Hootsuite Social Marketing cheat sheet

Social OS unifies five modules: Lumen (listening/analytics), Perch (publishing), Nest (inbox), Parliament (advocacy), and Amplify (paid).

Onboarding order: connect social accounts, set up the organization, invite users, assign roles, then configure streams and boards.

Supported networks typically include Facebook, Instagram, X, LinkedIn, TikTok, YouTube, and Pinterest; each has distinct permission scopes.

Role hierarchy: Super Admin > Admin > Member; permissions are scoped per social account and per module.

A 'board' is a saved view of streams; a 'stream' is a live column of content filtered by network, keyword, or list.

Social OS uses a single sign-on (SSO) option via SAML 2.0; SCIM provisioning is available for user lifecycle management.

The planner view shows scheduled and published posts across all connected accounts in a calendar layout.

Analytics exports are available as CSV or PDF; data retention varies by plan and module.

Integration options include Slack, Salesforce, HubSpot, Google Analytics, and Zapier for workflow automation.

Lumen is the listening and analytics engine inside Social OS, built on Talkwalker's data infrastructure.

A 'topic' defines the query scope; it contains keywords, Boolean operators, and source filters.

Boolean syntax: AND, OR, NOT, parentheses, and quotation marks for exact phrases; wildcards (*) match word variants.

Sentiment analysis classifies mentions as positive, negative, or neutral; emotion detection adds joy, anger, sadness, and surprise.

Data sources include social networks, news sites, blogs, forums, review sites, and video platforms.

A 'mention' is a single post or article matching the topic query; 'reach' estimates potential audience size.

Dashboards can be shared via public link or exported; widgets include volume charts, word clouds, and influencer lists.

Alert rules trigger on volume spikes, sentiment shifts, or specific keyword appearances; delivery is via email or webhook.

Historical data depth depends on the plan; standard topics typically cover up to two years of backfill.

Social OS AI features include post generation, hashtag suggestions, sentiment classification, and LLM-powered summaries.

AI-generated content must be reviewed by a human before publishing; the platform flags AI content for transparency.

Prompt structure: context (brand voice), task (what to generate), constraints (length, tone, platform), and examples.

The 'Working with Wisdom' framework emphasizes verifying AI outputs against source data and brand guidelines.

LLM Insights in Lumen summarize large mention volumes into themes, sentiment drivers, and emerging narratives.

AI models are probabilistic; identical prompts can yield different outputs, so reproducibility requires saved prompts.

Data privacy: AI features process only data within the organization's workspace; opt-out controls exist for model training.

Bias mitigation: review training data assumptions, test prompts across demographics, and audit outputs for stereotypes.

AI content detection is imperfect; always apply editorial judgment and fact-check statistics before publishing.

Perch supports scheduling, queueing, and immediate publishing across connected social accounts.

A 'queue' is a recurring time slot; content added to the queue publishes at the next available slot.

The content calendar displays scheduled, draft, and published posts; drag-and-drop reschedules posts.

Media requirements: Instagram supports JPG/PNG up to 8 MB; X images up to 5 MB; LinkedIn up to 5 MB.

Post approval workflows route drafts to approvers; roles determine who can approve and publish.

Best-time recommendations are based on historical engagement data per account and audience time zone.

UTM parameters can be appended automatically to track campaign traffic in Google Analytics.

Bulk scheduling is available via CSV upload; required columns include message, date, time, and network.

Failed posts generate alerts; common causes are expired tokens, permission changes, or media format violations.

Nest consolidates messages, comments, mentions, and reviews from all connected networks into one inbox.

Conversations can be assigned to team members, tagged, prioritized, and marked as resolved.

Saved replies (canned responses) insert pre-written text; placeholders personalize with contact name.

SLA rules can auto-escalate conversations based on response time or keyword triggers.

Sentiment and intent detection classify incoming messages to route to the right team.

Customer data from CRM integrations (Salesforce, HubSpot) appears alongside the conversation.

Message statuses: open, pending, resolved; resolved conversations can be reopened within a set window.

Automated rules can auto-reply, assign, or tag based on keywords, network, or time of day.

Reporting metrics include response time, resolution time, volume by network, and CSAT if surveys are enabled.

Parliament enables employees to share approved content to their personal social networks.

Content is curated into 'boards' by admins; advocates see available content and can share with one click.

Gamification features include leaderboards, points, and badges based on sharing activity and engagement.

Advocate posts can include suggested text, hashtags, and media; edits may be restricted by admin settings.

Analytics track reach, engagement, clicks, and top advocates; data is aggregated per campaign.

User roles: Admin (manages content and users), Advocate (shares content), and Moderator (reviews submissions).

SSO integration simplifies advocate onboarding; mobile apps support iOS and Android.

Compliance settings can require admin approval before an advocate shares specific content.

Best practice: curate 3-5 pieces per week per advocate to avoid fatigue and maintain authenticity.

Listening topics use Boolean queries; example: ("brand name" OR "product") AND (review OR complaint) NOT spam.

Source filters narrow results by network, country, language, and domain authority.

Sentiment scoring ranges from -100 to +100; neutral is near zero; thresholds are configurable.

LLM Insights generates narrative summaries, theme clusters, and anomaly explanations from mention data.

Influencer identification ranks authors by reach, engagement, and relevance to the topic.

Competitive benchmarking compares share of voice, sentiment, and volume across brands.

Alerts can be set on volume spikes, sentiment drops, or specific keyword emergence.

Data export formats include CSV, XLSX, and API access for BI tools.

Crisis detection uses anomaly detection on volume and sentiment; thresholds trigger alerts to response teams.