# Channel Purpose Guide

**Object type:** LPM Knowledge Object  
**Primary LPM layers:** Communication Architecture, Information Ecology, Governance Architecture, AI Amplification  
**Connected layers:** Decision Architecture, Ownership Map, Platform Structure, Identity & Incentives  
**Version:** 1.0  
**Owner:** LPM / Lapemo  
**Last reviewed:** 2026-06-24

## Purpose

The Channel Purpose Guide prevents the enterprise from using every channel for every kind of communication. It defines the official purpose, boundaries, owners, source-of-truth links, AI rules, retention expectations, and retirement triggers for each major communication channel.

For companies scaling AI, this matters because AI will amplify unclear channels, stale summaries, hidden decisions, duplicate truth, and unmanaged communication paths.

## What this object produces

- A channel taxonomy across chat, email, meetings, documents, dashboards, work systems, GRC, CRM, and AI-assisted channels.
- A clear rule for where decisions, approvals, risks, evidence, commitments, and execution records belong.
- A reusable checklist for channel health, drift, ownership, source-of-truth integrity, AI safety, and record retention.
- A machine-readable object that can be ingested into Lapemo and rendered as a website template, workshop worksheet, guided skill, or in-app module.

## Core principles

| Principle | Meaning |
|---|---|
| Every channel needs a job | Teams, Slack, email, meetings, dashboards, docs, Jira, CRM, GRC, and AI tools should not all be used for everything. |
| Chat is not memory | Fast discussion can happen in chat, but durable decisions, owners, evidence, and commitments must move into a system of record. |
| Email is not workflow | Email can notify, summarize, or approve when no workflow exists, but it should not become the hidden operating model. |
| Meetings need output | A meeting that does not create a decision, escalation, alignment, artifact, or next action should be redesigned or retired. |
| Dashboards are evidence, not interpretation | Metrics must have owners, definitions, freshness windows, and links to decision or governance context. |
| AI summaries are provisional | AI can summarize, route, classify, and draft, but material communications require source links, confidence, and human review rules. |
| Official records need owners | Each channel-purpose rule should identify who maintains it, when it is reviewed, and where exceptions are captured. |

## Knowledge object field model

| Field | Definition | Required |
|---|---|---|
| Channel ID | Unique identifier for the channel, forum, tool, workspace, group, distribution list, dashboard, agent, or knowledge base | Yes |
| Channel name | Human-readable name of the communication channel or forum | Yes |
| Channel type | Chat, email, meeting, work system, documentation, dashboard, GRC, service desk, CRM, AI tool, broadcast, or executive forum | Yes |
| Primary purpose | The one job this channel exists to perform | Yes |
| Allowed use | What the channel is approved to carry | Yes |
| Disallowed use | What must not live only in this channel | Yes |
| Primary audience | The people, teams, roles, vendors, executives, customers, or AI agents who use the channel | Yes |
| Accountable owner | Person or role accountable for channel health, accuracy, access, rules, and retirement decisions | Yes |
| Decision boundary | Whether the channel can discuss, recommend, approve, announce, or only reference decisions | Required when decision-related |
| Source of truth | Official system, artifact, decision log, evidence pack, dashboard, ticket, or policy that backs the channel | Yes |
| Record retention | Where records are stored, for how long, and under what privacy or regulatory rule | Required when material |
| Response expectation | Expected action: inform, acknowledge, discuss, decide, approve, execute, escalate, archive, or no response | Yes |
| SLA / cadence | Expected response window or communication frequency | Required when action-bearing |
| Escalation path | Where unresolved, high-risk, blocked, or time-sensitive items move next | Required when material |
| Sensitivity level | Public, internal, confidential, restricted, regulated, customer-sensitive, employee-sensitive, legal-sensitive | Yes |
| AI involvement | None, summarized, drafted, routed, classified, recommended, sent, or executed by AI | Yes |
| Human review rule | When a human must approve, validate, or correct the communication before use | Required when AI-assisted or material |
| Access rule | Who can read, post, approve, export, automate, or archive content in the channel | Yes |
| Retirement trigger | Condition under which the channel should be merged, archived, renamed, replaced, or deleted | Yes |
| Review date | Date the channel-purpose rule must be reviewed | Yes |
| Version | Artifact version, owner, last reviewed date, and change history | Yes |

## Channel purpose matrix

| Channel | Primary purpose | Allowed use | Disallowed use | System of record | AI rule |
|---|---|---|---|---|---|
| Teams / Slack chat | Fast collaboration and working-team coordination | Questions, triage, working notes, links, lightweight updates | Final decisions, official approvals, durable ownership, evidence of record | Decision Log, Jira, Confluence, GRC, CRM | AI summary allowed only when source-linked and owner-reviewed for material items |
| Email | Formal notification and external communication | Executive summaries, stakeholder notification, external messages, formal follow-up | Work tracking, hidden approvals, unresolved decision debates, long-term knowledge | CRM, Decision Log, document repository, workflow tool | AI drafting allowed. Human review required before sending material or sensitive messages |
| Meetings | Sense-making, tradeoff discussion, escalation, and decision forums | Debate, alignment, evidence review, decision readiness, issue resolution | Unstructured status readouts with no output artifact | Meeting notes, Decision Log, work system, escalation record | AI notes are draft only until owner validates decision, action, and evidence links |
| Jira / work system | Execution state and delivery accountability | Backlog, dependencies, blockers, owners, dates, acceptance criteria | Strategic decision rationale without decision log, undocumented governance exceptions | Jira, Azure DevOps, Asana, Monday, ServiceNow | AI classification and routing allowed. Human owner approves priority and commitments |
| Confluence / SharePoint / Wiki | Durable knowledge and operating memory | Policies, playbooks, requirements, architecture, standards, onboarding, decision context | Unowned stale pages, duplicate sources of truth, private notes posing as official guidance | Knowledge base with owner and review date | AI generated content must show owner, source links, and review date |
| Dashboards / BI | Evidence, trend visibility, and performance monitoring | Metrics, KPIs, adoption signals, risks, customer and financial trends | Unexplained interpretations, metric definitions hidden from users, stale snapshots | BI platform, data catalog, metric registry | AI insights must show source, freshness, confidence, and metric owner |
| GRC / risk tool | Control, audit, exception, policy, and risk records | Risks, controls, exceptions, approvals, audits, regulatory records | Casual discussion, unresolved debate, workarounds with no owner | GRC platform, audit system, risk register | AI can classify and summarize. Human control owner approves evidence and disposition |
| CRM / customer system | Customer commitments and market-facing record | Customer issues, account commitments, sales context, success plans, renewals | Internal-only decision debate, unsupported promises, undocumented AI commitments | CRM and customer support platform | AI drafts require human approval when customer-facing or commitment-bearing |
| AI assistant / agent channel | AI-enabled drafting, routing, summarization, classification, and task support | Summaries, first drafts, pattern detection, routing suggestions, knowledge retrieval | Final authority, unreviewed sensitive communication, autonomous policy exceptions | AI Owner Register, Decision Log, source systems | Must expose source, confidence, human owner, and action boundary |

## Communication class routing

| Communication class | Purpose | Preferred channel | Minimum required fields |
|---|---|---|---|
| Awareness | People need to know, but no action is required | Announcement channel, email, knowledge hub | Owner, summary, source link, retention rule |
| Input request | A person or team must provide context, feedback, or evidence | Work system, form, team channel, meeting pre-read | Request owner, deadline, required input, response path |
| Decision request | A decision owner must choose between options or approve direction | Decision forum, decision workflow, executive channel | Decision type, options, evidence, rights owner, due date |
| Decision outcome | A decision has been made and downstream teams must align | Decision Log, announcement channel, work system link | Decision, owner, rationale, affected teams, effective date |
| Escalation | A blocker, risk, exception, or conflict exceeds local authority | Escalation Map, governance forum, executive path | Severity, impact, attempted resolution, required decision |
| Evidence update | Metric, source, dashboard, or proof point changed | Dashboard, evidence checklist, knowledge base | Evidence owner, freshness, confidence, affected decisions |
| Incident / control event | Operational, compliance, customer, security, or AI event needs response | Incident tool, GRC, war room, executive alert | Severity, owner, timeline, control impact, closure record |
| AI-generated signal | AI detects, drafts, summarizes, recommends, routes, or executes something | AI control channel, source record, owner review queue | Source, confidence, risk tier, human reviewer, action boundary |

## Version A: 500+ employee company

The 500+ version is designed for a scaling company that has outgrown founder-led communication but does not need enterprise bureaucracy.

| Dimension | Design rule |
|---|---|
| Design intent | Keep the channel map simple enough to use every day. Prevent founder-led, relationship-based communication from becoming hidden infrastructure. |
| Core channels | Teams or Slack, email, a work system, a knowledge base, a dashboard, and one leadership decision forum. |
| Biggest risk | Decisions and ownership live in chats, meetings, and people heads instead of durable records. |
| Minimum rule | No material decision, priority change, customer commitment, or AI action is official until it is recorded in the right system. |
| Operating pattern | One-page channel guide, monthly review, clear owner for each channel, and simple escalation triggers. |
| AI focus | AI may summarize meetings and chats, but the human owner must validate decisions, actions, and commitments. |

## Version B: 5,000+ employee company

The 5,000+ version is designed for functional maturity with cross-functional friction.

| Dimension | Design rule |
|---|---|
| Design intent | Standardize channel purpose across functions so information does not fracture by department, tool preference, or leadership style. |
| Core channels | Function-specific work systems, enterprise knowledge base, portfolio dashboard, GRC, CRM, service desk, executive forums, AI owner register. |
| Biggest risk | Each function has its own communication logic, causing duplicate status, conflicting truth, and slow escalation. |
| Minimum rule | Every major communication type must have a primary channel, system of record, owner, escalation path, and review cadence. |
| Operating pattern | Channel taxonomy by function and communication type, quarterly governance review, and clear handoff rules between systems. |
| AI focus | AI routing and summarization must respect sensitivity, record retention, source of truth, and human approval boundaries. |

## Version C: 10,000+ employee company

The 10,000+ version is designed for a complex enterprise with regions, business units, regulated records, many platforms, and emerging human plus AI workflows.

| Dimension | Design rule |
|---|---|
| Design intent | Govern communication as enterprise infrastructure across regions, business units, vendors, regulated records, and AI-enabled workflows. |
| Core channels | Enterprise collaboration suite, records management, GRC, portfolio systems, data catalog, CRM, HRIS, service management, AI control plane. |
| Biggest risk | Communication volume overwhelms judgment, AI amplifies stale or unapproved messages, and official truth becomes fragmented across the enterprise. |
| Minimum rule | Channel purpose must be policy-backed, system-mapped, role-owned, retention-aware, sensitivity-aware, and monitored for drift. |
| Operating pattern | Enterprise channel taxonomy, automated stale-channel detection, metadata governance, AI controls, and executive visibility into communication health. |
| AI focus | AI can route, summarize, detect gaps, and draft communications, but cannot create final authority without governed human review and source-backed records. |

## Scoring model

| Score | Channel-purpose maturity |
|---|---|
| 0 | Channels are informal, duplicated, personality-driven, and not connected to official records or ownership. |
| 1 | Some channels have understood purposes, but rules are tribal, inconsistent, and not maintained. |
| 2 | Primary channels are documented, but decision boundaries, AI rules, retention, and escalation paths are incomplete. |
| 3 | Channels have owners, clear allowed uses, source-of-truth links, response expectations, and review cadence. |
| 4 | Channel purpose is governed as a live operating layer with AI-assisted routing, stale-channel detection, evidence links, and executive communication-health visibility. |

Suggested scoring: score each channel from 0 to 4 across purpose clarity, source-of-truth link, owner clarity, decision boundary, AI rule, sensitivity handling, retention rule, escalation path, and review cadence. Average across channels to generate a communication architecture maturity score.

## Validation rules

| Rule | Validation test |
|---|---|
| One-purpose rule | Each channel must have a primary purpose. If it has more than three unrelated purposes, split or redefine it. |
| Record rule | Material decisions, approvals, risks, commitments, and control events cannot live only in chat, email, or meeting notes. |
| Owner rule | Every official channel must have an accountable human owner and a review date. |
| AI rule | AI-generated communication must identify source, confidence, human reviewer, and whether it is draft, recommendation, or approved output. |
| Retention rule | Sensitive, regulated, legal, customer, employee, financial, or control-bearing communication must follow retention and access rules. |
| Escalation rule | Channels carrying blockers, incidents, risks, or dependency failures must link to an escalation path. |
| Retirement rule | A channel without owner, usage, purpose, or record value should be archived, merged, or deleted. |

## Lapemo mapping rules

| Lapemo field | Meaning |
|---|---|
| channel_id | Unique channel or forum identifier |
| channel_type | Chat, email, meeting, work system, knowledge base, dashboard, GRC, CRM, AI tool, broadcast, executive forum |
| primary_purpose | Main purpose of the channel |
| allowed_use | Permitted message types and communication behaviors |
| disallowed_use | Content that must move to another artifact or system |
| source_of_truth | System or artifact that carries durable truth |
| accountable_owner | Role accountable for channel quality and governance |
| decision_boundary | Discuss, recommend, approve, announce, reference, or none |
| ai_involvement | None, summarized, drafted, routed, classified, recommended, sent, executed |
| human_review_rule | Required review rule based on materiality and sensitivity |
| sensitivity_level | Public, internal, confidential, restricted, regulated, customer-sensitive, employee-sensitive, legal-sensitive |
| retention_rule | Archive and retention requirement |
| escalation_path | Linked escalation route |
| retirement_trigger | When the channel should be archived, merged, or redesigned |
| review_date | Date the purpose rule should be refreshed |

## AI prompts for guided use

| Prompt type | Reusable prompt |
|---|---|
| Classify channel purpose | Review this list of communication channels and classify each by primary purpose, allowed use, disallowed use, owner, source of truth, decision boundary, AI involvement, sensitivity, and review date. |
| Detect channel drift | Identify channels where the stated purpose no longer matches actual usage. Flag channels carrying decisions, approvals, risks, or commitments without links to official records. |
| Recommend channel consolidation | Find duplicate, stale, low-value, or overlapping channels and recommend which should be merged, archived, renamed, or governed. |
| Validate AI-safe communication | Check whether AI-generated summaries, drafts, or routing recommendations include source links, confidence, human review rules, sensitivity level, and action boundary. |
| Generate company-specific guide | Using company size, tool stack, regulatory needs, and operating model maturity, generate a channel purpose guide with 500+, 5,000+, or 10,000+ scale rules. |

## Workshop flow

1. Inventory the top 20 to 50 channels, forums, dashboards, work systems, and AI communication paths used by the company.
2. Classify each channel by primary purpose, allowed use, disallowed use, owner, and source of truth.
3. Mark where decisions, approvals, risks, commitments, or evidence currently live in the wrong place.
4. Define the official record for each material communication type.
5. Set AI review rules for summaries, meeting notes, customer communication, executive updates, routing, and automated action.
6. Assign review dates and retirement triggers for duplicate, stale, unmanaged, or low-value channels.
7. Load the results into Lapemo as channel-purpose records linked to owners, systems, decision logs, evidence, and escalation paths.

## Output artifacts

- Channel purpose guide for website download.
- Channel taxonomy for company onboarding.
- Communication rule set for Lapemo ingestion.
- AI-safe communication rules for summaries, routing, drafting, and agent-triggered messages.
- Channel retirement backlog for duplicate, stale, unmanaged, or low-value channels.
- Executive communication health snapshot.

## Usage note

This object should not be treated as a one-time document. It should be versioned, reviewed, and connected to systems of record. The future-state Lapemo skill should flag stale channels, unmanaged AI summaries, missing owners, decisions hiding in chat, duplicate sources of truth, and channels that no longer match their purpose.