The Modern Marketing Blog covers the latest in marketing strategy, technology, and innovation.

Getting Segmentation Right

Advancements in technology and data availability have rendered the one-size-fits-all approach to online marketing obsolete. Companies who are able to identify key characteristics of prospects or buyers and tailor website content to meet their needs have a competitive advantage and a more profitable digital experience.

Although historically it was difficult to drill down on individual user characteristics, today’s growing digital space has made this easy. Optimization tools allows marketers to segment their users and navigate away from the traditional approach and navigate towards a dynamic website that caters to each individual’s wants and needs.

If a random survey of users were asked to name their favorite baseball team, the answers will likely vary evenly among the 30 Major League Baseball teams. For this very reason, marketers have steered away from showing user-specific content on their sites. No marketer would want a Yankee to see a Red Sox banner on their homepage. Segment discovery, such as the kind Maxymiser clients achieve with our optimization solution, learns about each user based on personalization criteria (user attributes determined by a variety of insightful data) and shows the most relevant content, ensuring that Boston fans see a Red Sox banner every time. Segmentation and targeting are very powerful tools that can create unique digital experiences for every user.

There are many factors that influence the success of a segment discovery campaign - either with Maxymiser or another optimization solution. When a testing opportunity materializes, it is important to meet all of the requirements below to ensure the results are successful and applicable.

1. Difference Within Experiences

All experiences (different versions of a given webpage, funnel, landing page, mobile site, etc.) tested must be different enough to facilitate a particular group of visitors to always behave similarly. Interestingly enough, the most successful segment discovery campaigns have been product based rather than messaging based.

Bad Example

In this instance, an insurance company is updating the banner on the homepage. The default experience (or the existing banner) reads, “Switch and Save an Average of $625.” The company creates two “challenger” experiences (alternate banners) the first reading, “Users who Switch Save” and the second, “Switch Today and Start Saving Tomorrow.”

Good Example

In this case, the company has decided to default to different product offerings based on price. Their default experience is basic insurance coverage (bare minimum coverage), and the two challengers default to medium insurance coverage and stronger insurance coverage (the most expensive), respectively.

Understanding the Difference

The takeaway here is as follows: it is difficult to justify with the poor example that a group of users would choose a particular savings message every time, whereas it is easy to build a persona that is shopping for bare minimum auto insurance.

2. Meaningful Personalization Criteria

What we call personalization criteria, but can be understood as user characteristics identified through insightful data, must directly correlate with the hypothesis of the test and variant changes. In the example above, meaningful personalization criteria would build towards a persona and justify why a user should be defaulted to a different product offering. You might consider criteria such as vehicle make, vehicle year, prior policy value, or number of prior claims.

3. Traffic on Test Page

There must be enough traffic on the test page to allow for each personalization criteria value to have at least 100 actions. Since the comparison is done on the personalization criteria level, there must be adequate actions within each value to make a statistically valid comparison. A large number of variants and personalization criteria will cause a low action count within each personalization criteria value, making it difficult to reach statistical significance.

An Example of a Successful Segmentation Setup

This example uses an appropriate number of personalization criteria relative to website traffic whereby an adequate number of users will have the predefined characteristics.

Meeting the above criteria should result in a robust Segment Discovery report with many targetable micro-segments. With this tool a marketer should better understand behavioral differences and discover new opportunities to continuously improve the digital experience.
For more on segmentation and targeting, check out this blog post from the Maxymiser team!

Be the first to comment

Comments ( 0 )
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.