绘现 (GEO) Documentation

website·Indexed 31 days ago

https://docs.geo.hkting.com/

A monitoring and optimization platform for AI search citations in the Chinese market. It helps brands track their visibility across AI assistants like Doubao, Kimi, and DeepSeek, providing tools for citation scanning, AI-driven content generation, and automated publishing to platforms such as WeChat Official Accounts, Zhihu, and Baijiahao.

Tokens
21.6K
Snippets
9
Records
155
Agent score
50%

What's inside 绘现

  1. Overview of GEO platform capabilities

    The GEO platform is organized into six primary modules designed to take users from onboarding to advanced optimization:

    1. Getting Started: Account registration, project creation, and initial scanning.
    2. Monitoring & Analysis: Real-time AI platform scanning, citation rate tracking, GEO scoring, and competitor tracking.
    3. Content Generation & Publishing: Generating GEO-optimized content and distributing it across various channels.
    4. Billing & Credits: Managing subscription plans, credit consumption, and billing transactions.
    5. Prompts & AI Q&A: Managing monitoring prompts and exploring AI-generated answers.
    6. Account & Team: Managing user profiles, team members, and referral rewards.
  2. Overview of 绘现 (HuiXian) AI Search Monitoring and Optimization

    绘现 is an AI search citation monitoring and optimization platform for the Chinese market. It allows brands to track their appearance frequency in responses from major AI assistants (such as Doubao, Kimi, and DeepSeek) and improve AI visibility through a closed-loop process of content generation and publishing.
  3. Understand GEO (Generative Engine Optimization) vs SEO

    GEO (Generative Engine Optimization) focuses on getting content cited by AI Large Language Models (LLMs) like Doubao, Kimi, and DeepSeek, where users see a direct AI-generated answer. In contrast, SEO (Search Engine Optimization) aims for high rankings in search engine result lists (e.g., Google, Baidu) where users see a list of links. While SEO optimizes for keywords and backlinks, GEO optimizes for content authority, structure, and information density.
  4. Access the Project Dashboard

    The Dashboard provides an overview of brand AI exposure health, including GEO scores and citation trends. It is available to all users with 'Viewer' permissions or higher.

    Prerequisites:

    • At least one project must be created.
    • (Recommended) At least one scan must be completed; otherwise, GEO scores and trend charts will appear empty.

    Navigate to: https://geo.hkting.com/projects -> Click a project card OR While inside a project, click the 'Dashboard' (LayoutDashboard icon) in the left navigation bar.

  5. Access the scan configuration page

    There are two ways to enter the scan configuration page after creating a project:

    1. Top Navigation Bar: Click the "Immediate Scan" (立即扫描) or "Start New Scan" (发起新扫描) button located in the center.
    2. Project Overview/Workbench: Click the "Start Scan" (发起扫描) button on the card located on the right side of the page.
  6. Manage system alerts via the Notification Center

    v1-2
    The Notification Center aggregates system messages including scan completion alerts, citation rate changes, and insufficient credit warnings. Unread message counts are displayed as badges in the navigation bar. Users can enable email notifications by configuring their email address in the account settings.
  7. Manage team credit and project sharing

    Enterprise teams operate on a shared resource model:

    • Credit Sharing: All members share a single credit pool. All scans and content generation are deducted from this balance. Administrators can monitor usage and identify who performed specific actions via Account → Billing → Consumption Details.
    • Project Sharing: All team members have access to all projects. Project-level permission isolation is currently not supported.
  8. Understand the structure of AI visibility case studies

    Case studies for improving AI visibility using the 绘现 (Huixian) platform provide practical experience and data references. Each case study is structured to include the following components:

    • Industry Background: The brand's industry, scale, and competitive environment.
    • Core Challenges: Specific issues faced in AI search, such as low citation rates or competitor dominance.
    • Solutions: The specific operational path taken using 绘现 monitoring and content optimization.
    • Effectiveness Data: Metrics showing the increase in citation rates, changes in GEO (Generative Engine Optimization) scores, and period-over-period comparisons.
  9. Review and validate AI-generated content before publishing

    AI-generated content may contain inaccurate product descriptions, outdated/fictional data, or unauthorized case studies. Before publishing, manually verify:

    1. Feature Accuracy: Compare against the latest product features.
    2. Data Authenticity: Ensure cited numbers and reports are verifiable.
    3. Case Authorization: Confirm permission for any third-party brands or user cases mentioned.
  10. Filter notifications in the Full History page

    To find specific notifications in the Full History page:

    1. Navigate to the Full History page via the bell icon $ ightarrow$ "View All Notifications".
    2. Use the Type dropdown menu to select one or more notification types.
    3. Use the Time Range selector to set the start and end dates.

    The list will refresh in real-time to show only matching records.

  11. Access and requirements for Competitor Analysis

    The Competitor Analysis feature is exclusive to the Enterprise Edition. Standard Edition users will see the entry point but it will be locked. To use this feature, you must have completed at least one scan to establish a baseline. While all users with viewing permissions can see the data, only Administrators can add, edit, or delete competitors.

    To access the feature: Navigate to the left sidebar → Brand Monitoring (品牌监测) → Competitor Analysis (竞品分析, target icon).

  12. Interpret AI citation rates and handle response randomness

    AI responses are stochastic due to the 'temperature' parameter, meaning a single manual test may not represent the actual citation probability. To get an accurate measurement, use the scanning feature which performs repeated sampling (3–10 times) per query to calculate a statistical citation rate (e.g., 3 mentions out of 5 samples equals a 60% citation rate).