Choose standard vs. real-time calculated insights
A calculated insight (CI) is a metric Data Cloud derives from your data, like total spend or a loyalty score. Data Cloud offers two CI types — Standard and Real-Time — and recommenders, segments, and Personalization decisions read the result. This page compares the two and helps you pick one. It assumes you’ve already seen the Calculated insights overview page.
Compare the two CI types
| Standard CI | Real-Time CI | |
|---|---|---|
| Computation | Runs as a scheduled batch job | Computed on-the-fly, as data arrives |
| Refresh timing | On a recurrence you set (commonly daily) | Continuously, with no fixed schedule |
| Best for | Durable metrics reused across segments, reports, or scoring models (lifetime value, churn score, loyalty tier) | Live personalization decisions that need the newest value (last-session activity, in-session scoring) |
| Historical scope | Can join full historical and profile data | Narrower — tuned for freshness over deep history |
Pick a CI type
- From Data 360, open Calculated Insights for your data space.
- Start a new insight and give it a name that describes the metric.
- Set its type to Standard if the metric can refresh on a schedule, or Real-Time if a personalization decision needs the newest value.
- Define the metric’s logic (for example, a total, count, or score) over the fields it needs.
- Save and run it, then confirm the output values look right for a few sample profiles.
Discussion
Comments are not configured for local development. They appear once the site owner connects GitHub Discussions (Giscus).