# All agents
Source: https://docs.asklantern.com/agents/all-agents
Browse Lantern's specialist marketing agents and choose the right one for the outcome you need.
Lantern Agents are specialised marketing assistants. Choose one based on the outcome you need, give it a clear brief and the relevant brand context, then review its work before it is published or put into action.
## Choosing an agent
Use an agent when you need focused marketing work completed faster, with research and brand context brought together in one place. The best briefs explain the audience, goal, subject, deadline, and any essential messages, sources, or constraints.
Every output should receive human review. Your team remains responsible for final decisions, factual accuracy, legal or compliance approval, and publishing.
## Content and search
Create a researched, on-brand article, guide, or campaign content.
Plan an article, landing page, or resource before writing.
Build an authoritative, long-form resource on a strategic topic.
Answer common questions in clear, on-brand language.
Write stronger titles and descriptions for search results.
Prioritise search topics and shape your content plan.
Understand search results and find content gaps.
Prioritise the organic search improvements most likely to matter.
## Brand and strategy
Answer marketing questions, choose next steps, and coordinate larger work.
Assess a broad marketing question before committing time or budget.
Compare your brand with competitors and find ways to stand out.
Check that content sounds recognisably like your brand.
Check that important claims are accurate and supportable.
## Paid media
Understand paid-search performance and what is changing.
Get practical recommendations to improve existing campaigns.
Create new paid-search campaign concepts and ad copy.
Develop campaign concepts for advertising experiences in ChatGPT.
## PR and communications
Create media-ready communications for launches and announcements.
Find credible earned-media and expert-commentary opportunities.
## Older workflows
Continue work created in earlier versions of Lantern.
# Brand Voice QA Agent
Source: https://docs.asklantern.com/agents/library/brand-voice-qa-agent
Check whether content feels recognisably like your brand before it reaches an audience.
## Use it when
You want to check whether content feels recognisably like your brand before it reaches an audience.
## What to provide
The draft content and the brand voice, style guide, approved examples, or messaging principles it should follow.
## What you receive
A brand-voice assessment that highlights strengths, off-brand language, and practical improvements.
## Example request
```text theme={null}
Review this product launch email against our voice guidelines. We should sound direct, expert, and approachable, not overly technical.
```
Keep your [Brand Kit](/brand-kit/overview) up to date so voice reviews use your current guidelines.
# ChatGPT Ads Campaign Generator
Source: https://docs.asklantern.com/agents/library/chatgpt-ads-campaign-generator
Develop a campaign concept for advertising experiences in ChatGPT.
## Use it when
You want to develop a campaign concept for advertising experiences in ChatGPT.
## What to provide
The campaign brief, audience, destination page, offer, and the outcome you want the campaign to drive.
## What you receive
A campaign concept and ad content tailored to the brief and destination page.
## Example request
```text theme={null}
Create a ChatGPT Ads campaign concept for our executive AI-readiness assessment, aimed at UK mid-market CEOs.
```
ChatGPT Ads availability depends on your plan. See [Pricing](/pricing).
# Competitor Audit Agent
Source: https://docs.asklantern.com/agents/library/competitor-audit-agent
Understand how your brand compares with competitors and where it can stand out.
## Use it when
You need to understand how the brand compares with competitors and where it can stand out.
## What to provide
The brand, competitors, market, target audience, and any claims or messages you want to test.
## What you receive
A competitor analysis covering relative strengths, weaknesses, messaging gaps, and opportunities to differentiate.
## Example request
```text theme={null}
Compare our analytics platform with three enterprise competitors and identify the strongest position for data-governance buyers.
```
# Content Agent
Source: https://docs.asklantern.com/agents/library/content-agent
Create a researched, on-brand article, guide, or campaign content around a defined topic.
## Use it when
You need a useful, on-brand article, guide, or campaign content around a defined topic.
## What to provide
The topic, intended reader, business goal, target page or campaign, key messages, and any sources or points that must be included or avoided.
## What you receive
A researched article draft with a clear angle, supporting sources, key claims, and editorial review notes.
## Example request
```text theme={null}
Create a practical guide for operations leaders evaluating AI search visibility tools. Explain the decision criteria, include examples, and avoid promotional language.
```
Review every draft before it is published. Your team remains responsible for factual accuracy and final approval.
# Content Brief Agent
Source: https://docs.asklantern.com/agents/library/content-brief-agent
Give writers a clear plan before creating an article, landing page, or resource.
## Use it when
A writer needs a clear plan before creating an article, landing page, or resource.
## What to provide
The subject, audience, purpose, target search topic, and any existing materials or campaign goals.
## What you receive
A content brief covering audience needs, search intent, key messages, suggested structure, and research direction.
## Example request
```text theme={null}
Create a brief for an article that helps retail marketing teams prepare for seasonal search-demand changes.
```
# FAQ Generator Agent
Source: https://docs.asklantern.com/agents/library/faq-generator-agent
Answer common questions before they become barriers to conversion or customer confidence.
## Use it when
You want to answer common questions before they become barriers to conversion, sales conversations, or customer confidence.
## What to provide
The product, service, page, audience, and the questions customers commonly ask.
## What you receive
Clear, on-brand questions and answers that can be used on a website, in sales materials, or in support content.
## Example request
```text theme={null}
Create FAQs for a B2B software pricing page. Address implementation time, security, contracts, support, and integration requirements.
```
# Google Ads Analysis Agent
Source: https://docs.asklantern.com/agents/library/google-ads-analysis-agent
Understand what paid-search activity is achieving and where performance is changing.
## Use it when
You need to understand what paid-search activity is achieving and where performance may be improving or declining.
## What to provide
The campaign or account context, business goal, period to review, target audience, and any performance concerns.
## What you receive
A Google Ads insight report covering performance patterns, notable findings, and potential actions.
## Example request
```text theme={null}
Review Q3 paid-search performance for our UK lead-generation campaigns and explain why qualified leads fell in September.
```
# Google Ads Campaign Agent
Source: https://docs.asklantern.com/agents/library/google-ads-campaign-agent
Create new paid-search campaign concepts and copy without starting from a blank page.
## Use it when
You need new paid-search campaign concepts and copy without starting from a blank page.
## What to provide
The offer, audience, target market, destination page, key benefits, and campaign objective.
## What you receive
Three review-ready responsive search-ad campaign options with messages and ad copy.
## Example request
```text theme={null}
Create three Google Ads campaign options for our new fraud-prevention platform, aimed at payments leaders in the UK.
```
# Google Ads Optimisation Agent
Source: https://docs.asklantern.com/agents/library/google-ads-optimisation-agent
Get practical recommendations to make existing Google Ads campaigns work harder.
## Use it when
You want practical recommendations to make existing Google Ads campaigns work harder.
## What to provide
The campaign goal, performance data, target audience, budget constraints, and any changes already under consideration.
## What you receive
Recommendations for campaign, keyword, and search-term improvements, ready for review by the paid-media team.
## Example request
```text theme={null}
Recommend optimisations for our high-cost, low-conversion search campaigns without increasing monthly spend.
```
Recommendations are not applied automatically. Your paid-media team decides which changes to make.
# Keyword Research Agent
Source: https://docs.asklantern.com/agents/library/keyword-research-agent
Decide which search topics to prioritise and how they should shape your content plan.
## Use it when
You need to decide which search topics to prioritise and how they should shape your content plan.
## What to provide
The market, target audience, products or services, business priorities, competitors, and any existing search or analytics data.
## What you receive
A keyword opportunity report with search themes, audience intent, competitive context, and recommended content ideas.
## Example request
```text theme={null}
Find the best search opportunities for a cybersecurity consultancy focused on mid-market healthcare providers.
```
# LanternIQ
Source: https://docs.asklantern.com/agents/library/lanterniq
Answer marketing questions, choose the next step, and coordinate larger pieces of work.
## Use it when
You have a marketing question, need help choosing the next step, or want to coordinate a larger piece of work.
## What to provide
Your question, the outcome you are trying to achieve, any relevant campaign or brand context, and decision constraints.
## What you receive
A practical answer, a recommended plan, or help assigning the right specialist agent.
## Example request
```text theme={null}
We are launching into a new market in six weeks. What should the marketing team validate before we commit budget to content and paid search?
```
See the [LanternIQ overview](/lanterniq/overview) for how conversations, evidence, and approvals work.
# Marketing Analyst
Source: https://docs.asklantern.com/agents/library/marketing-analyst
Get an informed view of a broad marketing question before committing time, budget, or a campaign.
## Use it when
You need an informed view of a broad marketing question before committing time, budget, or a campaign.
## What to provide
The decision you need to make, business context, audience, market, timing, and any data or assumptions already available.
## What you receive
A concise research-led assessment of the situation, the most important findings, and the recommended next action.
## Example request
```text theme={null}
Assess whether our current messaging is strong enough for a move from founder-led sales to a demand-generation programme.
```
# Meta Titles Agent
Source: https://docs.asklantern.com/agents/library/meta-titles-agent
Write stronger page titles and descriptions for search results.
## Use it when
A page needs a stronger title and description for search results.
## What to provide
The page URL or page copy, its target audience, its main search topic, and any required wording.
## What you receive
Search-focused title tag and meta description options designed to accurately describe the page and encourage relevant clicks.
## Example request
```text theme={null}
Write title tag and meta description options for our enterprise analytics integration page, targeting marketing operations leaders.
```
# Older workflows
Source: https://docs.asklantern.com/agents/library/older-workflows
Continue content and keyword research work created in earlier versions of Lantern.
These agents support work created through earlier versions of Lantern. Use the current [Content Agent](/agents/library/content-agent) and [Keyword Research Agent](/agents/library/keyword-research-agent) for new work whenever possible.
## Content Agent (existing version)
**Use it when:** You are maintaining or continuing an existing content workflow created in an earlier version of Lantern.
**You receive:** A researched content draft and supporting material that remains compatible with that existing work.
## Keyword Research Agent (existing version)
**Use it when:** You are maintaining or continuing an existing keyword research workflow created in an earlier version of Lantern.
**You receive:** A keyword research report that remains compatible with that existing work.
# Pillar Article Agent
Source: https://docs.asklantern.com/agents/library/pillar-article-agent
Create a substantial, authoritative resource that supports long-term organic visibility.
## Use it when
You need a substantial, authoritative resource on a strategic topic that can support long-term organic visibility.
## What to provide
The central topic, audience, point of view, important subtopics, and relevant brand expertise or source material.
## What you receive
A long-form cornerstone article designed to answer a broad topic thoroughly and create a useful reference for readers.
## Example request
```text theme={null}
Create a comprehensive guide to digital trust for financial-services marketing leaders, including governance, measurement, and implementation examples.
```
Review every draft before it is published. Your team remains responsible for factual accuracy and final approval.
# PR Agent
Source: https://docs.asklantern.com/agents/library/pr-agent
Create media-ready communications for a launch, announcement, or news moment.
## Use it when
You need media-ready communications for a launch, announcement, executive point of view, or news moment.
## What to provide
The announcement, intended audience, key facts, spokespeople, proof points, timing, and desired response.
## What you receive
A press release, media pitch, or executive briefing tailored to the audience and communications goal.
## Example request
```text theme={null}
Draft a press release announcing our expansion into Germany, including a quote from the CEO and evidence of customer demand.
```
Confirm quotes, facts, and timing with the relevant stakeholders before distribution.
# PR Opportunity Agent
Source: https://docs.asklantern.com/agents/library/pr-opportunity-agent
Find credible earned-media and expert-commentary opportunities that improve brand visibility.
## Use it when
You want to find credible earned-media or expert-commentary opportunities that could improve brand visibility.
## What to provide
The brand, market, priority topics, target publications or audiences, and any areas where visibility needs to improve.
## What you receive
A prioritised list of PR and outreach opportunities connected to the brand's visibility gaps.
## Example request
```text theme={null}
Identify PR opportunities that could build our authority in supply-chain resilience among European manufacturing leaders.
```
# SEO Agent
Source: https://docs.asklantern.com/agents/library/seo-agent
Get a clear view of the organic search improvements most likely to matter.
## Use it when
You need a clear view of the organic search improvements most likely to matter.
## What to provide
The website or section to review, business priorities, important pages, target markets, and known concerns.
## What you receive
A prioritised SEO report with issues, opportunities, and recommended next actions.
## Example request
```text theme={null}
Review our resource centre for organic search opportunities and prioritise the fixes that could improve qualified traffic over the next quarter.
```
# SERP Research Agent
Source: https://docs.asklantern.com/agents/library/serp-research-agent
Understand what audiences see when they search for a topic and how to create a more useful result.
## Use it when
You want to understand what audiences see when they search for a topic and how to create a more useful result.
## What to provide
A topic, search query, target audience, market, and the page or content idea you are considering.
## What you receive
A report on search intent, competing content, common audience questions, and content gaps worth addressing.
## Example request
```text theme={null}
Research the search landscape for "customer data platform implementation" and identify what an enterprise buyer still needs to know.
```
# Trust QA Agent
Source: https://docs.asklantern.com/agents/library/trust-qa-agent
Check whether important claims are accurate, supportable, and safe to publish.
## Use it when
You need to check whether important claims are accurate, supportable, and safe to publish.
## What to provide
A page, article, document, or text draft, along with any sources that support key statements.
## What you receive
A trust report identifying claims, available support, evidence gaps, and points requiring review.
## Example request
```text theme={null}
Review this sustainability report for factual claims that need stronger support before publication.
```
A trust report supports your review process. It does not replace legal or compliance approval.
# Agents overview
Source: https://docs.asklantern.com/agents/overview
Run specialists for defined marketing deliverables and review their durable outputs.
Agents are reusable specialist workflows for defined marketing outcomes.
Each specialist has a bounded purpose, expected inputs, available evidence, and a durable output. LanternIQ can launch the right specialist from a clear request, or you can create and run an agent from the Agents workspace.
## Run an agent
1. Choose a specialist or existing agent.
2. Provide the required goal and inputs.
3. Review the expected run allowance.
4. Start the run.
5. Follow progress or continue working elsewhere.
6. Review the output in the run and Projects.
## Progress and results
Runs show meaningful stages, evidence work, specialist decisions, and the final output. Technical prompts, payloads, and hidden instructions are not part of the customer-facing record.
## Schedules
Reusable agents can run manually or on a configured schedule where scheduling is available. Confirm the goal, data access, output destination, and approval behavior before enabling recurring work.
## Publishing safety
Generating a draft is not the same as publishing. Live publication and other external actions depend on the connected system, workflow configuration, permissions, and approval boundary.
# Runs, progress, and recovery
Source: https://docs.asklantern.com/agents/runs-progress-recovery
Understand agent run states, usage, concurrency, and retries.
An agent run is one accepted execution of a specialist workflow.
## Run states
A run can be queued, running, awaiting review, complete, failed, or cancelled. Linked work and Project outputs remain available after the run finishes.
## Usage
One accepted run consumes one included run. Continuation turns and infrastructure retries do not consume another run. A manual retry starts new work and may consume another run.
## Concurrency
Your plan controls how many runs can execute at once across the organization. Additional accepted work remains queued until capacity is available.
## Stop and retry
Stop requests cancellation of active work. Some external operations may already have completed before cancellation is observed. Retry starts another attempt using the available saved context.
## Recovery
Lantern monitors queued and running work for stalled progress and can safely re-enqueue eligible work. Safeguards prevent the same accepted request from being charged or executed twice during infrastructure retries.
# Specialist catalog
Source: https://docs.asklantern.com/agents/specialist-catalog
Choose the smallest Lantern specialist that matches your marketing outcome.
Use the smallest specialist that matches the requested outcome.
| Specialist | Best for | Typical output |
| - | - | - |
| GEO Agent | AI-search visibility, prompts, citations, topics, and competitor pressure | Visibility analysis and prioritized opportunities |
| Keyword Research Agent | Search demand and keyword opportunities | Keyword research brief |
| FAQ Generator Agent | Useful, brand-aligned questions and answers | FAQ section |
| Competitor Audit Agent | Positioning, content, and AI-perception analysis | Competitor audit |
| Content Brief Agent | Evidence-backed planning before drafting | Content brief and outline |
| Pillar Article Agent | Long-form authority content | Article draft |
| Meta Titles Agent | Page titles and descriptions | Metadata set |
| Marketing Analyst | Analysis across connected marketing data | Cited analysis and recommendations |
| Content Agent | Researched content production | Content draft |
| Refresh Content Agent | Existing-page diagnosis and revision | Refreshed draft |
| Presentation Agent | Executive and client communication | Presentation narrative or deck artifact |
| Ads Agent | Campaign structure, copy angles, and optimization | Advertising draft or recommendation |
| SEO Agent | Site health, rankings, and priorities | SEO report |
| Trust QA Agent | Claim, statistic, source, and attribution verification | Trust report |
| Research Agent | Market, competitor, or topic research | Research report |
Some capabilities may depend on your plan, connected data, brand context, or release configuration.
# AI visibility
Source: https://docs.asklantern.com/analytics/ai-visibility
Review how often and how prominently your brand appears across tracked buyer questions.
AI Visibility shows how often and how prominently the active brand appears across its tracked buyer questions.
## Read the score responsibly
The score summarizes results from the current tracked prompt cohort and provider set. Use the underlying prompts, mentions, citations, and comparison period to understand why the score moved.
## Competitor comparisons
Competitor results use the competitors configured for the brand. A competitor mention is not an owned-brand mention, and total citations must not be presented as the brand's citations.
## Run analysis
Run Analysis refreshes AI visibility results for the tracked prompts and supported providers. It does not refresh Google Analytics or Search Console.
## Use LanternIQ
Ask LanternIQ to explain a movement, identify weak prompts, compare owned and competitor citation pressure, or recommend the next measurable action.
# Citations
Source: https://docs.asklantern.com/analytics/citations
Review the owned, competitor, and independent sources referenced in tracked AI answers.
Citations shows the sources referenced in tracked AI answers.
## Source categories
* Owned: pages on the active brand's domains.
* Competitor: pages associated with configured competitors.
* Independent: publications, communities, directories, and other external sources.
Keep these categories separate when comparing performance.
## Use citation evidence
Look for owned pages that already earn citations, competitor pages influencing answers, and independent sources that shape category trust.
A citation opportunity is a prioritization signal, not a guarantee that outreach or new content will produce a citation.
# Analytics overview and methodology
Source: https://docs.asklantern.com/analytics/overview
Understand how Lantern separates AI visibility, citations, prompts, traffic, and search performance.
Lantern Analytics connects AI visibility, citations, prompts, traffic, and search performance without treating them as the same metric.
## AI visibility
AI visibility measures how the brand appears across the tracked prompt cohort and configured providers. It is not website traffic, revenue, or a guarantee of future citation.
## Citations
Citations show which owned, competitor, and independent sources were referenced in tracked AI answers.
## Prompt analysis
Prompt analysis shows the answer and mention patterns associated with individual tracked questions.
## Traffic
Traffic uses connected Google Analytics data. It is not populated by an AI visibility run.
## Search Console
Search Console uses the connected Google property for queries, pages, clicks, impressions, CTR, and position. It is not populated by an AI visibility run.
## Comparing periods
Always confirm the date range, prompt cohort, provider set, and connected-data availability before interpreting movement. Changes in the tracked cohort can affect comparability.
# Prompt analysis
Source: https://docs.asklantern.com/analytics/prompt-analysis
***
title: "Prompt analysis"
description: "Connect tracked buyer questions to observed AI answers, mentions, and citations."
Prompt Analysis connects tracked buyer questions to the AI answers, brand mentions, competitors, and citations observed for each prompt.
## Prompt library
Organize prompts by topic and buyer intent. Use questions real buyers ask, not keyword fragments written only for scoring.
## Cohort changes
Changes to the tracked prompt cohort can affect trend comparability. Review proposed cohort changes before applying them and preserve the baseline used for later verification.
## Recommended coverage
Cover category discovery, use cases, comparisons, alternatives, risks, implementation, and direct brand questions. Avoid near-duplicate prompts that overweight one topic.
# Traffic
Source: https://docs.asklantern.com/analytics/traffic
***
title: "Traffic"
description: "Review website activity from a connected Google Analytics property, including AI referrals."
Traffic shows website activity from the connected Google Analytics property, with emphasis on referrals from AI assistants and answer engines.
## Requirements
Connect Google Analytics and select the correct property. Data availability and attribution depend on the source values recorded by that property.
## What to review
\- Total sessions and movement over the selected period.
\- AI referral sessions.
\- Landing pages.
\- Acquisition sources.
\- Geographic and engagement context where available.
AI referral classifications are based on recorded referral information. They do not capture every influence that occurred before a visit.
# Beam
Source: https://docs.asklantern.com/beam/overview
Apply bounded agent actions across structured rows of work.
Beam applies agent actions across structured rows. Use Beam when the same bounded task must run across many pages, keywords, records, or prompts.
## Create a grid
Start from a blank grid, a CSV file, a sitemap, or Google Search Console data.
## Add agent columns
Agent columns transform row inputs into durable outputs. Available actions include SERP research, content briefs, article drafts, page refreshes, meta titles, and brand-voice QA.
## Review usage before running
Each selected cell reserves one included agent run. Beam shows the selected cells and expected allowance before starting the batch.
## Batch limits
Your plan controls the maximum cells per batch. The platform supports grids up to 100 columns and 1,000 rows, subject to payload and plan limits.
## Failures and retries
Cells report their own status. Retry only the failed cells you want to run again. Manual retries consume additional runs.
# Knowledge bases
Source: https://docs.asklantern.com/brand-kit/knowledge-bases
***
title: "Knowledge bases"
description: "Add approved sources that give authorized Lantern work reliable brand context."
Knowledge bases give agents approved internal sources to retrieve when public web evidence is not enough.
## Supported sources
Add webpages or upload PDF, DOCX, and TXT files.
## Processing states
\- Pending: waiting to begin.
\- Processing: content is being extracted and prepared.
\- Ready: available for retrieval.
\- Failed: processing did not complete.
## How retrieval works
Lantern divides sources into searchable passages and retrieves relevant context for an authorized agent request. A knowledge base does not force every passage into every request.
## Good source hygiene
Use current product documentation, approved claims, brand guidance, research, and case studies. Remove obsolete sources and avoid uploading credentials or secrets.
# Brand Kit
Source: https://docs.asklantern.com/brand-kit/overview
Maintain the approved brand context Lantern uses for analysis and marketing work.
Brand Kit contains the approved context Lantern uses to understand and represent a brand.
## Brand foundations
Maintain the name, website, description, industry, size, aliases, audience, value proposition, products, and positioning.
## Voice and writing
Define persona, tone, writing style, preferred language, words to avoid, and approved examples.
## Market context
Review competitors and topics used for analytics, research, and planning.
## Visual preferences
Manage colors, fonts, logos, and presentation defaults where available.
## Archiving a brand
The current Delete action archives the brand workspace. It does not represent immediate permanent deletion of every associated record. Contact support for data-deletion requests until a dedicated deletion workflow is documented.
# Content
Source: https://docs.asklantern.com/content
***
title: "Content and CMS drafts"
description: "Review imported content, generated drafts, refresh workflows, and connected CMS behavior."
Content availability depends on the connected CMS and the workflows available to your workspace.
## Imported content
When supported by the connected CMS, Lantern can show existing content for review and planning. Availability depends on the connection, account access, and CMS configuration.
## Generated drafts
Specialists can produce content drafts and related artifacts, such as briefs, metadata, and FAQs. Review each output before moving it into an external publishing workflow.
## Refresh workflows
Use a refresh workflow to assess an existing page, identify opportunities, and prepare a revised draft. The result does not automatically publish changes.
## Publishing
Publishing behavior depends on the CMS connector, workflow configuration, permissions, and approval boundary. Contentful uses a draft-first workflow. See [Integrations](/integrations/overview) for connector guidance.
# Lantern Docs
Source: https://docs.asklantern.com/index
Understand AI search visibility, make evidence-backed marketing decisions, and create reviewable work with Lantern.
## Lantern documentation
Lantern is a governed AI marketing platform that connects brand context, AI-search intelligence, marketing decisions, and specialist execution.
Use Lantern to understand what matters, decide what to do next, create reviewable work, and measure what changed.
## Start with LanternIQ
LanternIQ is the conversational intelligence workspace on Home. Ask a question about your brand, visibility, traffic, search performance, content, competitors, or current work.
LanternIQ can answer from approved brand context and connected data, investigate a question with evidence, recommend a measurable next action, prepare work for review, and start bounded specialist work when you request a deliverable.
LanternIQ does not publish, spend, send outreach, change integrations, or make other consequential external changes directly.
## Explore the platform
Measure AI visibility, citations, prompts, traffic, and search performance.
Launch specialists for research, content, SEO, presentations, and other defined outcomes.
Find durable outputs, reviews, and execution history.
Maintain the context that guides Lantern's work.
Bring Lantern visibility, brand intelligence, and project context into Claude.
Browse every specialist agent and what each one delivers.
## Recommended first steps
1. Complete onboarding and review the generated brand context.
2. Connect Search Console and Google Analytics when available.
3. Review the initial AI visibility baseline and prompt library.
4. Ask LanternIQ what deserves attention first.
5. Review the evidence and approve or refine the proposed work.
# Contentful
Source: https://docs.asklantern.com/integrations/contentful
Connect Contentful to create structured content drafts from approved Lantern workflows.
Connect Contentful to create structured content drafts from approved Lantern workflows.
## Before you connect
Choose the Contentful space, environment, locale, and content type used for drafts. Confirm the field mapping for title, slug, body, excerpt, SEO metadata, publication date, and tags where those fields exist.
## Draft-first behavior
The current Lantern Contentful adapter creates drafts and rejects live publication. Review and publish the entry from Contentful using your normal editorial process.
## Required access
The connection uses Contentful content-management authorization. Use an account permitted to manage the intended space and environment.
## Troubleshooting
If a draft cannot be created, verify the space, environment, locale, content type, required fields, field types, and connection authorization. Reconnect when the authorization has expired or been revoked.
# Integrations
Source: https://docs.asklantern.com/integrations/overview
Connect approved data sources, publishing systems, communication channels, and advertising accounts.
Integrations connect Lantern to approved data sources, publishing systems, communication channels, and advertising accounts.
| Integration | Primary use | Typical access |
| - | - | - |
| Google Search Console | Search queries and page performance | Read only |
| Google Analytics | Traffic and AI referrals | Read only |
| Google Ads | Campaign context and supported campaign workflows | Read and workflow-dependent write |
| Contentful | CMS content drafts | Content management |
| Sanity | Structured CMS documents | Token-defined write access |
| WordPress | Content drafts or publication | Site-defined application access |
| HubSpot | Blog content | OAuth content access |
| Slack | Selected notifications | Incoming webhook |
## Before connecting
Confirm the account, property, site, environment, dataset, or channel. Grant only the access required for the intended workflow.
## Publishing behavior
Lantern can create CMS drafts in connected systems. Live publishing depends on connector configuration, workflow policy, user permissions, and an approval boundary. Do not assume that connecting a CMS enables unattended publication.
## Disconnecting
Disconnecting prevents future access but does not automatically remove data already imported into Lantern or work already created in the external system.
# Conversations and context
Source: https://docs.asklantern.com/lanterniq/conversations-and-context
Understand how LanternIQ conversations preserve context, background work, and review history.
LanternIQ conversations are durable workspaces. Messages, investigations, evidence, approvals, linked specialist runs, and initiative context remain available when you return.
## Conversation controls
Use Conversation history to reopen prior threads. You can also start a new conversation, archive the current thread, load earlier messages, or return to the originating product surface.
Leaving a conversation does not cancel background work.
## Product context
When LanternIQ opens from Analytics, Citations, Traffic, Search Console, Site Audit, Projects, or Home, the conversation can retain that surface as context. Remove a context chip when it is no longer relevant.
## Brand context
LanternIQ uses the active brand and your authorized access. It can incorporate approved Brand Kit information, connected integrations, knowledge sources, current initiatives, and approved strategic decisions.
## Conversation memory
Recent messages remain available verbatim. Older conversation content is summarized into bounded memory, and approved strategic decisions are retained separately.
Chat text is not automatically treated as an approved brand fact. Use Brand Kit or an approved strategic decision for information that must guide future work.
## Stop, retry, and background work
* Continue in background leaves the conversation while work continues.
* Stop run requests cancellation of the active LanternIQ run and keeps the conversation.
* Retry submits the latest user request as a new turn.
* Reconnect resumes live updates when the browser connection was interrupted.
Stopping a LanternIQ turn does not guarantee cancellation of a specialist that already started as separate work.
# Evidence and investigations
Source: https://docs.asklantern.com/lanterniq/evidence-and-investigations
Review the sources and limitations behind LanternIQ recommendations.
LanternIQ uses evidence when a useful answer depends on current brand or market information.
## Evidence sources
Depending on your request and connected accounts, an investigation can review:
* Brand Kit and knowledge bases.
* AI visibility and tracked prompt results.
* Owned, competitor, and independent citation sources.
* Google Search Console.
* Google Analytics.
* Existing pages and public web sources.
* Prior Projects and approved decisions.
## Investigation progress
The investigation panel shows what LanternIQ is reviewing. Expand it to see completed sources, unavailable sources, analysis steps, and timing.
An unavailable source does not automatically invalidate the full answer. LanternIQ should explain when missing evidence materially limits the conclusion.
## Evidence quality
Lantern distinguishes current evidence from stale, insufficient, or conflicting evidence. Recommendations should distinguish observed facts from interpretation.
Outcome reporting uses "Observed after this initiative" unless the measurement design supports a causal claim.
## Citations
Web citations and stored first-party evidence remain linked to the answer that used them. Open the source list when you need to verify a statement or understand the recommendation.
# Initiatives, approvals, and measurement
Source: https://docs.asklantern.com/lanterniq/initiatives-approvals-measurement
Plan, review, execute, and measure governed marketing work in LanternIQ.
Marketing initiatives connect a signal or request to a plan, specialist run, Project artifact, and measured outcome.
## Initiative lifecycle
An initiative can move through investigation, planning, review, approval, execution, measurement, completion, cancellation, and archive. The available next state depends on the current state.
Blocked work should record the blocker and return to planning when the plan needs revision.
## Review cards
LanternIQ creates review cards for governed changes such as a new initiative, an initiative revision, a strategic decision, an execution approval, or a tracked prompt cohort.
Review the summary, expected consequences, evidence, owner, success metric, and any allowance impact. Add an optional note, then approve or dismiss the proposal.
## Measurement plans
A measurable initiative defines the metric, source, baseline window, verification window, target direction, and materiality threshold before execution.
Results are reported as improved, flat, declined, or inconclusive. An initiative can state why measurement is not applicable when no responsible comparison is available.
## Attribution
Lantern does not claim that an initiative caused an outcome unless the measurement design supports causal attribution. Most results describe movement observed after the initiative.
# LanternIQ overview
Source: https://docs.asklantern.com/lanterniq/overview
Use LanternIQ to investigate marketing questions, review evidence, and coordinate specialist work.
LanternIQ is the conversational intelligence and coordination layer inside Lantern.
It combines your brand context, connected data, AI-search evidence, active work, prior decisions, and Project history to help you understand what matters and move work forward.
## What LanternIQ can do
LanternIQ can:
* Answer a concise marketing question.
* Investigate using read-only evidence.
* Prepare a measurable marketing initiative for review.
* Record a strategic decision for review.
* Start a bounded specialist for a requested deliverable.
* Direct you to a dedicated workspace for actions that need one.
## How investigations work
When a question depends on current data, LanternIQ reviews the smallest relevant evidence set. The conversation shows progress, reviewed sources, unavailable sources, and the resulting recommendation.
Evidence can include approved brand context, AI visibility, citations, Search Console, Analytics, knowledge bases, pages, and web research.
## How execution works
LanternIQ coordinates work but does not perform every task itself. A specialist owns substantial deliverables such as articles, briefs, audits, presentations, or trust reports.
Reversible internal work can start when your request is clear. External or irreversible actions use a dedicated workflow and review boundary.
## What LanternIQ does not do directly
LanternIQ does not directly:
* Publish content live.
* Send outreach or notifications.
* Activate advertising spend.
* Change integration settings.
* Delete customer data.
* Bypass approval requirements.
# Lantern MCP for Claude
Source: https://docs.asklantern.com/mcp/lantern
Connect Lantern to Claude for AI search visibility, brand intelligence, and project context.
Connect Lantern to Claude to bring your AI search visibility, brand intelligence, marketing briefs, citation opportunities, and project context into the conversations where your team plans work.
Lantern MCP gives Claude access to the Lantern information your account is already authorized to view. Use it to turn current marketing intelligence into focused analysis, preparation, and next-step recommendations.
## What Lantern MCP Does
Lantern helps teams understand how their brands are represented in AI search and organize the work needed to improve that representation. The Lantern MCP connector lets Claude work with that information in context.
After connecting, you can ask Claude to:
* Identify the Lantern brands you can access.
* Review a brand's positioning and marketing context.
* Summarize AI visibility over a chosen time period.
* Prioritize citation opportunities.
* Review the latest marketing intelligence brief and pending approvals.
* Search project work by topic and retrieve a specific project item.
The connector is designed for marketing leaders, SEO and content teams, agencies, and brand teams that need clear answers from their existing Lantern intelligence.
## What Lantern MCP Does Not Do
Lantern MCP is read-only. It cannot:
* Create, edit, publish, approve, or delete Lantern data.
* Change brand settings or user access.
* Run campaigns or make changes in third-party systems.
* Access brands, projects, or other information outside the signed-in user's Lantern permissions.
Use Claude's output to inform your work, then make any changes through Lantern and your normal team workflows.
## Connect Lantern to Claude
### From the Connectors Directory
When Lantern is available in the Claude Connectors Directory:
1. Open **Customize > Connectors** in Claude.
2. Search for **Lantern**.
3. Select **Connect**.
4. Sign in to your Lantern account when prompted.
5. Review the connection details and select **Allow connection**.
6. Enable Lantern in the conversation where you want to use it.
Your organization may require an administrator to approve or enable connectors before you can connect Lantern.
### Add Lantern as a Custom Connector
If Lantern is not yet available in your Claude connector directory, you can add it as a custom connector.
#### Claude Pro or Max
1. In Claude, open **Customize > Connectors**.
2. Select **Add custom connector**.
3. Enter this server URL:
```text theme={null}
https://api.asklantern.com/mcp
```
4. Open **Advanced settings**.
5. Set authentication to **Always required**.
6. Under **OAuth client**, select **Use your own OAuth client**.
7. Enter the following Client ID:
```text theme={null}
80913b5b-a933-498c-9dc6-dad08ca66a8b
```
8. Leave the Client Secret blank.
9. Keep **Streamable HTTP** selected, then select **Add**.
10. Sign in to Lantern and select **Allow connection**.
#### Claude Team or Enterprise
1. An organization owner opens **Organization settings > Connectors**.
2. Select **Add**, then **Custom**.
3. Add the Lantern server URL and OAuth settings above.
4. Save the connector for the organization.
5. Each member opens **Customize > Connectors**, finds Lantern, and connects their own Lantern account.
Each person authorizes access to their own Lantern account. Connecting Lantern does not grant access to data a person cannot already view in Lantern.
## Before You Connect
To use Lantern MCP, you need:
* An active Lantern account.
* Access to at least one Lantern brand or project.
* A Claude plan that supports connectors or a Claude organization where an administrator has enabled connectors.
* Permission to connect third-party services if your Claude organization requires administrator approval.
If you are unsure which Lantern account to use, sign in to Lantern first and confirm that you can view the relevant brand before connecting Claude.
## Getting Started
Start by asking Claude what Lantern data is available to you:
```text theme={null}
List my Lantern brands and give me a short description of each one.
```
Then choose a brand and ask a focused question:
```text theme={null}
For [brand name], summarize the strongest AI visibility opportunity and explain why it matters.
```
Claude may ask you to choose a brand when more than one is available. You can respond with the brand name, or ask Claude to list your brands again.
## Common Workflows
### Review AI Search Visibility
Use Lantern to understand how a brand is appearing in AI search and identify where attention is needed.
```text theme={null}
Review AI visibility for [brand name] over the last 30 days. Summarize the key trend, the most important gap, and the first action I should take.
```
```text theme={null}
Compare the last 30 days of AI visibility for [brand name] with the previous period. Focus on practical implications for the marketing team.
```
### Identify Citation Opportunities
Ask Claude to turn Lantern's citation intelligence into a prioritized outreach or content plan.
```text theme={null}
Show the top citation opportunities for [brand name]. Prioritize them by likely impact and explain the recommended next step for each.
```
```text theme={null}
Find citation opportunities for [brand name] that could support a content-led authority strategy. Group them into quick wins, strategic opportunities, and items to monitor.
```
### Prepare a Brand Briefing
Bring together positioning, visibility, and current marketing context before a meeting or planning session.
```text theme={null}
Create a concise briefing for [brand name] using its Lantern brand overview, AI visibility, citation opportunities, and latest marketing brief.
```
```text theme={null}
I am preparing for a strategy meeting about [brand name]. What does Lantern indicate is working, what is at risk, and what should we decide next?
```
### Review Marketing Intelligence
Use the latest Lantern brief to turn existing intelligence into a clear set of priorities.
```text theme={null}
Review the latest Lantern marketing brief for [brand name]. Identify the three highest-value actions that do not require a new campaign.
```
```text theme={null}
Summarize the pending approvals and open questions in the current marketing brief for [brand name].
```
### Find Project Context
Search existing Lantern project work before creating new plans or content.
```text theme={null}
Find project items about [topic] for [brand name], then summarize the most recent approved direction.
```
```text theme={null}
Search Lantern projects for [brand name] for work related to [topic]. Show the most relevant items and tell me what decisions have already been made.
```
### Turn Intelligence Into a Working Plan
Ask Claude to synthesize Lantern data into a plan, while keeping the plan separate from changes in Lantern.
```text theme={null}
Based on Lantern's current visibility data, citation opportunities, and marketing brief for [brand name], propose a 30-day plan with the highest-impact actions first.
```
```text theme={null}
Use Lantern information for [brand name] to draft a meeting agenda for the marketing team. Include decisions we need to make, evidence from Lantern, and suggested owners.
```
## Available Lantern Information
Lantern MCP can retrieve the following information for the connected user:
| Area | What Claude Can Help With |
| - | - |
| Brands | Find the Lantern brands available to you. |
| Brand overview | Review positioning and marketing context for a brand. |
| AI visibility | Summarize aggregated AI visibility for a selected time period. |
| Citation opportunities | Identify and prioritize citation opportunities. |
| Marketing brief | Review the latest marketing intelligence brief and pending approvals. |
| Projects | Search project items and retrieve the content of a specific accessible item. |
Claude selects the appropriate Lantern information based on your request. For best results, include the brand name, the question you want answered, and the outcome you need.
## Tips for Better Results
* Start with a clear brand name when you work with multiple brands.
* Ask for a specific time period when reviewing visibility.
* State the audience for the answer, such as a leadership team, content team, or client.
* Ask Claude to distinguish facts from recommendations.
* Use Lantern project search before starting a new strategy, brief, or content plan.
* Request concise output when preparing for a meeting and more detailed analysis when planning work.
For example:
```text theme={null}
For [brand name], give me a five-bullet executive summary of the current AI search position. Separate Lantern facts from your recommendations.
```
## Permissions and Data Access
Lantern MCP uses the Lantern account you sign in with during connection.
Claude can only retrieve information that account is already authorized to access. This includes the brands and project items visible to you in Lantern. If a brand or project does not appear in Claude, check your access in Lantern before retrying.
The connector is read-only. Claude does not receive permission to change Lantern data when you connect it.
Review the connection details before selecting **Allow connection**. You can remove access at any time.
## Manage or Remove Access
To disconnect Lantern from Claude:
1. Open **Customize > Connectors** in Claude.
2. Select Lantern.
3. Choose **Remove** or disconnect the connector.
You can also remove an authorized connection from Lantern:
1. Open Lantern.
2. Go to **Settings > Connected Apps**.
3. Find the Claude connection.
4. Revoke access.
Revoking access prevents Claude from using Lantern on your behalf. You can reconnect later and approve access again.
## Privacy and Security
Only connect Lantern to Claude when you trust the Claude account and workspace you are using.
* Do not include passwords, API keys, access tokens, or other credentials in your prompts.
* Review the information Claude returns before sharing it outside your organization.
* Use the connector with the Lantern account that has the intended level of access.
* Disconnect the connector when it is no longer needed.
For Lantern's data practices, see the [Lantern Privacy Policy](https://www.asklantern.com/privacy). For the terms that govern Lantern, see the [Lantern Terms](https://www.asklantern.com/terms).
## Troubleshooting
### I cannot find Lantern in the directory
Lantern may not yet be enabled for your Claude plan, workspace, or region. Ask your Claude organization administrator whether third-party connectors are available and approved. You can use the custom connector setup when it is available to your plan.
### Claude says the connector cannot connect
Confirm that the server URL is exactly:
```text theme={null}
https://api.asklantern.com/mcp
```
For a custom connector, confirm that you selected **Always required**, added the listed OAuth Client ID in Advanced settings, left the Client Secret blank, and selected Streamable HTTP.
Then remove the connector, add it again, and complete the Lantern sign-in flow in the same browser session.
### I signed in, but the connection did not finish
Return to Claude's connector settings and select **Reconnect** for Lantern. If the problem continues, record the time of the attempt and any Claude reference ID shown in the error, then contact Lantern support.
Do not send passwords, access tokens, Client Secrets, or other credentials to support.
### A brand or project is missing
Claude only sees the Lantern data available to the signed-in account. Open Lantern and confirm that you can access the relevant brand or project. If access was recently granted, reconnect the connector and try again.
### Claude's answer does not match what I expected
Ask Claude to identify the Lantern information it used, narrow the request to a specific brand or period, or request a factual summary before asking for recommendations.
For example:
```text theme={null}
For [brand name], summarize only the Lantern data available for the last 30 days. Do not make recommendations yet.
```
## Support
When contacting Lantern support about the connector, include:
* The approximate date and time of the issue.
* Whether you were adding, reconnecting, or using the connector.
* The brand or workflow you were trying to use, without sharing sensitive content.
* Any Claude reference ID shown in the error message.
Do not include access tokens, passwords, Client Secrets, API keys, or other credentials.
## Frequently Asked Questions
### Does Lantern MCP change my Lantern data?
No. The connector is read-only. It can retrieve authorized information but cannot create, edit, publish, approve, or delete Lantern data.
### Can everyone in my company see the same Lantern information?
No. Each person connects their own Lantern account. Claude can only retrieve the information that the connected account is allowed to access.
### Can I use Lantern MCP in any Claude conversation?
Enable Lantern for the conversation where you want to use it. Your Claude plan or organization settings may determine where connectors are available.
### Can I disconnect Lantern later?
Yes. Remove Lantern from Claude's connector settings or revoke the Claude connection from Lantern's **Connected Apps** settings.
### Does Claude automatically take action based on Lantern data?
No. Lantern MCP provides read-only context. Claude can analyze and recommend, but it does not make changes to Lantern or publish work on your behalf through this connector.
## Example Team Prompts
### Leadership
```text theme={null}
Give me an executive-ready update on [brand name]. Summarize AI visibility, the most important citation opportunity, and the decision we should make this week.
```
### Content and SEO
```text theme={null}
For [brand name], use Lantern to identify the most important AI search visibility gap. Recommend a content brief direction, but clearly separate Lantern facts from your proposed approach.
```
### Agency Account Management
```text theme={null}
Prepare a client check-in summary for [brand name] using the latest Lantern information. Include progress, opportunities, risks, and questions for the client.
```
### Marketing Operations
```text theme={null}
Search Lantern project items for [brand name] related to [topic]. Summarize the current status, prior decisions, and the next coordination step.
```
## Learn More
Visit [Lantern](https://www.asklantern.com/) to learn more about AI search visibility and marketing intelligence.
# Onboarding
Source: https://docs.asklantern.com/onboarding
Create your organization, first brand, and initial marketing context.
Onboarding creates your organization, first brand, and initial marketing context.
## Choose your workspace type
Select a direct workspace when you are marketing one organization. Select an agency workspace when you manage work across client brands.
## Add your first brand
Enter the brand website and your profile details. Lantern uses the website to prepare draft context, including a description, industry, topics, competitors, and tracked prompt ideas.
## Review generated context
Generated context is a starting point, not an approved fact set. Review and correct the brand name and website, description and positioning, products or services, audience and voice, competitors, topics, and prompt ideas.
## Connect data when ready
Search Console, Analytics, and CMS connections are optional during onboarding. Connect them later from Settings if another owner needs to approve access.
## Agency onboarding
Agency workspaces separate client brands while giving authorized agency members portfolio access. Add client users only after the client workspace and access level are ready.
# Pricing
Source: https://docs.asklantern.com/pricing
Compare Lantern plans, AI visibility coverage, agent runs, workspaces, and support.
Choose the plan that fits your team, client workspace, and execution needs.
## Plan comparison
| Plan | Price | Billing | Best for |
| - | - | - | - |
| Free | \$0 | Free forever | Seeing how your brand appears in ChatGPT and trying Lantern's marketing agents |
| Pro | \$89/month | Monthly or annual billing | Tracking AI visibility and turning insights into marketing work with your team |
| Agency | \$399/month | Billed monthly | Managing AI visibility and marketing work across a client portfolio, with Pro-level allowances for each client workspace |
| Enterprise | Custom | Tailored to your organization | Enterprise security, tailored capacity, and hands-on support |
## Included limits
| Capability | Free | Pro | Agency | Enterprise |
| - | - | - | - | - |
| Brands or client workspaces | 1 brand | 3 brands | 3 client workspaces | Custom |
| Seats | 1 seat | 3 team members | 5 agency seats | Custom |
| Agent runs | 10 per month | 100 per month | 100 per client workspace per month | Custom |
| Beam batch size | Not included | 100 cells | 100 cells per client workspace | Custom |
| Integrations | Not included | Available | Available | Available |
| Brand knowledge base | 1 | Unlimited | Unlimited | Unlimited |
| ChatGPT Ads | Not included | Included | Included | Included |
| Single sign-on | Not included | Not included | Not included | Included |
| Customer success manager | Not included | Not included | Not included | Dedicated |
## AI visibility
Free includes ChatGPT visibility monitoring. Pro, Agency, and Enterprise include tracking across ChatGPT, Claude, Google, and Gemini.
Each Agency client workspace includes the same Pro-level execution limits: 100 agent runs per month, Beam batches of up to 100 cells, available integrations, unlimited brand knowledge bases, and ChatGPT Ads.
## Support
Priority support is included with Pro, Agency, and Enterprise. Enterprise includes email, dedicated Slack, and a dedicated specialist with a 24-hour SLA.
## Trial
Trials include one brand, one seat, 50 agent runs for the trial period, and Beam batches of up to 10 cells. Integrations, knowledge bases, ChatGPT Ads, single sign-on, and priority support are not included in a trial.
## Understanding usage
An agent run is one accepted specialist-agent execution. Continuation turns and infrastructure retries do not use another included run. A customer-triggered retry uses another run.
Beam batch size is the maximum number of cells that can be selected and run in a single Beam batch.
Seats include current members and outstanding invitations. Included agent runs are tracked against the current billing period.
## Agency add-ons
Agency includes three client workspaces and five agency seats. Each client workspace receives its own Pro-level execution allowance. Add-ons expand workspace or seat capacity only. They do not increase the per-workspace agent-run or Beam limits.
| Add-on | Price | Billing | Effect |
| - | -: | - | - |
| Extra client workspace | \$299 | Per month, per workspace | Adds one client workspace |
| Extra agency seat | \$29 | Per month, per seat | Adds one agency seat |
Your monthly Agency total is the base price plus any selected client-workspace and seat add-ons.
## Enterprise
Enterprise plans are tailored by contract. Contact [Lantern support](mailto:collins@asklantern.com) to discuss your requirements.
# Projects
Source: https://docs.asklantern.com/projects
A library of all outputs produced by your agents.
Projects is a library of every final output your agents have produced — articles, presentations, ad campaigns, SEO reports, research documents, and PR materials.
## Project types
| Type | Description |
| - | - |
| Article | Blog posts, landing pages, thought leadership |
| Presentation | Slide deck outlines and scripts |
| Ads | Google Ads campaign copy and structure |
| PR & Comms | Press releases, pitches, announcements |
| SEO Report | Full site audits and keyword reports |
| Research | Market, competitor, or topic research |
## Project statuses
| Status | Meaning |
| - | - |
| Draft | AI has produced output; not yet reviewed |
| Ready | Reviewed and ready to publish |
| Approved | Approved by a team member |
| Published | Live on your website or platform |
| Archived | Kept for reference but no longer active |
## Opening a project
Click any project row to open it. This takes you to the agent run that produced it, where you can read the full output, edit it in the rich-text editor, and publish it to your CMS.
# Quick start
Source: https://docs.asklantern.com/quickstart
Set up a brand, establish an AI-search baseline, and begin useful work.
Set up a brand, understand its current AI-search position, and start useful work.
## 1. Create your workspace
Choose whether you are setting up Lantern for your own organization or for agency clients. Add your profile details and primary brand website.
Lantern researches the website to prepare initial brand context, topics, competitors, and suggested prompts. Review this information before relying on it.
## 2. Review Brand Kit
Confirm the brand name, website, description, positioning, audience, voice, competitors, and topics. Add supporting knowledge sources when the default website context is not enough.
## 3. Connect data
Connect the sources relevant to your work:
* Google Search Console for queries, impressions, clicks, CTR, and positions.
* Google Analytics for sessions, pages, acquisition, and AI referrals.
* A supported CMS for content drafts and publishing workflows.
* Slack for selected workflow and approval notifications.
## 4. Establish an AI visibility baseline
Open Analytics and review the tracked prompt set. Run or schedule visibility analysis only after the prompt set represents real buyer questions for the brand.
## 5. Ask LanternIQ for the next action
Try this request:
```text theme={null}
Review the latest visibility, citation, search, and brand evidence. What should we prioritize this week, and why?
```
Expand the investigation to see the sources reviewed. If LanternIQ prepares a plan or strategic decision, review it before approval.
## 6. Start specialist work
Ask for a concrete deliverable, such as a content brief, competitor audit, FAQ, metadata set, presentation, research report, or content draft. LanternIQ starts the appropriate specialist when the required inputs are known.
Specialist work continues in the background and remains linked to the conversation, initiative, and resulting Project item.
# Release notes
Source: https://docs.asklantern.com/release-notes
Customer-visible product updates and links to updated Lantern documentation.
Use this page for customer-visible product changes only.
Each entry includes the release date, product area, what changed, who is affected, required customer action, and links to updated documentation.
Release notes do not include internal details or infrastructure-only changes.
# Search Console
Source: https://docs.asklantern.com/seo/search-console
Your Google Search Console data, surfaced directly in Lantern.
If Google Search Console is connected, this page shows your full GSC performance data inline — no need to switch between tools.
## Metrics
| Metric | Description |
| - | - |
| Total impressions | How many times your pages appeared in Google results |
| Total clicks | How many times users clicked through |
| Average CTR | Click-through rate across all queries |
| Average position | Your mean ranking across all queries |
## Tables
* **Top queries** — keywords driving the most impressions and clicks, with position and CTR
* **Top pages** — your best-performing pages with their query breakdown
* **Ranking distribution** — percentage of queries ranking in positions 1–3, 4–10, 11–20, and 20+
**Date range** defaults to the last 28 days. Change it to 7 days, 90 days, or a custom range using the date picker.
## Connecting GSC
If GSC isn't connected yet, the page shows a connection prompt.
1. Go to **Settings → Integrations**
2. Click **Connect Google Search Console**
3. Sign in with Google and select your property
See [Integrations](/settings/integrations) for full setup steps.
Connect GSC before running SEO content agents. When connected, agents pull your actual ranking data and find keyword opportunities specific to your domain — making content plans far more targeted.
# Site Audit
Source: https://docs.asklantern.com/seo/site-audit
A detailed technical SEO crawl of your website with prioritised issues.
The Site Audit shows a full technical SEO crawl of your website. Data is populated by running an agent with the site health or site crawl audit tool.
## What the audit covers
| Category | What's checked |
| - | - |
| Crawlability | Pages Googlebot cannot access |
| Indexability | Pages blocked from index or with `noindex` tags |
| Page speed | Core Web Vital issues flagged at the page level |
| Title tags & meta descriptions | Missing, duplicate, or truncated |
| Heading structure | Pages with broken H1/H2 hierarchies |
| Broken links | Internal and external 404s |
| Redirect chains | URLs with multiple hops |
| Schema markup | Pages missing structured data |
| Duplicate content | Pages with near-identical content |
## Issue severity
Each issue is categorised as:
* **Critical** — actively hurting rankings or crawlability; fix immediately
* **Warning** — impacting performance; fix soon
* **Notice** — minor issue or best-practice recommendation
Every issue includes the affected URL and a plain-English recommendation.
# Billing, allowances, and limits
Source: https://docs.asklantern.com/settings/billing-and-allowances
Understand included runs, concurrency, seats, workspaces, and subscription management.
Billing shows the current plan, subscription status, included usage, seats, workspaces, and renewal timing.
## What consumes an agent run
One accepted specialist run consumes one included run. Internal continuation turns and infrastructure retries do not consume another run. Manual retries create new work and may consume another run.
Beam consumes one run for each selected cell.
## Concurrency
Concurrency is shared across the organization. Accepted work above the plan limit waits until capacity is available.
## Manage billing
Use the Stripe billing portal to manage supported payment methods, invoices, and subscription changes.
# Connected Apps
Source: https://docs.asklantern.com/settings/connected-apps
Review and revoke external applications authorized to access Lantern.
Connected Apps shows external applications authorized to access Lantern on your behalf.
Review the client name and authorization before granting access. Revoke a connection you no longer use or recognize.
Revocation prevents future authorized requests from that client. It does not delete information previously returned to the external application.
# Team and agency access
Source: https://docs.asklantern.com/settings/team-and-agency-access
Manage organization roles, client-brand access, and agency workspaces.
Lantern supports organization roles and client-brand access.
## Organization roles
* Owner: controls the organization and highest-level administration.
* Admin: manages organization work and supported settings.
* Member: works within the access granted by the organization.
## Client brand roles
* Client admin.
* Client editor.
* Client viewer.
Client roles scope access to the relevant brand workspace. Invitations expire after seven days and count against seat limits when accepted.
## Agency workspaces
Agency users can switch between authorized client brands. Keep client data, conversations, Projects, integrations, and approvals in the correct brand context.
# Troubleshooting
Source: https://docs.asklantern.com/troubleshooting
Resolve common issues with LanternIQ, integrations, runs, knowledge sources, and Beam.
## A conversation is not updating
Use Reconnect when Lantern reports that live updates were interrupted. Reloading the page keeps a durable conversation. Continue in background when you do not need to watch the active run.
## A source is unavailable
Expand View investigation to identify the unavailable source. Check the relevant integration and date range. Lantern can still answer from other evidence when the missing source does not materially change the conclusion.
## An integration stopped working
Reconnect the integration from Settings. Confirm the account, property, environment, dataset, portal, or channel and verify that the authorizing user still has access.
## A run is queued
Your organization may be at its concurrent-run limit. Queued work begins when capacity becomes available. Do not repeatedly submit the same request.
## A plan limit was reached
Review Billing for included runs, seats, workspaces, concurrency, and Beam limits. Upgrade or wait for the next billing-period reset when appropriate.
## A knowledge source failed
Confirm the source is a supported webpage, PDF, DOCX, or TXT file. Remove password protection, verify the file is readable, and try again.
## A Beam cell failed
Open the cell result and correct the input. Retry only the affected cells. Manual retries consume additional runs.
## Contact support
Include the approximate time, product area, brand, action attempted, and any visible reference ID. Never send passwords, access tokens, API keys, client secrets, or raw credentials.
# Who Lantern is for
Source: https://docs.asklantern.com/who-its-for
Lantern is built for marketing teams that need to win AI and organic search without growing headcount.
**CMO · VP Marketing · Head of Marketing · Head of Growth**
You need to scale pipeline without scaling headcount. Your current tool stack is fragmented, CAC is rising, and AI search is reshaping how buyers discover products — fast. Lantern gives your team the leverage to compete in that environment without adding bodies or budget.
**Head of Demand Gen · VP Growth · SEO Director · Content Director**
You own traffic and pipeline, and you're already hitting the ceiling of what your current SEO and content stack can do. AI search is both a threat to your existing rankings and the biggest acquisition opportunity in years. Lantern is built for exactly this moment — giving you the tools to act on it at scale.
**SEO Managers · Content Marketers · Growth Marketers · PR Managers · Performance Marketers**
You're the ones doing the work. Research, briefs, writing, optimising, publishing — across Ahrefs, Notion, a CMS, and whatever else your team has stitched together. Lantern replaces that entire loop with a single agent-driven workflow. You describe the goal, the agent executes it end-to-end, and you review before anything goes live.
## What each role uses Lantern for
| Role | Primary use |
| - | - |
| CMO / VP Marketing | Visibility into AI search performance; exec-level reporting |
| Head of Growth | AI citation monitoring, competitor gap analysis, pipeline attribution |
| SEO Director | Site audits, crawl reports, keyword research, content briefs |
| Content Director | End-to-end content pipelines from keyword to published article |
| Growth Marketer | Landing pages, competitor gap analysis, Google Ads copy |
| PR Manager | Press releases, journalist pitches, executive briefings |
| Performance Marketer | Ad copy, campaign structure, targeting briefs |