HubSpot Service Software cheat sheet

A ticket is a record of a customer request with a pipeline, stage, owner, priority, and source; the ticket pipeline stage replaces generic status labels and drives reporting.

Conversations inbox aggregates email, chat, and form channels into one thread view; replies sent from the inbox are logged on the contact record automatically.

Ticket creation sources include manual creation, form submissions, chat and chatbot flows, workflow automation, inbound email to a connected inbox, and API/CRM integrations.

Ticket association rules: a ticket can be associated with contacts, companies, deals, and other tickets; associations determine which records appear in the ticket sidebar and in cross-object reporting.

Inbox routing order is typically channel or inbox assignment first, then routing rules such as round robin or contact owner; unassigned conversations fall to the inbox's default assignee.

Ticket status vs. stage: stage reflects position in the pipeline lifecycle, while closed status is set when a ticket enters a closed stage; reopening moves it to an open stage and restores SLA clocks per settings.

SLA rules attach to tickets via pipeline and conditions; the two clocks are time to first response and time to close, and each has its own working-hours calendar.

Owner assignment can be manual, via workflow, via routing rule, or via round-robin in help desk; reassignment history is visible in the ticket activity log.

Shared inbox permissions are governed by inbox access settings plus user permissions; a user without ticket access can still see a conversation if they have inbox access.

Chatbot flow building blocks include welcome message, bot options (buttons), customer input capture, branching by contact property, handoff to a human, and ticket creation.

Handoff rules: a chatbot transfers to a live agent when a user selects a human option, when a keyword trigger fires, or when the bot cannot match an intent; the conversation then enters the inbox queue.

Knowledge base grounding lets the AI customer agent answer from published KB articles; unpublished or draft articles are excluded from the agent's answer set.

AI answer behavior is constrained by confidence thresholds; below threshold the agent escalates or offers a handoff rather than guessing, and conversations are logged for review.

Workflow automation for service uses ticket property enrollment triggers, delay steps, and actions such as assign owner, set priority, send email, or create a task.

Routing decision rules: round robin distributes evenly among eligible users, load balanced considers open ticket counts, and contact owner routes to the assigned owner when one exists.

Bot personalization uses contact properties and known conversation context; unknown visitors get generic flows until a property is captured or a contact is identified.

Automation testing rule: workflows and bot branches should be validated with test contacts before enabling, because enrollment criteria are evaluated at trigger time, not retroactively.

AI-assisted agent tools include suggested replies, conversation summaries, and content generation; outputs require agent review before sending unless auto-send is explicitly configured.

Knowledge base structure: knowledge base > category > subcategory > article; categories drive navigation and article URL paths.

Article lifecycle is draft, published, and scheduled; only published articles are visible on the public site and available to chatbots and AI agents.

Access control options are public (no login), private (registered visitors), and private with SSO; private content is gated by membership settings on the knowledge base.

SEO fields per article include title, meta description, URL slug, and canonical tags; category pages and article pages are indexed separately.

Article feedback and analytics track views, helpful/unhelpful votes, and search terms; high search-with-no-result terms indicate missing content.

Article templates (how-to, troubleshooting, FAQ) enforce consistent structure; template choice affects the default headings and body layout.

Multi-language knowledge bases create variant articles per language; each variant has its own publish state and URL.

Content reuse rule: articles can be linked inside chatbot flows and in ticket replies, but edits to a published article take effect immediately for all references.

Self-service deflection is measured by comparing KB article views and search success against ticket volume for the same topics.

Customer health scores are computed from weighted factors such as ticket volume, ticket sentiment, survey responses, product usage, and support response times; each factor has a configurable weight.

Health score grades map numeric ranges to labels (for example healthy, neutral, at risk); thresholds are configurable and should be documented before reporting.

Churn and retention reporting uses cohort analysis: group contacts or companies by start period, then track retained percentage over subsequent periods.

Ticket reporting dimensions include create date, close date, pipeline, stage, owner, priority, source, and associated company; measures include count, average time to close, and average time to first response.

SLA reporting measures the percentage of tickets meeting each SLA target, split by pipeline and by team, and highlights breached versus at-risk tickets.

CSAT is measured via survey with a 1-5 or 1-7 scale; the standard metric is the percentage of positive responses, and response rate should be reported alongside it.

NPS uses the 0-10 scale; promoters are 9-10, passives 7-8, detractors 0-6, and NPS equals percent promoters minus percent detractors.

Dashboard reporting rule: filters applied at the report level override dashboard-level filters; conflicting filters produce confusing results and should be consolidated.

Retention insight sources include ticket sentiment trends, repeat-contact rates, escalation frequency, and survey verbatims; these are leading indicators relative to churn events.

A ticket is a record of a customer request with a pipeline, stage, owner, priority, and source; the ticket pipeline stage replaces generic status labels and drives reporting.

Conversations inbox aggregates email, chat, and form channels into one thread view; replies sent from the inbox are logged on the contact record automatically.

Ticket creation sources include manual creation, form submissions, chat and chatbot flows, workflow automation, inbound email to a connected inbox, and API/CRM integrations.

Ticket association rules: a ticket can be associated with contacts, companies, deals, and other tickets; associations determine which records appear in the ticket sidebar and in cross-object reporting.

Inbox routing order is typically channel or inbox assignment first, then routing rules such as round robin or contact owner; unassigned conversations fall to the inbox's default assignee.

Ticket status vs. stage: stage reflects position in the pipeline lifecycle, while closed status is set when a ticket enters a closed stage; reopening moves it to an open stage and restores SLA clocks per settings.

SLA rules attach to tickets via pipeline and conditions; the two clocks are time to first response and time to close, and each has its own working-hours calendar.

Owner assignment can be manual, via workflow, via routing rule, or via round-robin in help desk; reassignment history is visible in the ticket activity log.

Shared inbox permissions are governed by inbox access settings plus user permissions; a user without ticket access can still see a conversation if they have inbox access.

Chatbot flow building blocks include welcome message, bot options (buttons), customer input capture, branching by contact property, handoff to a human, and ticket creation.

Handoff rules: a chatbot transfers to a live agent when a user selects a human option, when a keyword trigger fires, or when the bot cannot match an intent; the conversation then enters the inbox queue.

Knowledge base grounding lets the AI customer agent answer from published KB articles; unpublished or draft articles are excluded from the agent's answer set.

AI answer behavior is constrained by confidence thresholds; below threshold the agent escalates or offers a handoff rather than guessing, and conversations are logged for review.

Workflow automation for service uses ticket property enrollment triggers, delay steps, and actions such as assign owner, set priority, send email, or create a task.

Routing decision rules: round robin distributes evenly among eligible users, load balanced considers open ticket counts, and contact owner routes to the assigned owner when one exists.

Bot personalization uses contact properties and known conversation context; unknown visitors get generic flows until a property is captured or a contact is identified.

Automation testing rule: workflows and bot branches should be validated with test contacts before enabling, because enrollment criteria are evaluated at trigger time, not retroactively.

AI-assisted agent tools include suggested replies, conversation summaries, and content generation; outputs require agent review before sending unless auto-send is explicitly configured.

Knowledge base structure: knowledge base > category > subcategory > article; categories drive navigation and article URL paths.

Article lifecycle is draft, published, and scheduled; only published articles are visible on the public site and available to chatbots and AI agents.

Access control options are public (no login), private (registered visitors), and private with SSO; private content is gated by membership settings on the knowledge base.

SEO fields per article include title, meta description, URL slug, and canonical tags; category pages and article pages are indexed separately.

Article feedback and analytics track views, helpful/unhelpful votes, and search terms; high search-with-no-result terms indicate missing content.

Article templates (how-to, troubleshooting, FAQ) enforce consistent structure; template choice affects the default headings and body layout.

Multi-language knowledge bases create variant articles per language; each variant has its own publish state and URL.

Content reuse rule: articles can be linked inside chatbot flows and in ticket replies, but edits to a published article take effect immediately for all references.

Self-service deflection is measured by comparing KB article views and search success against ticket volume for the same topics.

Customer health scores are computed from weighted factors such as ticket volume, ticket sentiment, survey responses, product usage, and support response times; each factor has a configurable weight.

Health score grades map numeric ranges to labels (for example healthy, neutral, at risk); thresholds are configurable and should be documented before reporting.

Churn and retention reporting uses cohort analysis: group contacts or companies by start period, then track retained percentage over subsequent periods.

Ticket reporting dimensions include create date, close date, pipeline, stage, owner, priority, source, and associated company; measures include count, average time to close, and average time to first response.

SLA reporting measures the percentage of tickets meeting each SLA target, split by pipeline and by team, and highlights breached versus at-risk tickets.

CSAT is measured via survey with a 1-5 or 1-7 scale; the standard metric is the percentage of positive responses, and response rate should be reported alongside it.

NPS uses the 0-10 scale; promoters are 9-10, passives 7-8, detractors 0-6, and NPS equals percent promoters minus percent detractors.

Dashboard reporting rule: filters applied at the report level override dashboard-level filters; conflicting filters produce confusing results and should be consolidated.

Retention insight sources include ticket sentiment trends, repeat-contact rates, escalation frequency, and survey verbatims; these are leading indicators relative to churn events.

A ticket is a record of a customer request with a pipeline, stage, owner, priority, and source; the ticket pipeline stage replaces generic status labels and drives reporting.

Conversations inbox aggregates email, chat, and form channels into one thread view; replies sent from the inbox are logged on the contact record automatically.

Ticket creation sources include manual creation, form submissions, chat and chatbot flows, workflow automation, inbound email to a connected inbox, and API/CRM integrations.

Ticket association rules: a ticket can be associated with contacts, companies, deals, and other tickets; associations determine which records appear in the ticket sidebar and in cross-object reporting.

Inbox routing order is typically channel or inbox assignment first, then routing rules such as round robin or contact owner; unassigned conversations fall to the inbox's default assignee.

Ticket status vs. stage: stage reflects position in the pipeline lifecycle, while closed status is set when a ticket enters a closed stage; reopening moves it to an open stage and restores SLA clocks per settings.

SLA rules attach to tickets via pipeline and conditions; the two clocks are time to first response and time to close, and each has its own working-hours calendar.

Owner assignment can be manual, via workflow, via routing rule, or via round-robin in help desk; reassignment history is visible in the ticket activity log.

Shared inbox permissions are governed by inbox access settings plus user permissions; a user without ticket access can still see a conversation if they have inbox access.

Chatbot flow building blocks include welcome message, bot options (buttons), customer input capture, branching by contact property, handoff to a human, and ticket creation.

Handoff rules: a chatbot transfers to a live agent when a user selects a human option, when a keyword trigger fires, or when the bot cannot match an intent; the conversation then enters the inbox queue.

Knowledge base grounding lets the AI customer agent answer from published KB articles; unpublished or draft articles are excluded from the agent's answer set.

AI answer behavior is constrained by confidence thresholds; below threshold the agent escalates or offers a handoff rather than guessing, and conversations are logged for review.

Workflow automation for service uses ticket property enrollment triggers, delay steps, and actions such as assign owner, set priority, send email, or create a task.

Routing decision rules: round robin distributes evenly among eligible users, load balanced considers open ticket counts, and contact owner routes to the assigned owner when one exists.

Bot personalization uses contact properties and known conversation context; unknown visitors get generic flows until a property is captured or a contact is identified.

Automation testing rule: workflows and bot branches should be validated with test contacts before enabling, because enrollment criteria are evaluated at trigger time, not retroactively.

AI-assisted agent tools include suggested replies, conversation summaries, and content generation; outputs require agent review before sending unless auto-send is explicitly configured.

Knowledge base structure: knowledge base > category > subcategory > article; categories drive navigation and article URL paths.

Article lifecycle is draft, published, and scheduled; only published articles are visible on the public site and available to chatbots and AI agents.

Access control options are public (no login), private (registered visitors), and private with SSO; private content is gated by membership settings on the knowledge base.

SEO fields per article include title, meta description, URL slug, and canonical tags; category pages and article pages are indexed separately.

Article feedback and analytics track views, helpful/unhelpful votes, and search terms; high search-with-no-result terms indicate missing content.

Article templates (how-to, troubleshooting, FAQ) enforce consistent structure; template choice affects the default headings and body layout.

Multi-language knowledge bases create variant articles per language; each variant has its own publish state and URL.

Content reuse rule: articles can be linked inside chatbot flows and in ticket replies, but edits to a published article take effect immediately for all references.

Self-service deflection is measured by comparing KB article views and search success against ticket volume for the same topics.

Customer health scores are computed from weighted factors such as ticket volume, ticket sentiment, survey responses, product usage, and support response times; each factor has a configurable weight.

Health score grades map numeric ranges to labels (for example healthy, neutral, at risk); thresholds are configurable and should be documented before reporting.

Churn and retention reporting uses cohort analysis: group contacts or companies by start period, then track retained percentage over subsequent periods.

Ticket reporting dimensions include create date, close date, pipeline, stage, owner, priority, source, and associated company; measures include count, average time to close, and average time to first response.

SLA reporting measures the percentage of tickets meeting each SLA target, split by pipeline and by team, and highlights breached versus at-risk tickets.

CSAT is measured via survey with a 1-5 or 1-7 scale; the standard metric is the percentage of positive responses, and response rate should be reported alongside it.

NPS uses the 0-10 scale; promoters are 9-10, passives 7-8, detractors 0-6, and NPS equals percent promoters minus percent detractors.

Dashboard reporting rule: filters applied at the report level override dashboard-level filters; conflicting filters produce confusing results and should be consolidated.

Retention insight sources include ticket sentiment trends, repeat-contact rates, escalation frequency, and survey verbatims; these are leading indicators relative to churn events.

A ticket is a record of a customer request with a pipeline, stage, owner, priority, and source; the ticket pipeline stage replaces generic status labels and drives reporting.

Conversations inbox aggregates email, chat, and form channels into one thread view; replies sent from the inbox are logged on the contact record automatically.

Ticket creation sources include manual creation, form submissions, chat and chatbot flows, workflow automation, inbound email to a connected inbox, and API/CRM integrations.

Ticket association rules: a ticket can be associated with contacts, companies, deals, and other tickets; associations determine which records appear in the ticket sidebar and in cross-object reporting.

Inbox routing order is typically channel or inbox assignment first, then routing rules such as round robin or contact owner; unassigned conversations fall to the inbox's default assignee.

Ticket status vs. stage: stage reflects position in the pipeline lifecycle, while closed status is set when a ticket enters a closed stage; reopening moves it to an open stage and restores SLA clocks per settings.

SLA rules attach to tickets via pipeline and conditions; the two clocks are time to first response and time to close, and each has its own working-hours calendar.

Owner assignment can be manual, via workflow, via routing rule, or via round-robin in help desk; reassignment history is visible in the ticket activity log.

Shared inbox permissions are governed by inbox access settings plus user permissions; a user without ticket access can still see a conversation if they have inbox access.

Chatbot flow building blocks include welcome message, bot options (buttons), customer input capture, branching by contact property, handoff to a human, and ticket creation.

Handoff rules: a chatbot transfers to a live agent when a user selects a human option, when a keyword trigger fires, or when the bot cannot match an intent; the conversation then enters the inbox queue.

Knowledge base grounding lets the AI customer agent answer from published KB articles; unpublished or draft articles are excluded from the agent's answer set.

AI answer behavior is constrained by confidence thresholds; below threshold the agent escalates or offers a handoff rather than guessing, and conversations are logged for review.

Workflow automation for service uses ticket property enrollment triggers, delay steps, and actions such as assign owner, set priority, send email, or create a task.

Routing decision rules: round robin distributes evenly among eligible users, load balanced considers open ticket counts, and contact owner routes to the assigned owner when one exists.

Bot personalization uses contact properties and known conversation context; unknown visitors get generic flows until a property is captured or a contact is identified.

Automation testing rule: workflows and bot branches should be validated with test contacts before enabling, because enrollment criteria are evaluated at trigger time, not retroactively.

AI-assisted agent tools include suggested replies, conversation summaries, and content generation; outputs require agent review before sending unless auto-send is explicitly configured.

Knowledge base structure: knowledge base > category > subcategory > article; categories drive navigation and article URL paths.

Article lifecycle is draft, published, and scheduled; only published articles are visible on the public site and available to chatbots and AI agents.

Access control options are public (no login), private (registered visitors), and private with SSO; private content is gated by membership settings on the knowledge base.

SEO fields per article include title, meta description, URL slug, and canonical tags; category pages and article pages are indexed separately.

Article feedback and analytics track views, helpful/unhelpful votes, and search terms; high search-with-no-result terms indicate missing content.

Article templates (how-to, troubleshooting, FAQ) enforce consistent structure; template choice affects the default headings and body layout.

Multi-language knowledge bases create variant articles per language; each variant has its own publish state and URL.

Content reuse rule: articles can be linked inside chatbot flows and in ticket replies, but edits to a published article take effect immediately for all references.

Self-service deflection is measured by comparing KB article views and search success against ticket volume for the same topics.

Customer health scores are computed from weighted factors such as ticket volume, ticket sentiment, survey responses, product usage, and support response times; each factor has a configurable weight.

Health score grades map numeric ranges to labels (for example healthy, neutral, at risk); thresholds are configurable and should be documented before reporting.

Churn and retention reporting uses cohort analysis: group contacts or companies by start period, then track retained percentage over subsequent periods.

Ticket reporting dimensions include create date, close date, pipeline, stage, owner, priority, source, and associated company; measures include count, average time to close, and average time to first response.

SLA reporting measures the percentage of tickets meeting each SLA target, split by pipeline and by team, and highlights breached versus at-risk tickets.

CSAT is measured via survey with a 1-5 or 1-7 scale; the standard metric is the percentage of positive responses, and response rate should be reported alongside it.

NPS uses the 0-10 scale; promoters are 9-10, passives 7-8, detractors 0-6, and NPS equals percent promoters minus percent detractors.

Dashboard reporting rule: filters applied at the report level override dashboard-level filters; conflicting filters produce confusing results and should be consolidated.

Retention insight sources include ticket sentiment trends, repeat-contact rates, escalation frequency, and survey verbatims; these are leading indicators relative to churn events.

A ticket is a record of a customer request with a pipeline, stage, owner, priority, and source; the ticket pipeline stage replaces generic status labels and drives reporting.

Conversations inbox aggregates email, chat, and form channels into one thread view; replies sent from the inbox are logged on the contact record automatically.

Ticket creation sources include manual creation, form submissions, chat and chatbot flows, workflow automation, inbound email to a connected inbox, and API/CRM integrations.

Ticket association rules: a ticket can be associated with contacts, companies, deals, and other tickets; associations determine which records appear in the ticket sidebar and in cross-object reporting.

Inbox routing order is typically channel or inbox assignment first, then routing rules such as round robin or contact owner; unassigned conversations fall to the inbox's default assignee.

Ticket status vs. stage: stage reflects position in the pipeline lifecycle, while closed status is set when a ticket enters a closed stage; reopening moves it to an open stage and restores SLA clocks per settings.

SLA rules attach to tickets via pipeline and conditions; the two clocks are time to first response and time to close, and each has its own working-hours calendar.

Owner assignment can be manual, via workflow, via routing rule, or via round-robin in help desk; reassignment history is visible in the ticket activity log.

Shared inbox permissions are governed by inbox access settings plus user permissions; a user without ticket access can still see a conversation if they have inbox access.

Chatbot flow building blocks include welcome message, bot options (buttons), customer input capture, branching by contact property, handoff to a human, and ticket creation.

Handoff rules: a chatbot transfers to a live agent when a user selects a human option, when a keyword trigger fires, or when the bot cannot match an intent; the conversation then enters the inbox queue.

Knowledge base grounding lets the AI customer agent answer from published KB articles; unpublished or draft articles are excluded from the agent's answer set.

AI answer behavior is constrained by confidence thresholds; below threshold the agent escalates or offers a handoff rather than guessing, and conversations are logged for review.

Workflow automation for service uses ticket property enrollment triggers, delay steps, and actions such as assign owner, set priority, send email, or create a task.

Routing decision rules: round robin distributes evenly among eligible users, load balanced considers open ticket counts, and contact owner routes to the assigned owner when one exists.

Bot personalization uses contact properties and known conversation context; unknown visitors get generic flows until a property is captured or a contact is identified.

Automation testing rule: workflows and bot branches should be validated with test contacts before enabling, because enrollment criteria are evaluated at trigger time, not retroactively.

AI-assisted agent tools include suggested replies, conversation summaries, and content generation; outputs require agent review before sending unless auto-send is explicitly configured.

Knowledge base structure: knowledge base > category > subcategory > article; categories drive navigation and article URL paths.

Article lifecycle is draft, published, and scheduled; only published articles are visible on the public site and available to chatbots and AI agents.

Access control options are public (no login), private (registered visitors), and private with SSO; private content is gated by membership settings on the knowledge base.

SEO fields per article include title, meta description, URL slug, and canonical tags; category pages and article pages are indexed separately.

Article feedback and analytics track views, helpful/unhelpful votes, and search terms; high search-with-no-result terms indicate missing content.

Article templates (how-to, troubleshooting, FAQ) enforce consistent structure; template choice affects the default headings and body layout.

Multi-language knowledge bases create variant articles per language; each variant has its own publish state and URL.

Content reuse rule: articles can be linked inside chatbot flows and in ticket replies, but edits to a published article take effect immediately for all references.

Self-service deflection is measured by comparing KB article views and search success against ticket volume for the same topics.

Customer health scores are computed from weighted factors such as ticket volume, ticket sentiment, survey responses, product usage, and support response times; each factor has a configurable weight.

Health score grades map numeric ranges to labels (for example healthy, neutral, at risk); thresholds are configurable and should be documented before reporting.

Churn and retention reporting uses cohort analysis: group contacts or companies by start period, then track retained percentage over subsequent periods.

Ticket reporting dimensions include create date, close date, pipeline, stage, owner, priority, source, and associated company; measures include count, average time to close, and average time to first response.

SLA reporting measures the percentage of tickets meeting each SLA target, split by pipeline and by team, and highlights breached versus at-risk tickets.

CSAT is measured via survey with a 1-5 or 1-7 scale; the standard metric is the percentage of positive responses, and response rate should be reported alongside it.

NPS uses the 0-10 scale; promoters are 9-10, passives 7-8, detractors 0-6, and NPS equals percent promoters minus percent detractors.

Dashboard reporting rule: filters applied at the report level override dashboard-level filters; conflicting filters produce confusing results and should be consolidated.

Retention insight sources include ticket sentiment trends, repeat-contact rates, escalation frequency, and survey verbatims; these are leading indicators relative to churn events.

A ticket is a record of a customer request with a pipeline, stage, owner, priority, and source; the ticket pipeline stage replaces generic status labels and drives reporting.

Conversations inbox aggregates email, chat, and form channels into one thread view; replies sent from the inbox are logged on the contact record automatically.

Ticket creation sources include manual creation, form submissions, chat and chatbot flows, workflow automation, inbound email to a connected inbox, and API/CRM integrations.

Ticket association rules: a ticket can be associated with contacts, companies, deals, and other tickets; associations determine which records appear in the ticket sidebar and in cross-object reporting.

Inbox routing order is typically channel or inbox assignment first, then routing rules such as round robin or contact owner; unassigned conversations fall to the inbox's default assignee.

Ticket status vs. stage: stage reflects position in the pipeline lifecycle, while closed status is set when a ticket enters a closed stage; reopening moves it to an open stage and restores SLA clocks per settings.

SLA rules attach to tickets via pipeline and conditions; the two clocks are time to first response and time to close, and each has its own working-hours calendar.

Owner assignment can be manual, via workflow, via routing rule, or via round-robin in help desk; reassignment history is visible in the ticket activity log.

Shared inbox permissions are governed by inbox access settings plus user permissions; a user without ticket access can still see a conversation if they have inbox access.

Chatbot flow building blocks include welcome message, bot options (buttons), customer input capture, branching by contact property, handoff to a human, and ticket creation.

Handoff rules: a chatbot transfers to a live agent when a user selects a human option, when a keyword trigger fires, or when the bot cannot match an intent; the conversation then enters the inbox queue.

Knowledge base grounding lets the AI customer agent answer from published KB articles; unpublished or draft articles are excluded from the agent's answer set.

AI answer behavior is constrained by confidence thresholds; below threshold the agent escalates or offers a handoff rather than guessing, and conversations are logged for review.

Workflow automation for service uses ticket property enrollment triggers, delay steps, and actions such as assign owner, set priority, send email, or create a task.

Routing decision rules: round robin distributes evenly among eligible users, load balanced considers open ticket counts, and contact owner routes to the assigned owner when one exists.

Bot personalization uses contact properties and known conversation context; unknown visitors get generic flows until a property is captured or a contact is identified.

Automation testing rule: workflows and bot branches should be validated with test contacts before enabling, because enrollment criteria are evaluated at trigger time, not retroactively.

AI-assisted agent tools include suggested replies, conversation summaries, and content generation; outputs require agent review before sending unless auto-send is explicitly configured.

Knowledge base structure: knowledge base > category > subcategory > article; categories drive navigation and article URL paths.

Article lifecycle is draft, published, and scheduled; only published articles are visible on the public site and available to chatbots and AI agents.

Access control options are public (no login), private (registered visitors), and private with SSO; private content is gated by membership settings on the knowledge base.

SEO fields per article include title, meta description, URL slug, and canonical tags; category pages and article pages are indexed separately.

Article feedback and analytics track views, helpful/unhelpful votes, and search terms; high search-with-no-result terms indicate missing content.

Article templates (how-to, troubleshooting, FAQ) enforce consistent structure; template choice affects the default headings and body layout.

Multi-language knowledge bases create variant articles per language; each variant has its own publish state and URL.

Content reuse rule: articles can be linked inside chatbot flows and in ticket replies, but edits to a published article take effect immediately for all references.

Self-service deflection is measured by comparing KB article views and search success against ticket volume for the same topics.

Customer health scores are computed from weighted factors such as ticket volume, ticket sentiment, survey responses, product usage, and support response times; each factor has a configurable weight.

Health score grades map numeric ranges to labels (for example healthy, neutral, at risk); thresholds are configurable and should be documented before reporting.

Churn and retention reporting uses cohort analysis: group contacts or companies by start period, then track retained percentage over subsequent periods.

Ticket reporting dimensions include create date, close date, pipeline, stage, owner, priority, source, and associated company; measures include count, average time to close, and average time to first response.

SLA reporting measures the percentage of tickets meeting each SLA target, split by pipeline and by team, and highlights breached versus at-risk tickets.

CSAT is measured via survey with a 1-5 or 1-7 scale; the standard metric is the percentage of positive responses, and response rate should be reported alongside it.

NPS uses the 0-10 scale; promoters are 9-10, passives 7-8, detractors 0-6, and NPS equals percent promoters minus percent detractors.

Dashboard reporting rule: filters applied at the report level override dashboard-level filters; conflicting filters produce confusing results and should be consolidated.

Retention insight sources include ticket sentiment trends, repeat-contact rates, escalation frequency, and survey verbatims; these are leading indicators relative to churn events.