This handbook explains each major area of Octant, what it does, and how it works end to end.
On this page
- Getting started
- Core concepts
- Navigation and filters
- Data sources and ingestion
- Research studio
- Analytics
- Insights
- RAG and knowledge base
- Content engine
- Campaigns
- Reports
- Company, organization, product, and persona management
- Settings and administration
- Security and governance
- Troubleshooting
- Additional resources
Getting started
Account and workspace setup
- Sign up with your email and a strong password (8+ characters with upper/lowercase, numbers, and symbols).
- Confirm your email address using the verification link.
- On first login, create or confirm your workspace.
Workspaces are the top-level container for data, teams, and integrations. Each workspace has its own companies, organizations, personas, prompts, analytics, and reports.
Connect providers and integrations
- Go to Settings > Providers and connect your LLM and analytics providers.
- If you use CRM or PR tooling, connect them under Settings > Integrations.
- If you use DataForSEO, add credentials and review usage in the Admin Console.
Providers determine what sources are queried, how prompts are answered, and which analytics signals are available.
Complete onboarding
The onboarding wizard walks you through the core setup in sequence (company, organizations, Build persona funnel, products, hydration). It includes:
- A left sidebar progress tracker to show current step and completion status.
- Refreshed form inputs with inline AI buttons to autofill company and organization descriptions from your website content (cached for future fills) without citation markers.
- AI-labeled product positioning notes and competitor inputs.
- Validation that jumps to and highlights missing fields.
- The ability to return to completed steps to make edits before starting hydration.
Ingest your first data
- Create a research profile or upload documents in Knowledge Base.
- Run your first prompt or listening profile.
- Open the Dashboard to confirm visibility metrics are populating.
When onboarding completes, you are redirected to Prompt Forge to start prompt and listening runs.
Public pages and resources
The public landing page includes Pricing and Live Demo sections for plan review and walkthrough requests before signing up. Public resources are available at /about, /careers, /blog, /contact, /changelog, /legal/privacy, and /legal/terms. If you are already signed in, the landing page shows a Dashboard shortcut; signing out returns you to the landing page.
Core concepts
- Workspaces: isolate data, teams, and integrations. Switching workspaces changes what you see everywhere.
- Entities: companies, organizations, products, personas, and competitors define the context for research, analytics, and content.
- Research profiles: saved sets of prompts, keywords, sources, and schedules that drive ingestion.
- Prompts and listening profiles: the two main ways to query sources. Prompts focus on targeted questions; listening profiles monitor ongoing signals.
- Mentions: normalized references to your products, organizations, or competitors collected by research runs.
- Citations: sources tied to mentions, used to explain why a result appears.
- Visibility score: measures how often and how prominently your company or organization appears in relevant conversations.
- Share of voice (SOV): compares your visibility against competitors in the same space.
- Insights: synthesized trends and opportunities based on mentions and analytics signals.
- Brand kits: capture messaging, tone, and style rules used by content and campaign tooling, managed with search and pagination in the Brand Kits workspace.
Navigation and filters
Navigation
- Use the left navigation headings (Prompt Forge, Listening Engine, Recommendations, Analytics, Settings, Docs) to locate modules faster.
- Expand Insights to access the Organization and Persona dashboards; the Insights heading itself is not selectable.
- Insights provides an on-page switcher to move between Organization and Persona dashboards.
- Use the left navigation Insights dropdown (Listening Engine, Intelligence, Organization, Persona) plus Prompt Forge, Recommendations, Analytics, Settings, and Docs to locate modules faster.
- Listening Engine includes an on-page switcher to move between Listening profiles and Insights views.
- The left navigation SWOTT Library and Prompt Library links open the Listening Engine Insights and Prompt Library tabs and keep the URL in sync for shareable deep links.
- Listening Engine tab pills match the Prompt Forge tab styling for consistent navigation cues.
- Listening Engine now focuses on profile setup, run history, and scheduled runs; the Live Progress and Search History cards are no longer displayed.
- Listening Engine scheduling supports daily and monthly cadences; manual run controls are no longer shown in the UI.
- Listening Engine places the compact Create Search Profile editor on the left with Run History on the right for at-a-glance monitoring.
- Listening Engine opens directly in the Create Search Profile view to speed up first-time setup.
- Listening Engine channels/domains support AI suggestions based on your profile inputs.
- Press Cmd/Ctrl + K to open the command palette for quick navigation and search.
- The header keeps the command palette search, notifications, and account controls aligned on one row at intermediate widths by shortening the search input before wrapping.
- Overview pages for analytics, content, and campaigns include key metrics and deep links.
- Breadcrumbs remain visible on mobile in a condensed format for context.
Context and filtering
- The workspace switcher in the header changes the active workspace without logging out.
- Organization and persona selectors scope dashboards and create flows to the right context; the header labels this area with “Currently viewing:” aligned with the selector text in the dropdowns.
- If no workspace is active, use the workspace picker banner to select or manage a workspace.
- Most views include filters for time range, company, organization, product, topic, or channel.
- Dropdown selectors keep selected text fully visible in compact filters and forms.
- Reports include a visual filter builder for companies, organizations, products, and topics with optional advanced JSON.
- Save frequently used filters by bookmarking the page URL (filters persist in the URL).
Data sources and ingestion
Providers
- Connect LLM and analytics providers under Settings > Providers.
- Global defaults can seed internal workspaces.
- Provider health checks validate keys with a live provider ping.
DataForSEO
- DataForSEO powers SEO rankings, keyword volume, backlinks, reviews, and AI visibility.
- Octant dedupes DataForSEO signals alongside Perplexity to reduce overlap.
Integrations
- Connect CRM and PR tooling under Settings > Integrations.
- Integrations sync contact, lifecycle, and campaign data for dashboard metrics and reporting.
Research profiles
- Define sources, keywords, and cadence for unified ingestion.
- Profiles can be run manually, scheduled hourly/daily/weekly, or configured with cron.
Knowledge base ingestion
- Upload or sync documents to power RAG-assisted generation and context queries.
- Documents are parsed, chunked, embedded, and indexed for retrieval.
How ingestion works
- A profile run is scheduled or launched manually.
- Providers fetch sources and responses based on prompts or listening configurations.
- Mentions and citations are deduplicated and enriched with metadata (source, channel, topic, sentiment).
- Results are stored and surfaced to analytics, insights, and reports.
Research studio
Configure a research profile
- Go to Research.
- In Prompt Forge, review the AI Generate controls labeled “Generate, draft, or import prompts at scale.”, select a persona (and optional linked product), choose a category, add an optional goal, review real user queries, and run Deep Search (the query panel stays empty if DataForSEO is not configured).
- Prompt Forge includes tabs for Prompt Forge and Prompt Library; the Prompt Library tab contains the prompt list and recent runs, and its summary stats (Prompts, Active, Categories, Variables) sit at the top for quick context. Prompt Forge summary stats appear above the Prompt Intelligence panel.
- Manual prompt creation uses an optional Name field and a category dropdown with preset options (How Tos, Documentation, Market Trends, Products, Solutions & Services, Competitors, Social, Any/All); the Save action stays aligned at the bottom-right of the Manual tab, and the header no longer includes a “+ New Prompt” shortcut (use the Manual or Batch Import tabs instead).
- Batch imports accept one prompt per line without a separate category field.
- Use AI assist to draft prompt sets, keywords, sources, and competitors; suggestion previews render structured values for complex fields.
- The Goal input is a collapsible toggle in AI Generate; click Goal to expand and add optional context.
- Use Run Deep Search (all) to launch runs across visible prompts. Prompt Forge cards keep titles truncated so action menus stay aligned.
- Bulk runs queue all visible prompts in parallel for faster startup. Prompt generation defaults to 25 prompts per batch; attempting to exceed 25 at a time opens an upgrade modal to contact sales for higher limits. First-time visits after onboarding auto-generate 10 prompts and start Deep Search so results are ready on arrival.
- Set schedules (manual, hourly, daily, weekly, or cron) where needed.
- Run prompt or listening profiles to start collecting mentions (Perplexity or Google Search grounding handles citations while all active providers synthesize).
- Prompt cards require the Edit button for changes; reruns do not auto-open results, so use View results when needed.
Prompt creation is idempotent by name; submitting a prompt with an existing name returns the existing prompt instead of an error.
Monitor run history
- Review the run timeline for status, next run, and error messages.
- Run History now sits in the left column of the Listening tab, while Search Profiles filters live on the Insights tab.
- Research opens directly in Prompt Forge; prompt and listening run history stays within their respective workspaces instead of a shared tab.
- Listening runs execute query batches concurrently (with rate limiting) to reduce end-to-end run time.
- Due listening profiles start in parallel so scheduled ingestion kicks off immediately.
- Use manual run controls to kick off a run when you need fresh results.
- Live updates stay connected unless the network drops, so the status indicator should remain steady.
Review results
- Prompt citations are stored as mentions and flow into analytics, insights, and reports.
- The model comparison scorecard lists every provider execution in the run, not just the top scores.
- Run Results open full-screen; citations and sources show provider icons for each AI and wrap long URLs cleanly.
- Mentions are deduplicated and enriched with source metadata.
- Channels are inferred automatically (social, forum, news, blog).
- Use filters to narrow by topic, sentiment, or competitor.
Analytics
Analytics aggregates mentions, citations, and provider signals into dashboards and trend views. Tables support horizontal scrolling on smaller screens and the filter bar supports multi-select topics, platforms, and regions. Selections persist in the URL for sharing. The Analytics navigation tabs are ordered as Visibility, Citations, Competitors, and Sentiment.
Topics
- Track emerging themes and topic drift over time.
- Compare sentiment and SOV by topic.
- Paginate topics and expand a row to see entity mention breakdowns.
Regions
- See geographic shifts in attention and sentiment.
- Identify regional opportunities to prioritize.
- Use the interactive map to inspect region-level performance quickly.
Channels and communities
- Compare visibility across social, news, and community sources.
- Monitor channel growth and authority metrics.
Rankings
- Select a topic to see average SEO position trends, tracked keywords, and location averages.
- Review top-ranking domains and URLs for each topic keyword across locations.
- Monitor the topic AI visibility score with per-platform mention and impression totals.
- Review context snippets to understand how topics surface in AI responses.
Backlinks
- Review domain authority, referring domains, and backlink totals for the active workspace.
- Monitor backlink changes to surface meaningful shifts.
Reviews
- Follow review ratings and sentiment across supported platforms.
- Use the reviews timeline to spot reputation swings.
Sentiment
- Monitor the overall sentiment score and week-over-week change.
- Review positive and negative drivers derived from topic sentiment.
- Open entity-level prompt executions to inspect prompts, responses, and citations.
- The Sentiment tab description notes it shows sentiment for each citation per prompt.
AI visibility
- Monitor brand mentions across ChatGPT, Claude, Gemini, and Perplexity.
- Track top influencing domains that surface in LLM responses.
- Review multi-brand visibility and share-of-voice trends with adjustable time granularity.
DataForSEO usage (admin)
- Usage reporting is available in the Admin Console under Analytics.
Competitors
- Benchmark visibility and topic coverage.
- Identify strengths, weaknesses, and opportunity gaps.
Citations
- Review the sources feeding analytics and insights.
- Validate authority and filter out low-quality citations.
- Track citation share-of-voice, top domains, and most-cited URLs.
- Open domain analytics for trends and breakdowns across prompts, topics, platforms, and regions.
Insights
- Insights are generated automatically based on research and analytics signals and organized as a SWOTT library.
- Strength: share-of-voice exceeds 20% (saturated coverage).
- Weakness: share-of-voice is below 10% (coverage gaps).
- Opportunity: weaknesses with high impact potential if addressed.
- Threat: negative sentiment or declining visibility (≥20% mention drop).
- Trend: growing mentions or engagement momentum.
- Mentions show the estimated impact for each insight (displayed as mention counts).
- GEO Impact scores highlight predicted GEO performance along with the recommended asset type and topic; refresh impact from the insight detail page.
- Impact predictions depend on the Myles AI service; if it is unavailable, insights still render but impact fields remain empty until the service is restored.
- Assign owners and due dates to track follow-up actions.
- Regenerate insights after new data ingestion or campaign changes.
- Review evidence to understand the underlying data that drove each insight.
- Listening run completions auto-trigger insight generation once enough new mentions are ingested.
RAG and knowledge base
- Upload documents to enrich context-aware generation.
- Monitor processing status for uploads and re-embedding operations.
- Preview chunked passages to debug retrieval quality.
- Re-embed or download documents from the Knowledge Base manager when content changes.
- Use context query endpoints to retrieve relevant citations for reporting or content.
- Keep documents organized with tags, metadata, and version updates.
- AI assist can suggest document names, descriptions, and tags.
Content engine
Suggestions
- Generate content ideas tied to insights, topics, or personas.
- Review and accept suggestions to move them into production.
- GEO impact scores are shown on suggestions, and impact-recommended items are flagged in the list.
- Toggle between AI-assisted and rules-only suggestion generation.
- Generated content previews appear inline before you open the editor.
Content studio
- Generate or edit assets across formats (blog, email, social, PR).
- Apply brand kit tone and messaging automatically.
- Use the Content Studio tabs (Generate, SEO, Variants, Experiments) to focus each step.
- Draft in the rich text editor and apply AI-assisted edits with inline diff previews.
- Compare versions and track content changes.
- Send drafts directly into email campaigns when you are ready to activate.
SEO and optimization
- Score content for keyword coverage and readability.
- Generate optimization recommendations and internal link ideas.
- Use DataForSEO to generate meta tags, subtopics, and grammar checks.
Experiments and performance
- Run A/B tests on content variations.
- Track performance and feed results back into the learning loop.
Campaigns
Email campaigns
- Build email sequences with personalization tokens.
- Use the stepper (Details, Steps, Preview, Enroll) to move from setup to activation.
- Schedule sends and monitor delivery analytics.
Nurture sequences
- Design multi-step workflows with waits and conditions.
- Drag and drop steps to reorder the flow quickly.
- Use the flowchart view to visualize complex branching logic.
- Enroll contacts and monitor progress through the funnel.
PR campaigns
- Manage press releases and journalist outreach.
- Track distribution status and coverage.
- AI assist can draft press release briefs, titles, and content.
Campaign analytics
- Measure funnel impact, engagement, and PR coverage metrics.
- AI assist can draft campaign summaries, email step copy, and targeting keywords.
Reports
- Use templates to standardize recurring reports.
- Follow the report wizard (Template -> Date Range -> Recipients -> Review) for guided setup.
- Recipient emails are validated as you type, with invalid addresses highlighted.
- Generate PDFs or slide decks with charts and tables.
- Build filters with company/organization/product/topic selectors and optional advanced JSON.
- Schedule periodic reports for stakeholders.
- Share reports securely with expiring links.
- Status updates refresh automatically while reports are generating.
- AI assist can suggest report names and schedule filters/settings.
Company, organization, product, and persona management
Company and organization
- Companies establish the top-level context for executive dashboards.
- Organizations capture sub-entities, brands, or business units under the company.
- Company dashboards show SEO/SEM performance using real content metrics (organic traffic, top keywords, avg SEO score, CTR).
- CRM metrics in company dashboards use CRM lifecycle/stage data (HubSpot lifecycle stage, Salesforce lead status/converted) and display total contacts, new leads, qualified, and closed won counts.
Business units and dashboards
- Business units capture divisions or product lines under the company.
- Business unit dashboards support per-card filters for actions, conversations, and segments, plus time-range selection.
- Conversion & ROI Efficiency adds persona and date-range filters and treats open interviews as active CRM opportunities.
- Conversation filters narrow by source host and sentiment; action filters narrow by priority and action type.
- Time-range selection limits dashboard content to recent actions, mentions, and metric snapshots.
Brand kits and brand settings
- Brand kits define tone, positioning, and messaging pillars for content workflows.
- Brand kits can be linked to organization topics so AI-generated content stays aligned with the topics you care about; manage selections in the Topics tab when editing a kit.
- The Brand Kits page supports search, pagination, and quick actions to edit or delete kits.
- Use Brand Settings to update brand name, website URL, description, and aliases that guide AI content generation.
Products
- Products organize analytics and reporting by line-of-business.
- You can upload up to 5 PDF datasheets per product to enrich generation and reporting.
Personas
- Personas capture profiles, domain, industries, geography (including a Global option for worldwide targeting), application areas/use cases, market trends, and care-about prompts.
- AI suggestions can emphasize compliance topics and requirements.
- Persona Builder AI suggestions retry and fall back gracefully if a generation response is malformed, keeping the flow responsive.
- Persona Builder starts with structured inputs (organizations, profiles, ecosystem, domain, industries, geography, use cases, market trends) and includes AI assist to draft persona descriptions and suggest use cases.
- Step 2 confirms persona care-about prompts (3-5 recommended, max 10) with prompts that include application requirements, industry standards/regulations/certifications, and end-user requirements.
- Step 3 generates a preview with summary, goals, pain points, and an AI-generated avatar.
- Step 4 completes the flow and saves the persona (also used during onboarding).
- Persona avatars auto-generate after creation; regenerate them from the persona detail header if needed.
- External mode exposes a public persona builder at
/persona-builderwith the same Step 1-3 flow and business-email-gated exports. - Persona exports require a business email and deliver a secured PDF by download or email; full persona kits route through the contact form.
- Persona inference helps map CRM segments to personas (review before applying).
- Persona dashboards surface active prompts and recommended prompts that can be activated or dismissed.
- AI assist panels can draft company, product, persona, and brand kit fields to accelerate setup.
Settings and administration
Workspace settings
- Workspace settings: name, region, timezone, and defaults.
- Members and roles: invite teammates, resend/cancel invitations, and assign permissions.
- The Settings overview includes a Create panel with shortcuts to Organizations, Personas, Products, and Brand Kits.
- Theme: Octant uses a unified light theme across all pages (no dark mode toggle).
- Side panels and navigation sidebars share consistent light styling and allow scrolling for long forms.
- AI assist can help draft workspace descriptions.
Providers and integrations
- Providers: manage API keys and run health checks that validate keys with a live provider ping.
- DataForSEO: configure credentials, budgets, and review usage in the Admin Console.
- Integrations: connect CRM, PR, newswire, and email systems with inline OAuth status tracking.
- HubSpot integrations expose custom object schema/status endpoints for manual setup, plus sync status monitoring and manual sync triggers for the Organization Dashboard; the Integrations settings page shows status, schemas, and fallback warnings.
- HubSpot custom object and company property names use the
signalplane_prefix (replacingoctant_). - If custom objects are missing or scopes are unavailable, sync falls back to company properties, notes, and tasks; reauthorize with custom-object scopes to enable full custom object sync.
- Configure newswire provider credentials (PRNewswire, BusinessWire, PRWeb, or custom) and delivery modes.
- Choose an email delivery provider (SMTP, SendGrid, SES) for campaign sends.
Billing, notifications, and security
- Billing: review plan limits and usage.
- Notifications and security: configure email preferences, verify your email, and reset passwords.
Admin console (superadmins)
- The Admin Console lives at
/adminand is restricted to superadmins. - Track system KPIs (users, workspaces, subscriptions, MRR) and platform usage trends.
- Manage user roles, workspace status, billing plans, and global provider defaults.
- Review system audit logs, maintenance mode, and health indicators for incident response.
- Impersonate users to reproduce issues and invalidate sessions when access needs to be cut quickly.
- Audit logs capture action, entity, IP address, and user agent for compliance tracking.
- Monitor DataForSEO usage by workspace, endpoint, and spend window.
- Integration details and operational guidance live in
docs/knowledge_base/admin-console.md.
Security and governance
- Use role-based access controls for workspace permissions.
- Limit share links to approved domains and set expirations.
- Rotate provider keys and revoke unused integrations.
- Audit changes by reviewing workspace activity and insight history.
- Verify email addresses to confirm account ownership for sensitive actions.
- Password reset links are single-use and expire after a short window.
Troubleshooting
No data on the dashboard
- Verify the active workspace and confirm providers are connected.
- Ensure at least one search profile or document ingestion job has run.
- Check that a time range filter is not excluding all results.
Insights are stale
- Trigger a manual re-run for research profiles.
- Regenerate insights from the Insights page.
Reports stuck in queued
- Confirm report schedules are active and workers are running.
- Retry the report generation from the reports list.
Research runs stuck in queued
- Confirm providers are connected and workers are running.
- Research runs should move to running within about a minute; retry the run if needed.
Errors right after a deployment
- Refresh after a few minutes to allow database migrations to finish.
- If the deployment fails during build, ensure the build pipeline compiles TypeScript before pruning dev dependencies.
Redirected to login
- Your session expired or refresh failed; sign in again to continue.
Personas page freezes or keeps loading
- Confirm a workspace is selected, then refresh the page to reload persona data.
- If personas were recently linked to multiple business units, confirm the latest backend migrations are deployed; older deployments may only show personas tied to the primary business unit.
- If the issue persists, sign out and back in, or contact support.
Content generation fails
- Check provider health in Settings > Providers to validate API keys.
- Validate that your brand kit and prompts include required fields.
Integrations not syncing
- Reconnect the integration and verify OAuth tokens.
- Check integration-specific settings for scope or permissions.
- For HubSpot, review custom object status (schemas/status) and confirm the sync status panel shows recent successes; missing objects fall back to notes/tasks/company properties.
Additional resources
- User guide:
docs/knowledge_base/user-guide.md(in-app at/docs/guide) - Tutorials:
docs/knowledge_base/user-tutorials.md(in-app at/docs/tutorials) - Video tutorials:
docs/knowledge_base/video-tutorials.md - FAQ:
docs/knowledge_base/faq.md(in-app at/docs/faq) - Developer guide:
docs/knowledge_base/developer-guide.md - API docs:
docs/openapi.json