In data-driven marketing, precision and trust are essential. Marketers have limited budgets and need to have confidence in their segment before activating. We are introducing a new data previewing capability that allows you to visually validate segments.
Addressing uncertainty in Segmentation
A marketing manager crafting a highly targeted segment for an upcoming campaign will need to know if the filters defined truly represent the right audience for varying reasons: first, delivering an irrelevant message to customers can erode brand trust. Second, misallocating the marketing budget can result in wasted resources that could be deployed more effectively elsewhere.
Previously, users relied mainly on segment count to understand a segment’s potential impact before activation. While helpful, this often wasn’t enough. The alternative was to publish the segment and then analyze the results in a tool like Data Explorer. However, by then, publishing costs had already been incurred, and time had elapsed. The marketing manager lacked a native way to visually inspect a segment’s output before committing to it. This uncertainty frequently resulted in reduced trust in segmentation: a perception that the segment outputs were not easily verifiable.
Hence, a preview capability is essential to trust in the segment results by confirming they are targeting the right audience.
The Solution: Data Cloud Segmentation Preview
Data Cloud’s new preview functionality is designed to instill confidence and enable faster, more informed decisions. It allows business users to visualize a sample of their segment’s output as well as some data-level insights directly within the Segmentation UI before publishing or activating it.
How it Works: A Use Case
Let’s consider a marketing manager building a segment for a promotional email campaign targeting customers.
- Building the Segment: The manager defines the rules in the Data Cloud segmentation builder.
- Using the Preview: after getting the segment population count, the manager wants to review the data output, and clicks the new “Preview” button.
- Instant Insight with Preview: A modal window appears, showcasing the preview interface, powered by a new component called Data lens, which will be integrated across Data Cloud. This interface visualizes sampled data with :
- Profile Pane: Visualizes column-level summaries. Allows you to quickly see distributions and outliers, missing, null and unique values, and data types. You can brush data to highlight connections between related values. For example, when our manager clicks on a particular city in California, they can see how that value is distributed across other columns in the dataset.
- Data Grid: Displays row-level details for the individuals who qualify for the segment, and allows you to sanity check the segment composition, like confirming key customers are included. The table will show up to 1,000 rows and display attributes from the Segment On entity including individual names, IDs, and other direct attributes.
- Refine and Confirm: If the manager spots anything unexpected, he can go back to the segment rules, make adjustments, and rerun the preview. This iterative process ensures the segment is precisely tuned to his needs.
Key Benefits of Previewing
- Boosts Confidence and Informed Decisions: The marketing manager can confirm they’re targeting the right audience, and correct potential targeting errors upfront, minimizing the risk of irrelevant communications.
- Maximizes ROI, by ensuring the right profiles are included to maximize campaign results.
Technical Highlights
- Pre-Publish Execution: Previews run before segment publication, allowing for immediate feedback and iteration.
- Data Governance: Preview tables respect data governance requirements, ensuring data masking and compliance.
- Deterministic Results: Once generated, preview results are cached to ensure consistency unless segment rules change.
Looking Ahead
The Data Cloud Segmentation Preview is a significant step towards empowering business users with greater control and confidence. Future enhancements potentially include displaying related attributes used in segment qualification and further integration into the activation flow, ensuring a seamless and fully verifiable segmentation experience.
We are always looking for ways to enhance this feature. Please share your feedback in the comment section on how we can make it even better!
Written by
