How to Turn One Content Performance Export Into a Repurposing Priority Map With Gemini

The problem and who this is for

This workflow is for content strategists, creators, channel managers, and agency reporting teams who already have the core source material and need to turn one content performance export into a repurposing priority map, a priority memo, and a next-batch shortlist without rebuilding the same message from scratch. The goal is to use one approved source as the source of truth, then split it into channel-ready assets with the fewest steps possible.

Prerequisites

• Access to Gemini
• A csv or xlsx export with one row per content asset
• A place to save the finished outputs, such as a Google Doc, notes app, CMS draft, slide outline, or scheduler
• A human review pass before publishing anything outward-facing
• A clean export with the columns you actually trust

How to capture or gather the source material

1. Export the data as CSV or XLSX with one row per asset. Avoid merged cells, colored highlights, and stacked headers.
2. Keep the core columns that actually drive reuse decisions, such as title, format, publish date, views, clicks, saves, watch time, retention, conversions, or replies.
3. Normalize percentage columns so they are clearly percentages and not mixed decimal formats.
4. Remove duplicate rows, test rows, and assets that are too new to judge fairly.
5. Add one optional notes column if there is important context, such as paid promotion, seasonality, or a known distribution problem.

Step-by-step workflow

1. Upload a CSV or XLSX export with one row per content asset into Gemini. On mobile workflows, take or add the photo first, then explain what the image represents before asking for outputs.
2. Ask for extraction before drafting. For files, that means useful rows, patterns, or priorities. For photos, that means a clean reading of the board or notes.
3. Have Gemini create the repurposing priority map, priority memo, and next-batch shortlist in separate sections so you can accept or reject each one independently.
4. Do a quick verification pass against the source, especially if the source is visual or tabular. Numbers and handwritten text deserve extra scrutiny.
5. Ask for a final formatting pass that matches your real destination, such as carousel slides, a memo, a checklist, or a shortlist.
6. Save the final outputs outside the chat if they will be reused next week or handed to another person.

Tool-specific instructions

Primary tool: Gemini

• Gemini is a strong fit when the input already lives in a spreadsheet and the job is to turn performance data into a practical next-step decision map.
• Gemini is useful when the source already lives in the Google stack or starts as a photo, file, or spreadsheet rather than a polished text document.
• Be explicit about what the uploaded file or photo represents before you ask for analysis or drafting.
• When the source is visual, ask Gemini to transcribe or extract first, then turn the cleaned extraction into the final outputs.
• Keep the output tied to the source. Ask it to flag any uncertain reading rather than guess.

Alternative: ChatGPT

• Use ChatGPT when you already have a clean source file and mainly want fast rewriting or format conversion.
• It is a practical fallback when you need quick iteration on tone, length, or platform-specific packaging.

Alternative: Claude

• Use Claude when the source is dense and you want cleaner prose, calmer structure, or better long-form summarization.
• It is a good fallback when the first output feels too compressed or too social-first.

Copy and paste prompts

Primary repurposing prompt

{
  "role": "You are a content performance analyst.",
  "goal": "Turn one content performance export into a repurposing priority map and next-batch memo.",
  "rules": [
    "Use only the uploaded spreadsheet.",
    "Do not invent performance explanations that are not supported by the data.",
    "Separate observations from recommendations."
  ],
  "analysis_steps": [
    "Identify the rows or content types with the strongest reuse signals.",
    "Group assets by reuse potential, not just raw views.",
    "Recommend what should become an article, clip, thread, carousel, or email next."
  ],
  "output_format": {
    "sections": [
      "data_quality_notes",
      "priority_map",
      "top_assets_to_repurpose_next",
      "short_memo"
    ]
  }
}

Final packaging prompt

{
  "role": "You are formatting an internal decision memo.",
  "goal": "Package the spreadsheet analysis into a short memo a content lead can act on this week.",
  "rules": [
    "Keep the recommendations tied to the uploaded data.",
    "Avoid vanity-metric language unless the data clearly supports it.",
    "Keep the final memo concise and ranked."
  ],
  "return_format": {
    "priority_map_table": [
      "asset_or_format",
      "why_it_should_be_reused",
      "best_next_format",
      "effort_level",
      "priority_rank"
    ],
    "one_page_memo": "final"
  }
}

Quality checks

• Every important claim, quote, or metric still matches the source.
• The repurposing priority map, priority memo, and next-batch shortlist do not all sound like copies of one another.
• The outputs are short enough and structured enough to use in real work without another full rewrite.
• Any numbers, names, dates, or client details have been checked manually.
• You have spot-checked the extraction layer before trusting the rewrite layer.

Common failure modes and fixes

• The outputs all sound the same: Give each deliverable its own audience, length, and job-to-be-done before you request the rewrite.
• The tool makes the source too generic: Ask for an extraction pass first and tell it to preserve the strongest phrases, proof points, and examples.
• The recommendations do not match the numbers: Normalize the spreadsheet first and ask the model to show its reasoning per row or content type.
• The data includes too many weak rows: Filter out test content, duplicates, and assets that are too new to judge fairly.

Sources Checked

• Gemini Apps Help: Upload and analyze files in Gemini Apps: https://support.google.com/gemini/answer/14903178?hl=en (accessed 2026-03-25)
• Gemini Apps Help: Gemini Apps Help Center: https://support.google.com/gemini/?hl=en (accessed 2026-03-25)
• OpenAI Help: File Uploads FAQ: https://help.openai.com/en/articles/8555545-file-uploads-faq (accessed 2026-03-25)
• OpenAI Help: ChatGPT Image Inputs FAQ: https://help.openai.com/en/articles/8400551-chatgpt-image-inputs-faq (accessed 2026-03-25)
• OpenAI Help: Prompt engineering best practices for ChatGPT: https://help.openai.com/en/articles/10032626-prompt-engineering-best-practices-for-chatgpt (accessed 2026-03-25)
• OpenAI Help: ChatGPT Capabilities Overview: https://help.openai.com/en/articles/9260256-chatgpt-capabilities-overview (accessed 2026-03-25)
• Claude Help Center: Uploading files to Claude: https://support.anthropic.com/en/articles/8241126-uploading-files-to-claude (accessed 2026-03-25)
• Claude Help Center: What are projects?: https://support.anthropic.com/en/articles/9517075-what-are-projects (accessed 2026-03-25)
• Claude Help Center: Create and edit files with Claude: https://support.anthropic.com/en/articles/12111783-create-and-edit-files-wit (accessed 2026-03-25)

Quarterly Refresh Flag

Review by 2026-06-23. Re-check the current tool interface, upload behavior, supported file types, and any workflow changes that affect this article before republishing or refreshing it.