One project. Three source files. One monthly report.
Add the three starter files to a ChatGPT or Claude project once. Each month, the AI reads those project files and you replace only the files it updates. No computer folder connection is required.
Choose one
How will you collect the LinkedIn numbers?
Choose whether an AI browser collects them or an AI guide walks you through them. All three routes update the same Data Bank.
Monthly reporting · Your route: Manual
Monthly reporting · Your route: I’ll pull the numbers
Step 1 of 4
Add the starter files to your AI project.
The project—not a folder on your computer—will be the home for the current Data Bank and reports.
Choose ChatGPT or Claude. Create or open a project named LinkedIn Analytics.
Download the three starter files below. They will usually appear in Downloads.
In ChatGPT, open the project’s Sources section. In Claude, open the project’s Files section.
Add all three downloaded files to the project. Confirm that each filename appears once.
Working file · XLSX
Analytics Data Bank
This spreadsheet stores each reporting period without deleting prior periods.
No folder connection is needed. The downloads are only the transfer step. After the three files appear in Project Sources or Files, you may keep or delete the downloaded copies.
Before you move on
Your LinkedIn Analytics project shows one Data Bank, one Markdown backup, and one HTML report in its Sources or Files section.
Step 2 of 4
Set up the route you chose.
You will read the numbers from LinkedIn while ChatGPT or Claude guides you one screen at a time and updates the Data Bank for you.
Use ChatGPT Work inside your LinkedIn Analytics project. Work provides the browser needed to collect the LinkedIn numbers.
Use the Claude in Chrome side panel to collect the LinkedIn numbers. Your project Files hold the Data Bank and reports.
Use Claude in Chrome to collect the LinkedIn numbers, then take one clean handoff into your Claude project. The project chat—not the extension—updates the Data Bank and reports.
Choose your AI guide
Use ChatGPT
Choose this if ChatGPT is where you already work.
Open your LinkedIn Analytics project.
Confirm that the three files appear under Sources.
Start one new chat or Work task inside the project and keep it open through Steps 3 and 4.
Use Claude
Choose this if Claude is where you already work.
Open your LinkedIn Analytics project.
Confirm that the three files appear under Files.
Start one new chat inside the project and keep it open through Steps 3 and 4.
Open your LinkedIn Analytics project on the web or in the ChatGPT app.
Confirm that the three current files appear under Sources.
Start a new task in Work from inside the project. You do not need to connect a computer folder or install a browser extension.
Open your LinkedIn Analytics project on the web or in the Claude app.
Confirm that the three current files appear under Files.
Open your LinkedIn Analytics project and confirm that all three current files appear under Files.
Install Claude in Chrome and sign in to the same Claude account. Confirm that the extension is turned on.
When you collect the LinkedIn numbers in Step 3, open LinkedIn in Chrome and select the Claude icon in the Chrome toolbar. The side panel starts a Cowork chat.
The chat will appear in your regular Claude workspace. Open its three-dot menu and add the chat to the LinkedIn Analytics project.
Keep the project open in one tab and open the correct personal LinkedIn profile in another.
In Step 3, the Chrome side panel will collect the numbers and create a copyable handoff. A chat inside the project will use that handoff to update the spreadsheet.
Work is required for the browser collection step. Start the Work task from inside the project so the task can use the files in Project Sources.
Use one task from collection through reporting. The project holds the starting files. Keep the same task or chat open through Steps 3 and 4, then replace all three project files together at the end.
Before you move on
One chat or task is open inside your LinkedIn Analytics project, the three project files are available, and LinkedIn is open in another tab.
A new ChatGPT Work task is open inside the LinkedIn Analytics project, and the three project sources are available.
Your LinkedIn Analytics project contains all three files. Claude in Chrome is installed, turned on, and signed in.
Use one project conversation from collection through reporting. Keep it open through Steps 3 and 4, then replace all three project files together at the end.
Use one Work task from collection through reporting. Start it inside the project so it can use Project Sources, then keep it open through Steps 3 and 4.
Claude uses two clear stages. The Chrome side panel only collects the LinkedIn data. The project chat handles the spreadsheet and stays open for reporting.
Before you move on
One chat or task is open inside your LinkedIn Analytics project, the three project files are available, and LinkedIn is open in another tab.
A new ChatGPT Work task is open inside the LinkedIn Analytics project, and the three project sources are available.
Your LinkedIn Analytics project contains all three files, Claude in Chrome is installed, and the correct LinkedIn profile is open in Chrome.
Step 3 of 4
Let the AI guide you through the numbers.
Ask ChatGPT to update the Data Bank.
Ask Claude to update the Data Bank.
Collect in Chrome, then update the Data Bank in your project.
This step is only for collecting and checking the numbers. You will make the report in Step 4.
What you are doingCollecting the LinkedIn numbers for one complete reporting period and adding them to the Data Bank.
What you will haveAn updated XLSX Data Bank in the current task or chat. Keep that conversation open for the report step.
What you will haveAn updated XLSX Data Bank in the AI project conversation that handles the file. Keep that conversation open for the report step.
Keep LinkedIn open in one tab and your AI project chat open in another.
Confirm that LinkedIn-Analytics-Data-Bank.xlsx is available in the project’s Sources or Files. Do not download or open it.
Paste the guided collection prompt below.
Follow one instruction at a time. Read the requested number from LinkedIn and type it into the chat. If LinkedIn does not show it, reply Not shown.
Review the final summary in the chat. Correct anything that is wrong, then confirm it.
The AI will update the Data Bank and return the XLSX. Leave this same chat open and continue to Step 4.
Open the guided collection prompt
Paste this into the project chat
Use LinkedIn-Analytics-Data-Bank.xlsx already stored in this project's Sources or Files. Confirm that you can access it before we begin. Do not ask me to download, upload, open, or edit the workbook. I will keep LinkedIn open separately and give you the numbers. You will guide the collection and update the workbook after I confirm everything.
First, inspect the workbook's sheet names, column headers, existing reporting periods, formulas, and Field Guide.
Use the most recent fully completed calendar month. State its exact start and end dates and ask me to confirm them.
HOW TO GUIDE ME
- Give me only one short navigation instruction and one screen-sized metric group at a time. Then stop and wait for my reply.
- For each screen, tell me where to go, list only the fields needed from that screen, and give me a compact reply template.
- I may type the values, paste visible text, upload a screenshot, reply “Not shown,” or reply “I can't find it.”
- After each reply, briefly repeat what you captured. Ask only about missing or genuinely ambiguous values, then move to the next screen.
- If LinkedIn's labels or navigation differ, ask what I see and adapt. Do not pretend that you can see or control my LinkedIn tab.
- Never ask for my password, login code, or identifiable profile-viewer names.
COLLECT IN THIS ORDER
1. Combined content analytics
Guide me to Me → View Profile → Analytics → Show all, or the closest equivalent shown in my account. Set and confirm the exact reporting dates. Collect the workbook fields for impressions, members reached, network distribution, total engagement, and the visible engagement breakdown.
2. Audience and follower analytics
Collect total followers, new followers, and any other matching follower fields in the workbook. Use the exact calendar month only if LinkedIn offers that control on the current device. Total followers is a current snapshot, not a monthly total. Record the actual timeframe LinkedIn shows for new followers and demographics.
3. Profile analytics
Collect profile views, profile appearances, and the optional profile impression, click, or viewing-time fields only when LinkedIn shows them. Record the exact timeframe attached to every number; do not label a different timeframe as monthly.
4. Search appearances
Collect search appearances and the visible details about where the profile appeared, searcher companies, searcher job titles, and job titles the profile was found for. Record the exact timeframe LinkedIn shows.
5. Audience demographics
Take one visible demographic category at a time. Collect its audience type, category, ranked values, percentages, displayed source scope, and actual timeframe. A screenshot is acceptable instead of typing every row.
6. Posts published during the reporting period
First guide me to my posts and identify only the posts published within the confirmed dates. Then handle one post at a time.
For each post, first collect its date, time if shown, URL, opening line or identifier, and content format. Next, collect the matching reach, network, engagement, profile-action, and follower fields. Treat individual-post analytics as a snapshot taken on the collection date, not as activity limited to the calendar month. Ask for video metrics only for a video post, link metrics only when applicable, and article or newsletter metrics only for that content type. If there were no posts during the month, record that fact in the profile-level notes and do not create a placeholder post row.
DATA RULES
- Use only values I provide.
- Never guess, estimate, infer, or back-calculate a missing raw value.
- Record zero only when LinkedIn explicitly displays zero.
- “Not shown” is a chat response, not a numeric spreadsheet value. Leave the applicable numeric cell blank and record the omitted field in Data Quality Notes, Evidence / Missing Data Notes, or Availability Note.
- Do not silently replace an existing value with a blank or “Not shown.”
- The calendar month controls the combined-content period and which posts are included. It does not automatically apply to every other LinkedIn screen.
- Confirm the exact date range or timeframe on every range-sensitive screen. If a screen cannot use the confirmed calendar-month range, do not substitute another range or describe it as monthly. Record the actual timeframe in the appropriate notes field.
- Preserve all prior periods, sheet names, tables, formulas, formatting, and the Field Guide.
- Do not type into calculated rate columns. Preserve or extend their formulas.
FINAL CONFIRMATION
When collection is complete, show me one consolidated review containing:
- The reporting start and end dates
- All profile-level values
- All demographic rows
- Each post and its captured values
- Every field marked Not shown
- The actual timeframe for every value that does not match the calendar month
- Any date-range limitation or conflict with an existing value
Do not change the workbook yet. Ask me to reply “Confirm” or send corrections.
AFTER I CONFIRM
- Use Collection Method: Manual.
- Add the month if it is new.
- If the exact reporting period already exists, update the matching row or rows instead of creating duplicates. Preserve existing values unless I confirmed their replacement.
- Recalculate and verify the workbook.
- Return exactly one updated file named LinkedIn-Analytics-Data-Bank.xlsx.
- Do not create the Action Report yet. Wait for my report prompt in this same conversation.
- Do not return intermediate files, CSV files, ZIP files, or duplicate workbook versions.
- Tell me: “The Data Bank is updated. Keep this chat open and continue to the report step.”
Start with only the target date range and the first navigation instruction.
You never touch the spreadsheet. You read the numbers from LinkedIn; the AI asks the questions, checks your answers, and updates the Data Bank.
Open the LinkedIn Analytics project and start a new Work task inside it.
Confirm that LinkedIn-Analytics-Data-Bank.xlsx appears in the project’s Sources. You do not need to attach it again.
Paste the collection prompt below. ChatGPT will open LinkedIn in its browser. Approve access when ChatGPT asks.
If LinkedIn asks you to sign in, sign in inside the browser and then let ChatGPT continue.
When ChatGPT returns the updated XLSX, check at least three values against LinkedIn.
Keep this Work task open and continue to Step 4. Do not replace the project files yet.
Open the ChatGPT Data Bank prompt
Paste this into the new Work task
Use the browser available in this ChatGPT Work task. Open LinkedIn and update LinkedIn-Analytics-Data-Bank.xlsx from this project's Sources for the most recent complete calendar month.
YOUR TASK
1. Open LinkedIn and confirm that this is the correct personal LinkedIn profile.
2. Open Me → View Profile → Analytics → Show all.
3. In combined post analytics, use a custom date range for the most recent complete calendar month. Record the exact date range shown.
4. Collect the available audience, profile, and profile-search numbers. For each screen, record the exact timeframe LinkedIn actually shows; never describe a different timeframe as monthly.
5. Collect the analytics for each post published during that same period.
6. Add the exact values to the matching Data Bank tabs.
RULES
- Keep all prior rows, formulas, and the Field Guide.
- Do not type in formula cells.
- Do not combine different date ranges.
- Treat total followers and individual-post analytics as snapshots taken at collection time.
- Record any non-monthly audience, profile, search, or demographic timeframe in the appropriate notes field.
- If you check a field and LinkedIn does not show it, leave the numeric cell blank and record “Not shown” in the appropriate notes column.
RETURN TO ME
- Return the completed XLSX file as LinkedIn-Analytics-Data-Bank.xlsx so I can use it in the report step and replace the project source when the workflow is complete.
- A short note that lists the LinkedIn areas you checked and anything that was not shown.
Do not make the Action Report yet.
What if I opened a regular ChatGPT chat?
Stop before you paste the prompt. Open the LinkedIn Analytics project and start a new Work task there. The Work task provides the browser and can use the project sources.
Do I need a folder or Chrome extension?
No. ChatGPT uses the browser available in the Work task and reads the Data Bank from Project Sources. You do not connect a computer folder or install a Chrome extension for this route.
Open Chrome. Sign in to the correct LinkedIn profile.
Select the Claude icon in the Chrome toolbar. The Claude side panel starts a Cowork chat from the LinkedIn tab.
Open your regular Claude workspace. Find the new chat, select its three-dot menu, and add it to the LinkedIn Analytics project.
Return to the Cowork chat in the Chrome side panel.
Confirm that the chat can use LinkedIn-Analytics-Data-Bank.xlsx from the project’s Files. You do not need to attach it again.
Paste the Cowork collection prompt below. Let Claude use the open LinkedIn tab.
When Claude returns the updated XLSX, check at least three values against LinkedIn.
Keep this Cowork chat open and continue to Step 4. Do not replace the project files yet.
Select the Claude icon in the Chrome toolbar and paste the LinkedIn collection prompt below.
Let Claude collect the numbers. It does not need the spreadsheet or access to your Claude project.
When Claude returns the LINKEDIN ANALYTICS HANDOFF, review it and copy the entire block.
Open your LinkedIn Analytics project in Claude and confirm that LinkedIn-Analytics-Data-Bank.xlsx appears under Files.
If the extension conversation is available inside the project, you may continue there. If it is not, start a new project chat. Either way, use a chat inside the project for the remaining steps.
Paste the Data Bank update prompt below, followed by the complete handoff from Chrome.
When Claude returns the updated XLSX, check at least three values against the handoff. Keep this project chat open for Step 4.
Open the Claude Cowork Data Bank prompt
Paste this into the new Cowork chat
Use this open LinkedIn tab and LinkedIn-Analytics-Data-Bank.xlsx from this project's Files. Update the Data Bank for the most recent complete calendar month.
1. Open the LinkedIn collection prompt
Paste this into Claude in Chrome
Help me collect my personal LinkedIn analytics.
YOUR TASK
1. Confirm that Chrome is using the correct personal LinkedIn profile.
2. Open Me → View Profile → Analytics → Show all.
3. In combined post analytics, use a custom date range for the most recent complete calendar month. Record the exact date range shown.
4. Collect the available audience, profile, and profile-search numbers. For each screen, record the exact timeframe LinkedIn actually shows; never describe a different timeframe as monthly.
5. Collect the analytics for each post published during that same period.
6. Add the exact values to the matching Data Bank tabs.
IMPORTANT SCOPE
- Use only LinkedIn in this Chrome browser.
- You do not need access to any spreadsheet, Claude Project, project file, or computer folder.
- Do not create or update a spreadsheet.
- Do not create a report.
- Do not publish, edit, like, comment, message, follow, connect, or change anything on LinkedIn.
- Your only final output will be one structured, copyable data handoff.
RULES
- Keep all prior rows, formulas, and the Field Guide.
- Do not type in formula cells.
- Do not combine different date ranges.
- Treat total followers and individual-post analytics as snapshots taken at collection time.
- Record any non-monthly audience, profile, search, or demographic timeframe in the appropriate notes field.
- If you check a field and LinkedIn does not show it, leave the numeric cell blank and record “Not shown” in the appropriate notes column.
WORK ONE SCREEN AT A TIME
Guide me through one LinkedIn analytics screen at a time. Navigate the site when you have permission. If I must do something myself, give me one short instruction and wait.
RETURN TO ME
- Return the completed XLSX file as LinkedIn-Analytics-Data-Bank.xlsx so I can use it in the report step and replace the project file when the workflow is complete.
- A short note that lists the LinkedIn areas you checked and anything that was not shown.
Use the most recent fully completed calendar month as the target reporting period. Confirm its exact start and end dates before collecting anything.
Do not make the Action Report yet.
What if the chat is not in the project?
Start the Cowork chat in the Chrome side panel. Then open Claude in your usual workspace. Find the new chat, select its three-dot menu, and add it to the LinkedIn Analytics project.
COLLECT IN THIS ORDER
1. Combined content analytics
Open Me → View Profile → Analytics → Show all, or the closest equivalent available in this account. Open combined post or content analytics and set a custom range matching the exact target month. Collect every visible summary and engagement metric, preserving LinkedIn's labels exactly, including impressions, members reached, network distribution, total engagement, and the visible engagement breakdown.
2. Audience and follower analytics
Collect total followers, new followers, and every other relevant follower metric shown. Use the target calendar month only if LinkedIn offers that exact date control on this screen. Total followers is a current snapshot, not a monthly total. Record the actual timeframe LinkedIn displays for every other value.
3. Profile analytics
Collect profile views, profile appearances, and any profile impressions, clicks, or viewing-time metrics LinkedIn displays. Preserve the exact label, value, and timeframe shown.
4. Search appearances
Collect the search-appearance total and the visible breakdowns, including where the profile appeared, searcher companies, searcher job titles, and job titles the profile was found for. Record the exact timeframe shown. Do not collect or list individual viewer names.
5. Audience demographics
Collect each visible demographic category one at a time. Preserve the audience type, category name, ranked values, percentages, displayed scope, and actual timeframe.
6. Posts published during the target month
Find only posts whose publication dates fall within the confirmed target dates. Handle one post at a time.
For every included post, collect:
- Publication date and time, when shown
- Post URL
- Opening line or another clear identifier
- Content format
- Every analytics metric visible for that post, preserving LinkedIn's exact labels
- The date on which these post analytics were collected
Ask for video-only, link-only, article, or newsletter metrics only when they apply to that post. Treat individual-post analytics as snapshots taken on the collection date, not as activity limited to the calendar month.
If there were no posts during the target month, record that fact. Do not invent a placeholder post.
ACCURACY RULES
- Copy only values LinkedIn actually displays.
- Never guess, estimate, infer, calculate, or back-calculate a missing raw value.
- Record zero only when LinkedIn explicitly displays zero.
- If a requested field is unavailable, record “Not shown.”
- Do not silently convert percentages, abbreviations, or units.
- The target month controls the combined-content range and which posts are included.
- It does not automatically apply to audience, follower, profile, search, demographic, or individual-post analytics.
- For every range-sensitive screen, record the exact timeframe LinkedIn displays.
- Never describe a different timeframe as monthly.
- Keep differently timed values separate.
FINAL REVIEW
Before creating the handoff, show me a concise review of the reporting dates, profile-level values, demographic rows, posts, fields marked Not shown, and every non-monthly timeframe. Ask me to reply “Confirm” or send corrections.
FINAL OUTPUT
After I confirm, return one fenced plain-text block using this structure, with no commentary before or after it:
LINKEDIN ANALYTICS HANDOFF
Collection method: Claude in Chrome
Collection date:
LinkedIn profile:
Target reporting period:
Target start date:
Target end date:
COMBINED CONTENT
Displayed date range:
- [LinkedIn label]: [value]
AUDIENCE AND FOLLOWERS
Displayed timeframe:
Snapshot date for total followers:
- [LinkedIn label]: [value]
PROFILE ANALYTICS
- [LinkedIn label] | [value] | [displayed timeframe]
SEARCH APPEARANCES
- [LinkedIn label] | [value] | [displayed timeframe]
AUDIENCE DEMOGRAPHICS
- [audience type] | [category] | [ranked value] | [percentage] | [displayed scope] | [displayed timeframe]
POSTS PUBLISHED DURING THE TARGET PERIOD
POST 1
Publication date:
Publication time:
URL:
Identifier or opening line:
Content format:
Analytics snapshot date:
- [LinkedIn label]: [value]
[Repeat for every included post, or state: No posts found during the target reporting period.]
NOT SHOWN
- [screen]: [field]
TIMEFRAME NOTES
- [metric or screen]: [actual timeframe and any limitation]
OTHER NOTES OR CONFLICTS
- [note]
END HANDOFF
Include no assumptions and no spreadsheet instructions inside the handoff.
2. Open the Data Bank update prompt
Paste this into a chat inside the Claude project, then add the handoff
Update the existing LinkedIn Analytics Data Bank using the structured handoff pasted below.
The current project Files contain LinkedIn-Analytics-Data-Bank.xlsx. Use that exact workbook as the starting file. Do not ask me to download, open, edit, or re-upload it.
FIRST
1. Open and inspect LinkedIn-Analytics-Data-Bank.xlsx from this Project's Files.
2. Identify its existing sheets, tables, headers, formulas, formatting, Field Guide, notes fields, and prior reporting periods.
3. Map the supplied handoff values to the matching workbook fields.
4. If you cannot access the existing project workbook, stop and tell me. Do not construct a replacement workbook from memory.
UPDATE RULES
- Preserve every prior reporting period.
- Preserve all sheet names, tables, formulas, formatting, instructions, and the Field Guide.
- Set Collection Method to Claude in Chrome.
- If the target reporting period already exists, update its existing records instead of creating a duplicate.
- Match posts by URL when available, then by publication date and identifier.
- Do not type into calculated-rate columns. Preserve or extend their formulas.
- Use only values explicitly supplied in the handoff.
- Never guess, estimate, infer, calculate, or back-calculate a missing raw value.
- Record zero only when the handoff explicitly reports zero.
- For “Not shown” fields, leave the applicable numeric cell blank and record the missing field in the appropriate Data Quality, Evidence, Missing Data, or Availability notes field.
- Do not replace a nonblank existing value with a blank or “Not shown.”
- If a supplied value conflicts with a nonblank value already stored for the same metric and timeframe, show me the conflict and wait before replacing it.
- The target calendar month applies to combined-content analytics and determines which posts are included.
- Do not force that monthly range onto other LinkedIn screens.
- Preserve the actual displayed timeframe for audience, follower, profile, search, and demographic values.
- Store total followers as a snapshot taken on the stated collection date.
- Store individual-post analytics as snapshots taken on the stated collection date.
- Keep values from different timeframes clearly separated.
- If a supplied LinkedIn label has no matching workbook field, preserve it in the appropriate evidence or notes field rather than discarding it.
WHEN FINISHED
- Return exactly one updated workbook named LinkedIn-Analytics-Data-Bank.xlsx.
- Keep the updated workbook available in this conversation for the report step.
- Briefly list the reporting period updated, the number of posts added or updated, and any values stored in notes because no matching field existed.
- Do not create the Action Report yet.
- Do not attempt to replace the file stored in Project Files. I will replace all completed project files at the end.
The handoff begins below:
[PASTE THE COMPLETE LINKEDIN ANALYTICS HANDOFF HERE]
What if the extension conversation is not in my project?
That is fine. The extension conversation only collects the LinkedIn data. Copy its complete LINKEDIN ANALYTICS HANDOFF, open a new chat inside the LinkedIn Analytics project, and paste the Data Bank update prompt followed by the handoff. You do not need to move the extension conversation.
What if the Claude side panel does not open?
Confirm that Claude in Chrome is installed, turned on, and signed in. Refresh the LinkedIn tab and select the Claude icon again.
Your current task or chat contains the updated LinkedIn-Analytics-Data-Bank.xlsx for the most recent complete calendar month. Keep that conversation open.
Before you move on
The AI project conversation you will use for reporting contains the updated LinkedIn-Analytics-Data-Bank.xlsx for the most recent complete calendar month. Keep that project conversation open.
Step 4 of 4
Update the two Action Report files.
The AI will read the checked Data Bank, explain what the numbers mean, and return the updated HTML report plus a lightweight Markdown backup.
What you are doingTurning the checked numbers into one short analysis and three practical next moves.
What you will haveA polished HTML report and a matching lightweight Markdown backup.
Stay in the same project chat that guided you through the LinkedIn numbers. Do not start a new chat.
Stay in the same ChatGPT Work task that collected the LinkedIn numbers. Do not start a new task.
Stay in the same Claude Cowork chat that collected the LinkedIn numbers. Do not start a new chat.
Stay in the Claude project chat that received the Chrome handoff and updated the Data Bank. You do not need the Chrome side-panel conversation anymore.
Confirm that the updated Data Bank is available in this conversation and the current Markdown and HTML report templates are available in the project’s Sources or Files.
Paste the reporting prompt below.
When the AI finishes, download the updated Data Bank, Markdown report, and HTML report from this conversation.
In the project’s Sources or Files, replace the old Data Bank and both old reports with the three updated files.
Confirm that the project contains exactly one Data Bank, one Markdown backup, and one HTML report. Open the reports and check that the reporting period, main conclusion, and three actions match.
Open the Action Report prompt
Paste this in the same task or chat
Use the updated LinkedIn Analytics Data Bank created earlier in this conversation and the current Action Report files from this project's Sources or Files. Update both reports for the most recent complete calendar month. Do not ask me to upload the Data Bank again.
Open the Action Report prompt
Paste this in the conversation that contains the updated Data Bank
Use the updated LinkedIn Analytics Data Bank created earlier in this conversation and the current Action Report files from this project's Sources or Files. Update both reports for the most recent complete calendar month. Do not ask me to upload the Data Bank again.
The reporting cadence is monthly. Use the exact calendar-month date range consistently in both reports.
YOUR GOAL
Explain what happened during this reporting period and give me exactly three useful actions for the next period.
ANALYSIS RULES
- Use only evidence in the Data Bank.
- Compare only matching date ranges.
- Clearly separate facts from interpretations.
- Do not call one post or one reporting period a trend.
- If no prior matching period exists, label this “Baseline Period.” Do not claim an increase or decrease.
KEEP THESE REPORT SECTIONS
Period in one sentence; What changed; What created movement; Audience signals; What the data does not prove; Repeat, Refine, Retire; Three moves for the next period; Evidence drawer; Rolling history.
RETURN THESE TWO FILES
- An updated HTML report named linkedin-living-action-report.html. Treat this as the primary, designed report.
- An updated Markdown backup named linkedin-living-action-report.md. Keep it clean and readable with clear headings, short paragraphs, and simple bullets.
- Use the existing HTML report design when the current file is available in the project. Keep its working “Print or save as PDF” button and print styling.
- If the HTML file is not available in the project, create a clean replacement that matches the Markdown report. Use a white and ivory background, deep red accents, black text, generous white space, and text no smaller than 14px. Include a working “Print or save as PDF” button.
- Give me a short confirmation in the chat with the three actions for the next reporting period.
Preserve the prior Rolling History. Before you finish, confirm that the Markdown and HTML reports have the same reporting period, evidence, conclusions, and three actions.
What if the HTML file is missing from the project?
Continue with the Data Bank and Markdown backup. The reporting prompt tells the AI to create a replacement HTML report that matches the Markdown file. Add the new HTML file to the project when the task is complete.
What if the two reports do not match?
Reply: “Regenerate both files from the same final analysis. The reporting period, headings, evidence, conclusions, and three actions must match.”
What if this is my first reporting period?
The report must say Baseline Period. It can describe the current evidence, but it cannot claim that a result increased, decreased, or became a trend.
You are finished when
Your LinkedIn Analytics project has one current Data Bank and two matching current Action Report files. The reports give exactly three actions for the next reporting period.
Quick help
File questions
Open only the answer you need.
Where did my downloads go?
They will usually appear in your device’s Downloads area. Use the project’s Sources or Files section to add the returned file, then remove the older project version.
Does the AI project connect to a folder on my computer?
No. A normal ChatGPT or Claude project does not need a folder connection. You explicitly add the starter files to the project once, then replace a project source only when the AI returns an updated version.
Why use Project Sources or Files?
Files stored with the project are available to new tasks and chats in that project. That means you do not need to locate and attach all three files for every reporting period.
How does Claude in Chrome use the Data Bank?
It does not use the Data Bank while it is browsing LinkedIn. Claude in Chrome collects the numbers and gives you one copyable handoff. Paste that handoff into a chat inside your LinkedIn Analytics project; that project chat can read the Data Bank from Files, update it, and create the reports.
How do I install Claude in Chrome?
Open Google Chrome. The extension is not supported in other browsers.
Open the official Claude in Chrome setup guide below and follow its link to the Chrome Web Store.
Select Add to Chrome and approve the installation.
Sign in with your Claude account when prompted.
Select Chrome’s puzzle-piece icon and pin Claude so its icon stays in the toolbar.
Open LinkedIn, select the Claude icon, and approve the permissions needed for the current task.
Double-click the file that ends in .html. It opens as a designed page in Chrome, Edge, or Safari. Use its Print or save as PDF button when you want a familiar file to share.
How do I open the Markdown backup?
The file ending in .md is a simple text backup. Open it in any notes or text app, or add it back to your AI project as a source. The HTML file is the polished version.
How do I share the report?
Open the HTML report and choose Print or save as PDF for the easiest file to send. You can also send the HTML file itself; the recipient downloads it and double-clicks it to open it in a browser.
What if the download has a number in its filename?
Rename the downloaded file so it matches the original filename. Then remove the old project source and add the renamed replacement.
Manage Consent
To provide the best experiences, we use technologies like cookies to store and/or access device information. Consenting to these technologies will allow us to process data such as browsing behavior or unique IDs on this site. Not consenting or withdrawing consent, may adversely affect certain features and functions.
Functional
Always active
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
Preferences
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
Statistics
The technical storage or access that is used exclusively for statistical purposes.The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
Marketing
The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.