For example, in order to increase sales of surfboards, we may determine that personalizing content, or recommending products, based on certain geographic locations is more likely to meet a customer’s needs.
To understand customer segmentation, there are a number of common metrics that can provide value. These may include, but are not limited to:
- Geographic location: Can be inferred from network location, or collected with permission based on device location.
- Referer: Can be identified by search engine terms, external websites addresses, campaigns, advertisements or newsletters.
- Browsing history: When allowed, tracking cookies or other unique identifiers can help to identify browsing trends, or on-site browsing history can be easily inferred.
- Profile information: With a connected social account or known account information, we can use specific personally identifiable information to determine preferences.
- Duration: The amount of time that a user spends on a specific page, or overall session duration.
- Page views: Each individual page request from a browser.
- Events: Custom actions tracked on a web page or from an application, such as clicking a specific button, or expanding an accordion.
In order to validate and optimize the personalization of content for a particular segment, it is important to test each interaction through collected analytics against the desired conversion.
- A/B or multivariate testing: With a large dataset, we can determine segmentation profile preferences for multiple alternative content options.
- Heatmaps: Mouse or trackpad movements, or touch-based tracking can help to identify the most popular areas of a particular page that a user might interact with.