used and tested in a variety of channels. With so much data at our fingertips, marketers have never been in a better position to conduct data-driven tests.
For more “Heard on the street” posts click here >>
