01
Observe
Capture relevant customer signals with their source, context, and appropriate permissions.
What actually happened?
Conceptual model · Version 0.1
From signals to outcomes
A proposed five-stage cycle for developing better-supported interpretations of customer goals and using them to inform decisions.
Research status: This is an exploratory synthesis, not a scientifically validated model or a claim that its component practices are new. Empirical testing and comparison with existing scholarship remain necessary.
Behavior describes an action. Prediction estimates a possible outcome. Intent concerns the customer’s goal. A responsible decision respects the limits of each.
Five stages
Observe → Interpret → Validate → Decide → Measure
02
Interpret
Consider competing hypotheses about the customer’s goal. Keep observations separate from explanations.
What might the customer be trying to accomplish?
03
Validate
Compare evidence, seek feedback, examine contradictions, and assess uncertainty and bias.
What supports or challenges the interpretation?
04
Decide
Choose a proportionate action, assign ownership, and require human review when warranted.
What should change, and who is accountable?
05
Measure
Evaluate whether the intervention helped the customer accomplish their intended goal.
Did the decision improve the outcome?
Hypothetical application
One behavior. Several possible goals.
A customer repeatedly visits a payment-assistance page but does not submit an application. This example is hypothetical, not a finding from an employer or customer dataset.
| Stage | Possible application |
|---|---|
| Observe | Determine whether the customer actually started an application. |
| Interpret | Consider eligibility research, document gathering, comparison, or application friction. |
| Validate | Examine relevant site search, feedback, support themes, and user research. |
| Decide | If supported, clarify eligibility guidance or simplify the difficult step. |
| Measure | Test whether customers find answers or complete intended tasks more easily. |
Enterprise safeguards
Evidence without overconfidence
Data quality
Trace source, completeness, and limitations.
Responsible AI
Separate generated interpretations from validated findings.
Governance
Define access, oversight, escalation, and decision rights.
Measurement
Distinguish changes in metric definitions from actual customer improvements.
Framework Documentation
The Customer Intent Intelligence Framework
A downloadable reference describing the proposed approach, its stages, and its practical applications.
Download the frameworkRelationship to the broader work
CII, ECII, and ECIIS
Customer Intent Intelligence is the public-facing research focus. Enterprise Customer Intent Intelligence is the broader scholarly and professional specialty concerning organizational application at scale. ECIIS is a separately developed enterprise intelligence system; this conceptual framework does not replace its existing architecture or disclose internal workplace implementation.
The precise relationship between these approaches is an area for future articulation and validation.
Read the CII Manifesto → Return to the CII Hub