A high risk does not establish a good recovery opportunity.
Investigate the likely obstacle, remaining engagement, and the time available. The right response depends on the situation.
Guides & product knowledge
Understand the system.
Make better
retention decisions.
Choose the question you are trying to answer. Explore the product, prepare a connection, understand the numbers, or review how customer data is handled.
THE PRODUCT / START HEREHow does a behavior
What makes an
What does the AI
How is customer
How does a behavior
become a decision?
Follow the five stages from an approved connection to an observed outcome.
Read the product guide ↗YOUR ENGINEERING TEAMWhat makes an
integration useful?
Prepare identity, meaningful activity, complete coverage, and approval boundaries.
Open the developer guide ↗THE INTELLIGENCEWhat does the AI
actually learn?
Explore product meaning, personal patterns, and the role of later outcomes.
Meet the Brain Engine ↗SECURITY & DATAHow is customer
information protected?
Inspect source access, workspace isolation, sanitization, and model-use boundaries.
Review the safeguards ↗A shared language
Know what the numbers mean.
Know what to ask next.
A useful dashboard makes its terms understandable. These definitions help your team read the same result with the same expectations.
- Revenue at risk
- The revenue exposed to a supported loss estimate within a defined horizon. Read it with its currency, coverage, and calculation basis.
- Recoverable revenue
- An estimate of what an intervention could preserve. It needs evidence of impact beyond the underlying risk estimate.
- Meaningful behavior
- An approved activity that represents customer value, such as finishing useful work rather than opening a screen.
- Customer baseline
- The available personal pattern used to put current behavior in context. Short or incomplete history limits what it can support.
- First loss
- The first supported loss event within the modeled horizon, such as cancellation, downgrade, payment loss, non-renewal, or trial lapse.
- Source readiness
- Whether identity, mapping, freshness, coverage, and the required evidence support the next step.
- Available-at time
- When evidence became available to the system, which can differ from when the underlying event happened.
- Depth Dossier
- The deeper customer investigation: history, source evidence, feature changes, and context for a decision.
Three distinctions that matter
Better interpretation.
Better conversations.
Keep the evidence, the forecast, and the effect of an intervention separate when you decide how to help.
A renewal does not, by itself, prove the intervention caused it.
A renewal is an observed outcome. Demonstrating the effect of a response requires additional evidence.
A connected source does not guarantee enough learning history.
Setup can finish while individual histories and later outcomes are still developing. Readiness should stay visible.
Put the concepts to work
Open a customer.
Follow the evidence.
See how the terms connect to a practical investigation in the complete Revioli workspace.