LinkedIn Content Audit: Plan 30 Days of Posts With Claude

There are two things that make LinkedIn hard. First, you try to post consistently but have no idea what to post. Then, when you do post, it doesn't get the results you expected.

The answer to both is already sitting in your account. Every post you've published is data: which hooks stopped the scroll, which topics your buyers cared about, which CTAs actually got clicks. This guide shows you how to pull that data out, hand it to Claude, and turn it into an exact plan for the next 30 days.

Watch the walkthroughPlan 30 Days of LinkedIn Content With Claude11 min · YouTube

What the audit gives you

• What actually performed: your posts ranked by the metric closest to your goal, against your own median, not someone else's benchmarks.
• Your strongest hooks, word for word: with a reusable formula for each and a new hook written in your voice.
• Your best topics, angles and formats: with how many posts sit behind each finding, so you don't chase one-off flukes.
• CTAs to reuse, rewrite, test or drop: with replacement copy for the ones that need fixing.
• A repost and repurpose shortlist: up to 10 posts worth bringing back, what to change, and the earliest sensible date to repost each.
• A dated 30-day plan: one row per post with the exact hook, CTA and success metric, delivered as a Claude Artifact with checkboxes.
• A 30-day scorecard: the numbers that tell you whether to repeat, revise or drop each idea.

What you need

• At least 20 to 30 posts. More is better, but see the note on time ranges below.
• Your post copy, publish dates and post URLs.
• Your post analytics: impressions at minimum, plus reactions, comments, reposts, saves and clicks if you have them.
• A Claude account (the free plan works for smaller exports, a paid plan handles bigger files and longer analysis).
• Your primary goal, who you want to reach and how often you want to post.

Step 1: Get your LinkedIn post data

You need two things: the copy of each post and its numbers. There are two ways to get them.

Option A: Download your data from LinkedIn (free)

Part 1: your post copy (data archive)

1. On LinkedIn (desktop), click your profile photo in the top right, then Settings & Privacy.
2. Go to Data privacy, then Get a copy of your data (sometimes labelled "Download your data").
3. Choose Download larger data archive, then Request archive.
4. LinkedIn says it can take up to 24 hours. It often arrives within a couple of hours. You'll get an email with a download link.
5. Unzip the archive. Your posts are in the file called Shares (Shares.csv), with the post text, date and link for each post.

Part 2: your post analytics (content export)

1. Go to your profile and open your Analytics section (click "post impressions").
2. Set the date range to match your audit, for example the past 90 or 365 days.
3. Click Export. You'll get a spreadsheet with your top posts (with links and dates), plus impressions and engagement over time.
4. For any important post that isn't in the export, open it on LinkedIn, click View analytics under the post, and note its numbers.

LinkedIn moves these menus around occasionally, so if a label looks slightly different, look for the closest match. Claude matches the copy file and the analytics file using each post's URL, so you don't need to merge them yourself.

Option B: Export everything with MagicPost (fastest)

1. Sign up for MagicPost and connect your LinkedIn account.
2. Go to Metrics and select your posts.
3. Choose your time range, for example the last 6 months or 90 days.
4. Click Export to CSV. Done.

The CSV includes each post's URL, full copy, type (image, video, article, carousel), impressions, likes, comments, the day and hour you posted, and more. That's everything Claude needs in one file.

Disclosure: the MagicPost link above is a referral link. Other LinkedIn tools and Chrome extensions can export similar data, and Option A is completely free.

Pick the right time range

More data helps Claude find patterns, but LinkedIn changes often. What worked a year ago can look very different from what works today.

• 90 days: best for spotting what's working right now. Use this if you post 3 or more times a week.
• 6 months: a good balance for most people.
• 12 months: use if you post less often and need enough posts to see patterns. Tell Claude to weight recent posts more heavily.

Step 2 (optional): Tidy the file in Google Sheets

You can drop your files straight into Claude. If you want to check or combine them first:

1. Open a new Google Sheet and go to File → Import → Upload.
2. Upload your CSV or spreadsheet and choose Append to current sheet (or insert new sheets for each file), then Import data.
3. Check every post has its URL, copy, date and numbers. Delete reposts of other people's content and anything private.
4. Download it again with File → Download → CSV.

Step 3: Copy the content audit prompt

Click Copy prompt, then fill in the [brackets] under "My inputs" with your real details. Leave the rest exactly as it is.

Act as my LinkedIn content strategist. I’m giving you my LinkedIn post copy and analytics. Find the posts worth bringing back, identify the patterns I should repeat, and turn your findings into an exact 30-day publishing plan.

## My inputs

- **Analysis period:** [Last X days or specific date range]
- **Primary goal:** [Qualified leads, booked calls, subscribers, follower growth, reach, or another goal]
- **Audience and offer:** [Who I want to reach and what I sell or promote]
- **Publishing cadence:** [Posts per week and preferred days, if any]
- **Time zone:** [Your time zone, optional]
- **Post copy and analytics:** [Paste below or attach files. Include the full copy, publish date, post URL or ID, format, and any available metrics such as impressions, reactions, comments, reposts, saves, clicks, leads, or booked calls.]

## 1. Establish what performed

Confirm the date range, number of posts, available metrics, and any gaps that affect the analysis. Match each post to its analytics using the URL or ID when available. Flag uncertain matches.

Rank posts by the metric closest to my primary goal. If that metric is missing, explain which available metric you’re using as a proxy. For example, clicks can show interest in an offer, while impressions show reach. If I leave the goal blank, use reach and meaningful engagement as separate views.

Where the data supports it, compare raw totals and rates. Show the denominator for every rate. Use my own median as the baseline for each metric you rely on, and account for differences in how long posts have been live. If I provide only lifetime totals, flag recent posts that haven’t had a comparable window to accumulate results. Treat blank metrics as unknown. Cite the post date or URL whenever you make a recommendation based on a specific post.

## 2. Find my strongest hooks, topics, and CTAs

**Hooks.** Pull the strongest hooks word for word, preserving their line breaks. For each, give me:

- The exact hook and source post
- The hook style and the result it earned
- A reusable formula with placeholders
- One new hook in my voice using that formula

Group similar hooks and show which styles performed well across multiple posts. Call out one-off winners as ideas to test. Flag recurring weak hooks and explain how you’d rewrite them.

**Topics and angles.** Tag each post by its subject, angle, and format. For example, “LinkedIn outreach” is a subject, “a behind-the-scenes walkthrough” is an angle, and “carousel” is a format. Show which subjects, angles, formats, and combinations performed best. Include the number of posts behind each finding, a typical result such as the median, and representative examples. Flag combinations that performed well once and deserve another test.

**CTAs.** Quote the exact CTAs from the strongest examples and group them by the action they ask people to take. Include posts with no CTA as a group. Judge each CTA against its intended action when that outcome is available. A keyword CTA can be assessed with keyword comments, a newsletter CTA with signups, and a booking CTA with booked calls. When you only have indirect evidence, say what it shows. Tell me which CTAs to reuse, rewrite, test, or drop, and give me replacement copy for the ones you’d fix.

## 3. Build my repost and repurposing shortlist

Rank up to 10 posts worth bringing back. For each, give me:

| Rank | Original post and date | Result vs. my baseline | Recommended action | What to keep or change | Earliest sensible repost date |
| ---- | ---------------------- | ---------------------- | ------------------ | ---------------------- | ----------------------------- |

Choose the action for each post: repost as-is, refresh the hook or CTA, update the proof, change the format, or turn the idea into a follow-up. Explain why each post still has value for my audience. Check for dated claims, expired offers, and recent posts that could feel repetitive. If a winner is too recent to repost during the next 30 days, schedule a follow-up angle or give it a later date.

Then give me four short lists: **Do more, fix, test, and eliminate.** Tie every recommendation to a metric or example. Reserve “eliminate” for patterns with enough comparable posts to justify dropping them. Put promising or weak patterns with limited evidence under “test.”

## 4. Give me my exact 30-day LinkedIn plan

Start with today’s calendar date in my time zone if I supplied one. If today’s date is unavailable, ask me for it before assigning dates. Use today through 29 days from today as the planning window. State the start and end dates. Follow my cadence and preferred days. If I leave them blank, use my recent posting frequency and state the assumption.

Give me one row per planned post:

| Publish date and weekday | Repost, refresh, repurpose, or new test | Source post | Topic, angle, and format | Exact opening hook | Exact CTA or no CTA | Success metric |
| ------------------------ | --------------------------------------- | ----------- | ------------------------ | ------------------ | ------------------- | -------------- |

Schedule the strongest eligible ideas first and space out repeats of the same topic, story, or offer. For each repurposed post, specify what to carry over and what to change. Use the remaining slots to test the most promising unproven patterns. When possible, change one main variable per test so I can learn whether the hook, angle, format, or CTA made a difference.

Finish with my **first three actions this week** and a **30-day scorecard** with the metrics to track, when to capture them so posts are compared at the same age, and what result would justify repeating, revising, or dropping an idea. Use my historical baseline for benchmarks where the data supports one.

Make every recommendation specific enough that I can open the original post and start writing. Please provide me with a Claude Artifact that I can easily utilize with checkboxes for timeline, as well as the timeline of to-do items.

## Paste my post copy and analytics here

[PASTE DATA HERE OR ATTACH FILES]

Step 4: Run it in Claude

1. Open claude.ai and start a new chat. Use the most capable model available on your plan for the best analysis.
2. Paste the prompt and fill in your inputs: analysis period, primary goal, audience and offer, posting cadence and time zone.
3. Drag your export files into the chat (your CSV, spreadsheet or both).
4. Make sure today's date is clear, either in your time zone line or by adding "Today is [date]", so the 30-day plan starts on the right day.
5. Send it and let Claude work. A big export can take a few minutes.
6. Open the Artifact Claude creates. It has your plan with checkboxes, findings, hooks, CTAs and the repost shortlist in one place.

Want to reuse it every month? Create a Claude Project called "LinkedIn Content Audit", save the prompt in the Project instructions, and drop in a fresh export each time.

How to read your results

Findings

Start with your medians: impressions, comments, saves, profile views and link clicks. These are your personal baseline. A post "performed well" if it beat your median, not someone else's.

Hooks

Look for hook styles that won across several posts, not just once. Use the formulas Claude gives you, but fill the brackets with new material instead of copying an old hook word for word. For more hook ideas, see LinkedIn hooks.

CTAs

Claude sorts your calls to action into reuse, rewrite, test and drop. Take the "drop" list seriously: if a CTA consistently underperforms, scrap it and use the ones that work.

Repost shortlist

Respect the "earliest sensible repost date". Repost too soon and your audience notices you're recycling. Use refreshes and follow-up angles to keep winners working in the meantime.

Do more, fix, test, eliminate

Your strategy in four lists. "Eliminate" only includes patterns with enough posts behind them, so promising but unproven ideas go to "test" instead.

One post, one purpose

As you work through the plan, remember that every post needs one specific purpose: one topic, one theme and one call to action. Send people to your profile, a landing page, a booked call or nowhere at all, but never to two or three places at once. Too many options confuse your audience, and they end up choosing none of them.

Using the 30-day plan and scorecard

1. Post on the dates in your plan and tick each one off in the Artifact.
2. Measure every post at the same age, for example 7 days after publishing, so you're comparing like with like.
3. Compare each result to the scorecard: above the "repeat" line, do more of it; in the middle, revise the copy; below the "drop" line twice, drop it.
4. Change one main thing per test (the hook, the format or the CTA) so you know what made the difference.
5. At the end of the 30 days, export again and rerun the audit. Each month your baseline gets sharper.

For a quick weekly version of this review, see LinkedIn analytics explained.

Content audit mistakes

• Auditing too few posts. Under 20 posts, most patterns are noise.
• Comparing a post from yesterday to a post from six months ago. Lifetime totals favour older posts.
• Copying old hooks word for word instead of reusing the formula.
• Reposting winners too soon.
• Dropping a pattern after one bad post.
• Forgetting the goal. A post with huge reach and zero leads isn't a winner if leads are what you want.

Want every prompt and blueprint from this system in one place, plus feedback on your plan? Join The Growth Innovator community on Skool and start with the Start Here module. It's free.