# Social Listening Insight Log

Turn observed public conversations into bounded, sourced insights, competing explanations, open questions, ethical research notes, and responsible content actions.

## Usage note

Use this log for ethically gathered material your organization is permitted to access and analyze. Public visibility does not remove privacy, consent, platform, cultural, or safety concerns. Minimize personal data, avoid profiling individuals, and do not quote identifiable people outside their expected context without a defensible basis. A loud or highly engaged conversation is evidence of that observed conversation, not automatic evidence of population-wide demand.

## How to use this template

1. Define the content decision, listening question, platforms, communities, dates, languages, access basis, exclusions, and privacy controls.
2. Sample relevant conversations consistently and record de-identified observations, context, counterexamples, event effects, and source limitations.
3. Form a bounded insight that clearly separates observed evidence from interpretation, assumptions, and unknown audience coverage.
4. Compare alternative explanations and select proportionate validation through research, operational data, or a reversible content change.
5. Assign an action and owner, monitor outcomes and harms, then retain or delete the underlying material under the approved rule.

## Blank template

### Listening frame

- **Decision to inform:** [Specific content or research decision]
- **Listening question:** [What you are trying to learn]
- **Platforms or communities:** [Named scope]
- **Period and time zone:** [Start and end]
- **Languages or locales:** [Included scope]
- **Collection method:** [Manual, approved tool, or another]
- **Access and ethical basis:** [Why observation is appropriate]
- **Exclusions:** [Private spaces, sensitive topics, minors, or another]
- **Known coverage gaps:** [Who or what is missing]

### Observation record

- **Observation ID:** [Non-identifying reference]
- **Situation or trigger:** [Event and context]
- **Question, barrier, or language:** [Paraphrased]
- **Interaction pattern:** [Reply, correction, repetition, disagreement, or another]
- **Source type:** [Original, quote, promotion, automation, or unclear]
- **Counterexample:** [Contradictory evidence]
- **Potential sensitivity:** [Privacy or safety concern]
- **Analyst note:** [Interpretation clearly labeled]
- **Evidence location:** [Restricted record if retained]

### Insight and validation

- **Bounded insight:** [Among observed material, what appears true]
- **Supporting observations:** [IDs or summary]
- **Contradicting observations:** [IDs or summary]
- **Alternative explanation:** [Event, sampling, algorithm, campaign, or another]
- **Confidence:** [Low, medium, or high with reason]
- **What remains unknown:** [Gap]
- **Validation method:** [Interview, analytics, support data, test, or another]
- **Decision threshold:** [What would change the conclusion]
- [ ] The insight does not claim population representativeness without evidence.
- [ ] No sensitive trait or intent has been inferred from a fragment.
- [ ] Publication examples are de-identified or separately approved.

### Action and stewardship

- **Proposed content action:** [Improve, create, test, research, monitor, or none]
- **Expected reader benefit:** [Specific value]
- **Potential harm:** [Targeting, exposure, stereotyping, or another]
- **Owner and due date:** [Role and date]
- **Outcome measure:** [Useful signal]
- **Recheck date:** [When context may change]
- **Retention rule:** [Delete or preserve with date]
- **Learning recorded:** [Result and next question]