Skip to main content

Persona Slide Deep Dive Guide

Table of Contents​

  1. Feature Overview
  2. Use Cases
  3. Data Preparation
  4. Operation Workflow
  5. Result Interpretation
  6. Export Functions
  7. Frequently Asked Questions

Feature Overview​

What is Persona Slide?​

Persona Slide is one of the three core features of this system, designed to help Google advertising sales teams generate professional two-page slides for each audience persona based on Google Ads big data analysis, used to present detailed audience characteristics, interests, values, media consumption habits, and creative recommendations to clients.

Core Value​

  1. Automated Generation: Reduces manual creation work from 2 hours to 10 minutes
  2. Bilingual Support: Automatically generates bilingual content (Chinese-English), meeting internationalization needs
  3. Image Generation: Uses Google Imagen 4.0 to automatically generate scenario illustrations, no manual design required
  4. Ready to Use: Generated slides can be directly used for client presentations, meeting Google DMS delivery standards

How It Works​

The system receives Excel files or Google Sheet links containing persona data, parses and translates the data through Google Gemini AI, and generates for each persona:

  • Slide 1 (Overview Page): Contains audience characteristics, core affinity tags, top in-market tags, interests, values, challenges, and AI-generated scenario illustrations
  • Slide 2 (Creative & Insights Page): Contains top search terms, favorite YouTube channels, sample text creatives, and sample video creatives

Use Cases​

Typical Application Scenarios​

  1. Client Proposal Preparation: Quickly generate professional slide content when presenting personas to clients
  2. Creative Strategy Development: Generate targeted copy and video creative recommendations based on persona characteristics
  3. Media Consumption Habit Analysis: Understand target audience search behavior and media preferences to optimize advertising strategies
  4. Multilingual Delivery: Generate localized persona content for clients in different regions

Applicable Data Types​

  • Persona Dimension: Different audience personas (e.g., "Esports Enthusiasts," "Professional Creators," etc.)
  • Characteristic Dimension: Demographic characteristics, occupation, income, interests, values, etc.
  • Behavior Dimension: Search behavior, media preferences, purchase intent, etc.

Data Preparation​

Data Source: Connect Benchmark (CBX)​

The data required for the Persona Slide feature comes exclusively from Google's internal Connect Benchmark (CBX) system. It shares the same data source file as the "Audience Signal Analysis" feature. This is typically a complex Excel file or Google Sheet containing dozens of worksheets.

InsightHub is specifically designed to intelligently automatically locate and extract the core data needed to generate personas from this complex data file. You don't need to manually perform any data preprocessing.

Data Format Requirements​

The system will automatically search for a specific worksheet named AffinityGeminiOutput in your uploaded file.

Required Worksheet NameDescription
AffinityGeminiOutputContains raw text content for all personas. The system reads column A of this table row by row (A1, A2, A3...), parsing each cell's content as an independent persona.

Format Requirements for Column A of AffinityGeminiOutput Worksheet:

Each cell should contain complete persona information consisting of multiple fields. Format example:

1 - The Esports Enthusiast

Demographic: 25-54, Male (60%), Female (40%), United States, United Kingdom, Canada
Occupation: Student, Entry-level professional
Income: $30,000 - $60,000
Affinity: Gaming Enthusiasts, Technology Early Adopters, Mobile App Users
In-Market: Video Game Purchases, Gaming Accessories, Streaming Services
Interests: Esports tournaments, Gaming hardware, Live streaming
Values: Competition, Community, Innovation
Challenges: Finding time for gaming, Balancing gaming with responsibilities
Top search terms: Best gaming laptop, Esports tournaments, Gaming accessories
Favorite YouTube Channels: PewDiePie, Markiplier, Ninja
Sample Text Creative:
- "Level up your gaming experience with the latest technology"
- "Join the esports community and compete with the best"
Sample Video Creative:
- Visual: Fast-paced gaming montage with competitive scenes
- Soundtrack: Energetic electronic music
- Styling: Modern, dynamic, high-energy
- Call to Action: "Start your gaming journey today"

Important Notes:

  • Intelligent Recognition: You don't need to worry about the AffinityGeminiOutput worksheet name being truncated by Excel or renamed and unable to be recognized. The system's intelligent matching algorithm can handle these situations.
  • Automatic Ignoring: All other irrelevant worksheets will be automatically ignored by the system.

Data Quality Checks​

The system automatically performs the following checks before analysis:

  1. Required Worksheet Validation: Ensures the AffinityGeminiOutput worksheet exists
  2. Data Completeness Validation: Checks whether column A contains valid persona data
  3. Field Identification: Automatically identifies and parses various fields (Demographic, Occupation, Income, etc.)
  4. Industry Name Extraction: Reads the industry name from cell B4 of the input worksheet; uses default value if empty

Operation Workflow​

Step 1: Log In to the System​

  1. Access the system URL (provided by administrator)
  2. Click "Sign in with Google"
  3. Authorize the system to access your Google account information (for authentication and Google Sheet access)

Permission Notes:

  • The system needs access to your Google account information (name, email)
  • If using Google Sheets, read-only access to Google Sheets is required
  • All permissions are used only for feature operation and will not be used for other purposes

Step 2: Select Feature​

  1. Click the "Persona Slide" card on the system homepage (Dashboard)
  2. Enter the Persona Slide page

Step 3: Upload Data​

The system supports two data input methods:

Method A: Upload Excel File​

  1. Click the "Upload Excel File" button
  2. Select your Excel file in the file selection dialog
  3. The system will automatically validate the file format and display file information

File Limitations:

  • File size: Recommended not to exceed 10 MB
  • File format: .xlsx (Excel 2007 and above)
  • Worksheet requirement: Must contain the AffinityGeminiOutput worksheet
  1. Copy Link: Directly copy the target Google Sheet's URL in your browser.
  2. Paste Link: Paste the link in the InsightHub page input box.
  3. No Sharing Required: Thanks to the Google OAuth permissions you granted during login, the system can directly and securely access spreadsheet data you have permission to view. You don't need to configure any "sharing" settings, such as "Share with anyone who has the link."
  4. Click Load: Click the "Load" button, and the system will begin reading data.

Permission Check:

  • The system will automatically verify whether your currently logged-in Google account has access to that Sheet.

Step 4: Configure Analysis Parameters​

  1. Set Query Language: Select the query language you want Gemini AI to use
  2. Set Output Language: Select the language for the final report
  3. Edit Prompt: The system has built-in targeted prompts; generally no modification needed

Step 5: Start Analysis​

  1. Click Analyze: Click the "Start Analysis" button to launch the analysis process
  2. Wait for Results: The analysis process is relatively complex and may take 5-10 minutes, please wait patiently

Results Page Details​

The Persona Slide results page is the most information-dense and visually rich page among the project's three features. It aims to comprehensively depict target user groups and directly provide creative materials usable for marketing communication, serving as a bridge connecting data insights with marketing activities.

1. Page Layout​

The results page uses a multi-tab layout, with each tab representing an independent Persona identified and generated by AI. Each Persona contains two pages of carefully designed slide content, with information progressing layer by layer from macro personas to specific marketing touchpoints.

Slide Page 1: Persona Overview​

The core goal of this page is "Who is this person?", quickly establishing a three-dimensional understanding of this user group through highly visual means.

  • 2x2 Grid Layout: The page uses a classic four-quadrant grid layout with clear structure and prominent focus.
  • AI-Generated Scenario Image:
    • Design Philosophy: The most prominent position in the top-left corner is a scenario image automatically generated by AI (Imagen 4.0 model) based on this Persona's core characteristics. This image is not a simple material collage but rather AI's understanding and creation of "what kind of life or work scenarios this type of person might appear in," greatly enhancing the persona's vividness and immersion.
    • Technical Implementation: During the analysis process, the system extracts key characteristics of this Persona (such as "tech enthusiast," "outdoor sports fan"), dynamically generates a descriptive prompt, sends it to the image generation model, and thus obtains this unique image.
  • Three Core Characteristics:
    • Demographics: The top-right corner displays the basic profile of this group, such as age, gender, geographic location, language, etc.
    • Top Affinity: The bottom-left corner lists the most prominent interest tags of this group, helping understand what they love.
    • Top In-Market: The bottom-right corner reveals the product or service categories they are most likely to purchase recently, which are key signals for marketing conversion.
  • Deep Insight Supplements:
    • Below the grid, AI further provides text descriptions of this group's Interests, Values, and Challenges, revealing their inner motivations and pain points from a deeper level.

Slide Page 2: Activation & Creatives​

The core goal of this page is "How to reach and move him/her?", providing actionable (executable) strategies and materials for specific marketing activities.

  • Three-Column Layout: The page is divided into three columns: "Media Consumption," "Text Creatives," and "Video Creatives," with clear logic.
  • Media Consumption:
    • Top Search Keywords: Shows the keywords this group most frequently searches on Google, which are direct inputs for SEO and SEM strategies.
    • Top YouTube Channels: Reveals the content creators they most follow, providing precise targets for influencer marketing and video ad placement.
  • AI-Generated Marketing Creatives:
    • Design Philosophy: The two columns on the right are completely generated by Gemini AI based on all characteristics of this Persona, automatically generating Ad Copy Samples and Video Ad Script Ideas that can be directly used for advertising.
    • Value: This content greatly shortens the path from user insights to creative execution. Marketers no longer need to conceive from scratch but can directly use or optimize based on this, significantly improving work efficiency and creative relevance.

2. Interactive Features and Highlights​

  • In-Place Content Editing:
    • Like the first two features, all text elements on the results page, including AI-generated insights, interest points, values, and even ad copy and video scripts, all support click-to-edit. This gives users the power of final review and personalized adjustment.
  • Rich Export Options:
    • Download Session (ZIP): One-click package of all Personas' complete data and images.
    • Export as PPTX: Complete export of all Personas' slides (two pages per person) into a single .pptx file.
    • Export as Google Slides:
      • Unique Implementation: Uses "image insertion" mode. The system creates a brand new blank Google Slides document, then inserts each slide rendered on the frontend (including AI-generated images and complex layouts) as high-resolution images one by one into the new document.
      • Advantage: This approach can 100% guarantee that the final exported Google Slides exactly matches what you see on the webpage, perfectly solving the complex layout restoration problem.
    • Single Page Export/Copy: Also supports exporting any single slide as an image or copying to clipboard.
  • AI Image Refresh:
    • In the top-right corner of the AI-generated image on the first page, there is a "Refresh" button. If you're not completely satisfied with the currently generated image, you can click this button, and the system will re-call the image generation model with the same prompt to provide you with a new visual creative.

Result Interpretation​

Slide 1: Overview Page Interpretation​

Slide 1 uses a 2x2 layout, containing the following four areas:

Top-Left: Characteristics​

Contains the following fields:

  • Age: Such as "18-34 years" / "18-34"
  • Gender: Such as "Male (75%), Female (25%)" / "Male (75%), Female (25%)"
  • Location: Such as "United States, Europe, Asia" / "United States, Europe, Asia"
  • Occupation: Such as "Student, Entry-level professional" / "Student, Entry-level professional"
  • Income: Such as "$30,000 - $60,000"

Key Interpretation Points:

  • All fields include bilingual content (Chinese-English)
  • Data is based on Google Ads big data analysis
  • Used to understand basic demographic characteristics of target audiences

Bottom-Left: Scenario Illustration (Image)​

  • Content: AI-generated 1:1 aspect ratio scenario illustration (250px x 250px)
  • Generation Method: Automatically generated using Google Imagen 4.0 based on persona characteristics
  • Purpose: Intuitively displays typical scenarios and lifestyle of this persona

Image Characteristics:

  • Style: Modern, professional, meeting business presentation standards
  • Content: Reflects core characteristics and interests of the persona
  • Quality: High resolution, suitable for printing and presentation use

Top-Right: Top Affinity Tags​

  • Content: 3-5 core affinity tags
  • Format: Bilingual (Chinese-English), such as "Gaming Enthusiasts" / "Gaming Enthusiasts"
  • Purpose: Used for brand building and user cultivation

Key Interpretation Points:

  • Tags are based on Google Ads big data analysis
  • Indicate users' long-term interest in this field
  • Suitable for brand advertising and content marketing

Bottom-Right: Top In-Market Tags​

  • Content: 3-5 top in-market tags
  • Format: Bilingual (Chinese-English), such as "Video Game Purchases" / "Video Game Purchases"
  • Purpose: Used for conversion optimization and sales promotion

Key Interpretation Points:

  • Tags are based on Google Ads big data analysis
  • Indicate users are actively seeking related products or services
  • Suitable for performance advertising and conversion optimization

Bottom: Interests, Values, Challenges​

  • Interests: 3-5 interest points, such as "Esports tournaments, Gaming hardware, Live streaming"
  • Values: 3-5 values, such as "Competition, Community, Innovation"
  • Challenges: 3-5 challenges, such as "Finding time for gaming, Balancing gaming with responsibilities"

Key Interpretation Points:

  • All content includes bilingual (Chinese-English) content
  • Used to understand psychological characteristics and behavioral motivations of target audiences
  • Can be used to develop more precise marketing strategies

Slide 2: Creative & Insights Page Interpretation​

Slide 2 uses a 3-column layout, containing the following three areas:

Left: Media Habits​

Contains the following content:

Top Search Terms:

  • Content: 5-10 top search terms
  • Format: Bilingual (Chinese-English)
  • Purpose: Understand target audience search behavior, optimize keyword strategy

Favorite YouTube Channels:

  • Content: 3-5 YouTube channel names
  • Format: English only (preserving original channel names)
  • Purpose: Understand target audience media preferences, optimize video ad placement

Key Interpretation Points:

  • Data is based on Google Ads big data analysis
  • Can be used to develop content marketing and video advertising strategies
  • Helps select suitable media channels and KOL collaborations

Middle: Sample Text Creative​

  • Content: 3-5 text creative samples
  • Format: Bilingual (Chinese-English)
  • Purpose: Provide inspiration and reference for ad copy creation

Key Interpretation Points:

  • Creative samples are generated based on persona characteristics and interests
  • Match the language style and values of this persona
  • Can be directly used for ad copy creation or as reference

Right: Sample Video Creative​

Each video creative sample contains the following four dimensions:

  • Visual: Description of video visuals, such as "Fast-paced gaming montage with competitive scenes"
  • Soundtrack: Background music style, such as "Energetic electronic music"
  • Styling: Overall visual style, such as "Modern, dynamic, high-energy"
  • Call to Action: Action to encourage users to take, such as "Start your gaming journey today"

Key Interpretation Points:

  • All content includes bilingual (Chinese-English) content
  • Creative samples are generated based on persona characteristics and interests
  • Can be used for video ad script creation and production guidance

Export Functions​

Export Options​

The system supports multiple export formats:

1. Download Session ZIP​

Contents:

  • YAML configuration file (contains all analysis metadata)
  • XLSX detailed data tables (one worksheet per persona)
  • PPTX presentation (contains slides for all personas)
  • PNG image files (all slides and images)
  • Original data file (if Excel was uploaded)

File Naming Convention:

InsightHub-YYYYMMDD-HHmmss-{sessionUUID}-{fileTypeLabel}-{languageLabel}.zip

Use Cases:

  • Complete backup of analysis results
  • Share with team members
  • Archive storage

2. Export PPTX​

Contents:

  • Two slides per persona
  • All images and content

Use Cases:

  • Direct use for client presentations
  • Integration into existing presentations

3. Export Images​

Contents:

  • Slide PNG images (two pages per persona)
  • Scenario illustration PNG images (one per persona)

Image Quality:

  • Resolution: Automatically optimized based on screen resolution
  • Format: PNG (supports transparent background)
  • Clarity: Suitable for printing and presentation use

Use Cases:

  • Insert into other documents
  • Use for social media sharing
  • Print output

4. Copy to Clipboard​

Supported Formats:

  • Images (PNG format)
  • HTML code (for web embedding)

Use Cases:

  • Quick paste into other applications
  • Insert images in emails
  • Insert content in documents

Export Operation Steps​

  1. On the results page, find the content to export (slides or images)
  2. Click the corresponding export button ("Download," "Copy," etc.)
  3. Complete the operation according to prompts

Notes:

  • Exporting large amounts of data may take some time, please wait patiently
  • If export fails, the system will display error information; please retry according to prompts

Frequently Asked Questions​

Q1: Why did my Excel file upload fail?​

Possible Reasons:

  1. Incorrect file format (must be .xlsx format)
  2. Missing required worksheet (AffinityGeminiOutput)
  3. Incorrect column A data format
  4. File size exceeds limit
  5. File is corrupted or unreadable

Solutions:

  1. Confirm file format is .xlsx
  2. Check if the AffinityGeminiOutput worksheet is included
  3. Confirm column A contains valid persona text content
  4. Try re-saving the file in Excel
  5. Check if the file is being used by another program

Possible Reasons:

  1. Incorrect link format
  2. No access permissions
  3. Sheet has been deleted or moved
  4. Missing required worksheet (AffinityGeminiOutput)

Solutions:

  1. Confirm link format is: https://docs.google.com/spreadsheets/d/...
  2. Check the Sheet's sharing permissions
  3. Confirm the Sheet contains the AffinityGeminiOutput worksheet
  4. Try logging in to the system again

Q3: What if image generation fails?​

Possible Reasons:

  1. Image generation service temporarily unavailable
  2. Network connection issues
  3. Image generation quota exhausted

Solutions:

  1. Wait a while and retry
  2. Check network connection
  3. Contact system administrator to check image generation service status
  4. If image generation fails, slides will still be generated, just missing the image portion

Q4: How do I edit slide content?​

Operation Steps:

  1. On the results page, find the persona you want to edit
  2. Click the "Edit" button
  3. Modify card title or content in the edit dialog
  4. Click the "Save" button to save changes

Notes:

  • Edited content is only saved in browser local storage
  • If browser data is cleared, edited content may be lost
  • It's recommended to export ZIP files for backup after editing

Q5: Can I generate multiple personas simultaneously?​

Yes. The system supports:

  • Single Excel file containing multiple persona data
  • Each persona occupies one cell in column A
  • System automatically identifies and parses all personas

Recommendations:

  • If there are too many personas (>10), it's recommended to generate in batches to ensure optimal performance
  • Each persona's generation time is approximately 2-3 minutes

Q6: How do I save analysis results?​

The system automatically saves analysis results to browser local storage (IndexedDB). You can:

  1. View History: View all historical analyses in the Dashboard
  2. Export ZIP: Download the complete analysis results package
  3. Reload: Click on history records to view results again

Note: Data stored in browser local storage may be lost after clearing browser data. It's recommended to regularly export ZIP files for backup.

Q7: Can generated images be customized?​

Currently not supported. Images are automatically generated by Google Imagen 4.0 based on persona characteristics and cannot be manually specified.

Future Plans:

  • The system may support image customization features in future versions
  • You can influence image generation results by editing persona data

Technical Details (Optional Reading)​

Data Processing Workflow​

  1. File Parsing: Uses XLSX library to parse Excel files, or reads data through Google Sheets API
  2. Data Extraction: Extracts all cell content from column A of the AffinityGeminiOutput worksheet
  3. AI Parsing: Parses and translates each persona's text content through Google Gemini API
  4. Data Structuring: Converts parsed data into structured JSON format
  5. Slide Generation: Uses pure code conversion functions to convert structured data into slide content
  6. Image Generation: Generates scenario illustrations through Google Imagen 4.0 API
  7. Result Storage: Saves results to browser local storage (IndexedDB)

AI Parsing Mechanism​

The system uses Google Gemini 2.5 Pro model for the following parsing:

  1. Field Identification: Automatically identifies and extracts various fields (Demographic, Occupation, Income, etc.)
  2. Data Translation: Translates all fields to target language while preserving original English content
  3. Data Validation: Validates whether parsed data meets expected format
  4. Error Handling: If parsing fails, system uses default values or skips that field

Image Generation Mechanism​

The system uses Google Imagen 4.0 API to generate scenario illustrations:

  1. Prompt Generation: Generates image prompts based on persona characteristics and interests
  2. Image Generation: Calls Imagen 4.0 API to generate 1:1 aspect ratio images
  3. Image Optimization: Automatically adjusts image size and format to ensure suitability for slide use
  4. Error Handling: If image generation fails, system uses placeholder images

Slide Layout Algorithm​

The system uses pure code conversion functions to convert structured data into slide content:

  1. Layout Calculation: Automatically calculates optimal layout based on content length
  2. Text Formatting: Automatically formats text to ensure bilingual content displays correctly
  3. Image Embedding: Embeds generated images into slides, maintaining correct proportions and positions
  4. Style Application: Applies unified styles and themes to ensure all slides have consistent styling

Related Documentation: