Visualize Your LinkedIn Data Export
An experiment by Logan Currie using Claude Code
January 20, 2026
LinkedIn knows everything about your professional life, but their interface only shows you a fraction of it. The data was always there. I just couldn't see it. So I downloaded all 50 files and made AI read them.
Note: This uses my real LinkedIn data. I've removed identifying info about other people and simplified some details so you can see what these visualizations could look like with your own export. I found this exercise fascinating. Read my full thoughts on Substack →
I'm sure I'm just scratching the surface here.
If you try this and get interesting results, DM me. Send this page to a friend. It's all free. Let's see how far this travels.
How to Get Started
- Request your LinkedIn data export. Go to Settings → Data Privacy → Get a copy of your data. Select "Download larger data archive" for the complete export. Takes about 24 hours.
- Unzip the file. You'll get ~50 CSV files with everything LinkedIn knows about you.
- Open Claude Code and point it at your export folder. I really encourage you to try Claude Code if you haven't. This is a great way to see what all the hype is about.
- Ask questions. Use the prompts below to generate each visualization. Copy/paste them directly.
- Get creative. There's so much more you could do with this data. These are just some examples.
💡 On AI tools: I recommend Claude Code because it handles large context windows well. I use Claude Code Max ($100/mo) which let me analyze all 50 CSVs in extended sessions. On the Pro plan ($20/mo), you can absolutely do this. Just focus on 2-3 CSV files per session and spread it across a few days. The prompts below work great either way.
1. Your Network as a Universe
Every node = a cluster based on current employer or industry. Size = number of connections. Drag to explore.
Note: LinkedIn only exports current employer — not full work history or education. So this is a snapshot, not the complete picture.
Founders
Academia
AI/ML
EdTech
VC/Investors
HR/Talent
Big Tech
Prompt to try
"Analyze my Connections.csv file. Group my connections by their current company or industry. Show me the top 10 clusters by size and create a force-directed network visualization."
Uses: Connections.csv
2. LinkedIn's Inferences vs Reality
What LinkedIn thinks they know about you for ad targeting vs. what your data actually shows.
What LinkedIn Thinks
Hostage Negotiation
Bathing
Fisheries Management
Hunting Land
Archery
Garden Design
Toy Sales
Harley Davidson
What Your Data Shows
Future of Work
AI & Hiring
Career Development
Educational Technology
Content Strategy
Startup Building
Prompt to try
"Open my Inferences_about_you.csv and Ad_Targeting.csv. What does LinkedIn think they know about me? Compare this to my actual interests based on Company_Follows.csv and my content."
Uses: Inferences_about_you.csv, Ad_Targeting.csv, Company_Follows.csv
3. Lost Opportunity DM Search
High-potential messages buried in your inbox that deserve a reply. These are real opportunities you might be missing.
23
unanswered high-value DMs
from the last 6 months
7
someone followed up
they messaged you twice — still no reply
4
still salvageable
sent within the last 2 weeks
!!
Collaboration offer from a founder in your space
They followed up once already. Mutual connections: 12
Collaboration
18 days
!!
Podcast interview invitation
Host has 25K followers. Sent a follow-up 5 days later.
Opportunity
12 days
!
Warm intro from a mutual connection
Introduced by someone you message frequently. In your industry.
Warm Intro
9 days
!
Meeting request — specific dates offered
They proposed 3 time slots. VP-level at an edtech company.
Meeting
6 days
·
Someone sharing your post with their take
Reshared your content and DM'd you about it. 8K followers.
Engagement
3 weeks
·
Advisor intro — someone vouched for you
A 2nd-degree connection said "you should talk to [you]." No reply yet.
Warm Intro
4 weeks
They followed up — reply ASAP
High-value — don't let this go cold
Worth a reply when you have time
Prompt to try
"Go through my messages.csv and find unanswered messages that look like real opportunities — collaboration offers, meeting requests, warm intros, podcast invitations. Prioritize ones where someone followed up or where a mutual connection made the intro. Show me who's waiting and how long they've been waiting."
Uses: messages.csv
4. Does Posting Actually Work?
Correlation between posting frequency and new connection requests.
0.71
correlation coefficient
(that's strong)
1833%
more connections/month
after consistent posting
5 → 100
avg connections/month
before vs after
Prompt to try
"Is there a correlation between posting frequency and new connections? Compare my Shares.csv (posts) with Connections.csv (connection dates) by month. Calculate the correlation coefficient and visualize both on a chart."
Uses: Shares.csv, Connections.csv
5. Who You Actually Follow
The companies you follow reveal your real professional interests — often more accurately than LinkedIn's ad targeting.
EdTech / Learning
~80
companies
Job Search / Career
~60
companies
AI Companies
~50
companies
Future of Work / HR
~45
companies
Women's Networks
~35
companies
VC / Startups
~30
companies
Prompt to try
"Cluster the companies I follow by theme. Analyze my Company_Follows.csv and group them into categories like EdTech, AI, VC, etc. Show me how many companies are in each cluster."
Uses: Company_Follows.csv
6. Inbound vs Outbound Connections
Are people finding you, or are you chasing them?
3.69×
more people find you
than you chase
78.6%
acceptance rate
vs 25-35% avg
129
peak month inbound
Sep 2025
Prompt to try
"Compare my inbound vs outbound connection requests over the last 6 months. Using Invitations.csv, show me how many people requested to connect with me vs how many I reached out to. Calculate my acceptance rate and visualize the trend."
Uses: Invitations.csv
7. Connections vs Real Relationships
Of your connections, how many have you actually messaged in the last year?
~100
Active
5+ DMs this year
~400
Some contact
1-4 DMs this year
~800
Going stale
6-12 mo since DM
~500
Dormant
1-2 yrs since DM
~1,580
Never messaged
Just... a name
Prompt to try
"Of my connections, how many are actual relationships vs. names I've collected? Cross-reference Connections.csv with messages.csv to categorize connections by how recently we've messaged (active, some contact, going stale, dormant, never messaged)."
Uses: Connections.csv, messages.csv
8. Connection Timeline
LinkedIn shows "Connected." The data export shows when, who reached out, and whether you ever spoke again.
C
They added me • never spoke
E
They added me • never spoke
Prompt to try
"Show me a timeline of recent connections. For each person, show when we connected, who initiated, and how many messages we've exchanged since. Use Connections.csv, Invitations.csv, and messages.csv."
Uses: Connections.csv, Invitations.csv, messages.csv
9. Inbox Quality Analysis
Does your inbox feel noisier? The data shows what percentage is genuine vs spam.
91%
of inbound messages are genuine
~9%
actual noise (spam, pitches)
Prompt to try
"Analyze my messages.csv. What percentage of inbound messages are genuine professional outreach vs. spam/sales pitches? Show the breakdown by half-year and visualize the trend."
Uses: messages.csv
10. Your Career in Layers
Each layer = a chapter of your career. The width = how many connections you made during that period.
2008-2012: The Early Days
First professional network • 50 connections
50
2013-2016: Building Overseas
International work, expat network • 284 connections
284
2017-2022: Heads Down
Company building, raising family overseas • 53 connections
53
2023: The Wake-Up Call
Rebuilding network after years overseas, launched company • 124 connections
124
2024: Building in Public
Startup launch, content creation begins • 903 connections
903
2025-2026: The Visibility Era
Viral content, public thought leadership • 1,093 connections
1,093
Prompt to try
"Analyze my Connections.csv by connection date. Group connections by year and show me how my network grew during different periods of my career. Visualize this as layers showing network expansion over time."
Uses: Connections.csv
11. What If You Had Better Data?
Everything above came from a LinkedIn data export — the breadth of your professional life. But breadth without depth is just a contact list. Here's what changes when you add real experience data.
What is careerspan?
careerspan is a free tool that helps you capture the depth of your professional experience — the stuff LinkedIn doesn't know about you. Create an account, walk through structured prompts about your actual work, and get data you can combine with your LinkedIn export for a much richer picture.
The problem with LinkedIn data
What LinkedIn knows about you
- ✓ Who you're connected to
- ✓ When you connected
- ✓ What you post about
- ✓ Who messages you
- ✗ What you've actually done
- ✗ How you think about problems
- ✗ What you're capable of next
What a behavioral interview captures
- ✗ Your network graph
- ✗ Your engagement patterns
- ✗ Your industry connections
- ✓ Specific accomplishments with metrics
- ✓ How you approach ambiguity
- ✓ Skills proven in context (not endorsed)
- ✓ Gaps relative to target roles
Neither data source alone tells you what to do next. Together, they do.
Same experience. Two very different levels of data.
What a resume bullet captures
"Spearheaded a user program reaching 1,200 community members, onboarding 35 participants to validate product-market fit and enhance product alignment with career counselor workflows"
What a structured behavioral interview captures
SITUATION
Transitioning to B2B2C model. Career counselors needed to integrate external product into existing workflows — historically a hard sell. Three pilots launching within weeks, no playbook existed.
TASK
Field-test the product in counselor capacity before pilots launched. Prove functionality, effectiveness, and ease of use. Design the customer success process from scratch.
ACTION
Personally messaged 1,200 people across Slack/WhatsApp. Fought co-founder to expand from 10 to 35 participants. Ran 30-min onboarding calls with each. Weekly check-ins for 2 months. Built user personas from patterns across passive seekers, active searchers, and graduating students.
RESULT
Referenced in 20+ sales meetings. Directly influenced 3 signed pilots (200-student nonprofit, 300-student university, 1,000-student community org). Co-founder: "You were totally right." Completed Dec–Feb, outsized impact on business trajectory.
Skills proven in context (not just endorsed)
User Research
Product Strategy
Community Engagement
Zero-Budget Execution
Stakeholder Persuasion
Customer Success Design
Sales Enablement
Now combine both: Gap Analysis × Your Network
Target role: VP of Product, EdTech (Series A-B). careerspan identifies gaps from your experience data. LinkedIn shows who in your network could help close them.
User Research
Required: Advanced
92% — Strong match
35-person pilot, 3 user personas built
12 connections in UX research
Product Strategy
Required: Advanced
85% — Strong match
Drove B2B2C pivot, admin dashboard specs
8 connections in product
Revenue Ownership
Required: P&L experience
55% — Gap: no P&L ownership cited
Positioned as influence, not ownership
3 connections in finance
Engineering Management
Required: Team leadership
30% — Gap: no direct eng reports
Managed product, not engineers
0 connections in eng mgmt
Scale (500K+ users)
Required: Growth stage exp
40% — Gap: early-stage experience only
Pilots of 200-1,000, not 500K
5 connections at scale-ups
What I see when I look at this
My LinkedIn data tells me I have 2,400+ connections and I'm most engaged in EdTech and AI. Cool. But it can't tell me which experiences actually qualify me for what I want to do next — or where the real gaps are. The resume bullet version of my experience flattens everything into a line item. The behavioral interview version captures the actual story: I fought my co-founder on scope, ran 35 onboarding calls with zero budget, and directly influenced $100K+ in pilot contracts. That's the difference between a contact list and a career strategy.
2,400+
LinkedIn connections
7
skills proven with evidence
Try this yourself
"I'm going to paste a job description and my experience from careerspan. Show me where I'm strong, where the gaps are, and which of my LinkedIn connections could help me close those gaps."
Uses: careerspan export + LinkedIn Connections.csv
How This Was Made
A human + AI collaboration, start to finish.
1
Data Export
Downloaded my LinkedIn data (Settings → Data Privacy → Get a copy). Got 50+ CSV files covering connections, messages, posts, invitations, ad targeting, and more.
2
Exploratory Analysis
Pointed Claude Code at the folder and started asking questions: "What's in here?" → "Show me patterns" → "What would be interesting to visualize?" We iterated through dozens of ideas.
3
Visualization Development
Claude Code generated the HTML, CSS, and JavaScript (Chart.js + D3.js). I gave feedback on design, asked for revisions, and we refined together until the visualizations told the story I wanted.
4
Privacy Scrub
Removed names and identifying details of other people. Kept my data real so you can see what's actually possible with your own export.
Tools used: Claude Code (Max plan) for analysis and code generation, Chart.js for bar/line charts, D3.js for the force-directed network graph. Total development time: ~4 hours across multiple sessions.
Style Guide
Want your visualizations to look like these? Here's what we used.
Color Palette
Muted, sophisticated tones that work on light backgrounds.
Typography
Font: Inter (Google Fonts) — clean, modern, great for data.
Headers: 300-500 weight, tight letter-spacing (-0.02em to -0.03em).
Body: 400 weight, 1.7 line-height for readability.
Libraries
Chart.js — bar charts, line charts, simple stuff.
D3.js — force-directed network graph (the cool floating one).
Everything else is vanilla HTML/CSS.
Prompt for matching this style
"Generate an HTML visualization using this color palette: gold #b8943d, sage #5a8384, lavender #7a6a8a, taupe #8a7060, moss #6a8362, rose #9a6a6a, slate #5a6a7a. Use Inter font from Google Fonts. Light background (#ffffff), muted professional aesthetic. Use Chart.js for simple charts, D3.js for network graphs."
Publish Your Own
Made something cool? Here's how to share it.
1
Save your HTML file
Ask Claude Code to export your visualization as a single HTML file. All CSS and JavaScript should be inline (no external files needed).
2
Privacy scrub
Remove names and identifying info of other people before publishing. Your data, your choice — but be respectful of others.
3
Host it free
GitHub Pages: Push to a repo, enable Pages in settings, done. Free custom domain support.
Netlify: Drag and drop your HTML file. Instant URL.
Vercel: Same deal — free tier is plenty for a single HTML page.
4
Share it
Post on LinkedIn, Twitter, wherever. Screenshots of the cool visualizations work great. Link to the live page for anyone who wants to dig deeper.
Made something? I want to see it!
Share your weirdest LinkedIn inference, your coolest visualization, or just tell me if this was useful. I'm building in public and love seeing what other people create.
Tag me on LinkedIn
If you publish your own version, link back here. Let's see how far this travels.