CRM Analytics

(CRA)

The reporting layer on top of your customer data: pipeline health, forecast accuracy, cohort retention, churn signals and customer lifetime value. It reads the same records the CRM writes, so there is no export and no lag.

Pipeline Forecasting Cohorts Churn LTV Attribution

Key Features

The reporting layer on top of your customer data: pipeline health, forecast accuracy, cohort retention, churn signals and customer lifetime value.

Pipeline Analytics

Conversion by stage, time in stage, and where deals actually stall. Win and loss reasons aggregated instead of buried in notes.

Forecasting

Weighted and committed forecasts by rep, team and period, measured against what actually closed.

Cohort Retention

Group customers by acquisition period, channel or segment and watch how each cohort behaves over time.

Churn Signals

Rules and scores that flag accounts going quiet before they leave, based on order cadence and engagement.

Customer Lifetime Value

LTV per customer, segment and channel, built from real order history rather than an assumed average.

Campaign Attribution

Tie revenue back to the campaign, source and territory that produced it, across every channel the customer used.

Kubrik OS — agent layer

CRM Analytics, assisted and operated by AI

Analytics is where the AI stops being a chart and starts being an answer — with the transactions that produced it attached.

AI-assisted

What the AI advises on

Available on every plan, Starter included.

Answers questions in plain language against pipeline, cohorts, churn and lifetime value
Explains why a metric moved, tracing it back to the individual transactions that caused it
Flags forecast drift and cohorts decaying faster than their peers

AI-operated

What the AI does under governance

Professional and Enterprise. Starter runs the AI in advisory mode only.

Builds and schedules recurring reports, and delivers them to the people who act on them
Maintains segment and cohort definitions as the underlying behaviour changes
Raises an alert when a metric breaches its threshold, with the diagnosis already attached

Governed tools an agent can call in this module

answerAnalyticalQuery() buildReport() scheduleReport() defineCohort() raiseMetricAlert()

Autonomy is a setting, not a promise — choose the rung per role and per module

Suggest

The agent proposes. Nothing changes until a person acts on it.

Prepare

The agent builds the transaction and leaves it waiting for approval.

Execute below threshold

Acts on its own under a value or risk limit you set.

Execute by policy

Acts inside a written policy, every action fully audited.

Escalate

Anything outside the policy stops and goes to a person.

Same rules, either operator. An agent obeys the same permissions, validation and business rules a person does, and every action lands in the audit trail: who or what initiated it, the inputs, the rationale, the approvals, the resulting transaction and how to reverse it.