I Created 115 Fake LinkedIn Profiles for My Agency - Here's What Actually Happened

I'm Mike. I run a LinkedIn outreach agency in Austin.

This is what happened to me before I decided to engage the team at LinkedSDR, I’m sharing so you can avoid the painful journey.

.

.

.

Two years ago, I decided to create fake LinkedIn profiles to scale my agency. The plan was 100 profiles.

I actually lost count of how many I created. Definitely over 100. Maybe 120.

Most died within weeks. A few survived long enough to connect to automation. But in the end, none of it mattered.

This is the complete story—the plan, the panic, and what actually happened.

Why I Decided to Create Fake Profiles

My agency was handling LinkedIn outreach for clients. We'd been using client accounts, but one got restricted and the client was furious.

I realized: I can't keep risking client accounts. I need profiles we control.

The math seemed simple:

• 100 profiles = massive outreach capacity
• No client account risk
• Full control over campaigns

I researched online. Found guides. The process seemed straightforward.

I was wrong.

The Process (And All My Theories About What Was Wrong)

Step 1: Generate Identities

I used AI tools to create fake identities:

• Photos: ThisPersonDoesNotExist.com for AI-generated faces
• Names: Random name generators for American-sounding names
• Backgrounds: Created job histories at real companies

Time per profile: ~30 minutes to make it look real

Step 2: Set Up Infrastructure (And Constant Troubleshooting)

This is where things got complicated.

Email setup:

• Started with Gmail accounts
• Switched to ProtonMail thinking it was not as "generic"
• Tried custom domains with professional email addresses
• Even tried warming email domains before creating LinkedIn accounts

I kept thinking: "Maybe the email service is the problem."

Domains:

• Bought multiple domains to spread profiles across
• Thought maybe LinkedIn flagged accounts created with the same domain
• Spent extra money on "aged" domains

Phone numbers:

• Virtual phone verification services
• Different providers to avoid patterns

Cost kept climbing. I was now at $15-20 per profile just for infrastructure.

Step 3: Build Out Profiles (Small and Steady)

I learned to space everything out. Controlled. Measured. Small and steady wins the race.

Not all at once:

• Week 1: Create 5 profiles
• Week 2: Create 8 more
• Week 3: Create 10 more
• Keep going gradually

For each profile:

• Professional headshot
• Detailed work experience
• Education background
• "About" section
• Skills listed

Time per profile: ~1 hour to make it complete

My thinking: "If I space them out, LinkedIn won't see the pattern."

Step 4: The Warm-Up

60-75 days of warm-up per profile:

• Daily logins
• View profiles
• Like posts
• Gradual connection requests starting at 5-10 daily

I kept adjusting based on what I thought was wrong—maybe warming up too fast, maybe the names were too generic, maybe I needed more variation.

I changed the warm-up process multiple times, thinking each iteration would be the fix.

Month 1: The Panic Begins

More than 30 profiles dead within the first month.

Not gradual restrictions. Sudden bans. Sometimes within 48 hours of creation.

My theories about what went wrong:

Theory #1: We scaled too fast on same IP/fingerprint

• "Maybe creating multiple accounts from similar locations flagged us"
• Invested in more sophisticated anti-detection tools
• Better browser fingerprinting
• More advanced IP masking
• Didn't help

Theory #2: The email domains

• "Maybe ProtonMail is flagged"
• Switched back to Gmail, then to custom domains
• Still getting banned

Theory #3: The photos are too fake

• "Maybe LinkedIn's AI detects AI-generated faces"
• Started using stock photos
• Started editing AI photos to make them look more real
• Still getting banned

Theory #4: The names

• "Maybe the names sound too generic"
• Used more unique names, added middle initials
• Varied ethnic backgrounds
• Still getting banned

Theory #5: The connections

• "Maybe I'm connecting with the wrong people"
• Changed targeting to only highly relevant profiles
• Only connected with 2nd degree connections
• Still getting banned

I was troubleshooting constantly. Every theory led to another dead end.

Month 2: It Got Worse

Dead accounts > new accounts being created.

I was creating 10-15 profiles per week. But losing 15-20 per week.

The math wasn't working. I was going backwards.

New theories:

Theory #6: The spacing wasn't enough

• Reduced creation to 5 profiles per week
• Didn't matter

Theory #7: The warm-up sequence was wrong

• Extended warm-up to 90 days before any outreach
• Profiles still got banned during warm-up

Theory #8: The infrastructure needed to be more sophisticated

• Bought better proxies (residential, not datacenter)
• Used more advanced anti-detection browsers
• Spent more money on setup per profile
• Still getting banned

Theory #9: Something about the profile content

• Rewrote job descriptions to be more detailed
• Added more skills
• Changed profile photos multiple times
• Nothing helped

At this point, I'd lost count of how many profiles I'd actually created. Definitely over 100. Maybe 110? 120?

I just knew that almost all of them were dead.

A Few Survivors (But Still...)

Out of 100+ profiles, maybe 10-15 survived long enough to actually use.

These profiles:

• Made it through warm-up (60-90 days)
• Had 50-100+ connections
• Got connected to automation tools
• Started sending connection requests for actual client campaigns

I thought: "Finally! These ones work! I figured it out!"

Reality:

• Lasted 1-2 weeks under automation
• Then got restricted
• All of them eventually banned

Even the "successful" profiles weren't successful. They just died slower.

The Real Cost

Direct costs:

• Virtual phone numbers: ~$1,200
• Email services and domains: ~$400
• VPNs, proxies, anti-detection tools: ~$200
• Total: $1,800

Time costs:

• Creating 100+ profiles: 50+ hours
• Warming up profiles: 40+ hours
• Troubleshooting and testing theories: 20+ hours
• Infrastructure setup and changes: 10+ hours
• Total: 120+ hours

Working profiles after 4 months: Zero

Some survived temporarily. None lasted long enough to generate actual ROI.

What I Finally Understood

After months of theories, troubleshooting, and failures, here's what I learned:

1. There was no single "problem" to fix.

It wasn't the email service. Or the photos. Or the names. Or the IP masking. Or the warm-up sequence.

The problem was fundamental: they were fake profiles. LinkedIn's AI doesn't just look at one variable. It looks at everything—patterns, behaviors, authenticity signals across hundreds of data points.

2. LinkedIn's detection is always learning.

While I was creating profiles and testing theories, LinkedIn's AI was learning from each ban. The detection got better. Faster. More sophisticated.

By month 2, profiles were getting caught faster than month 1. The system was evolving while I was trying to outsmart it.

3. Small and steady doesn't help when the foundation is fake.

I thought spacing out creation would help. Slowing the warm-up would help. Being measured and controlled would help.

None of it mattered. Slow or fast, the profiles were still fake. The pace didn't change the outcome.

4. A few survivors don't mean success.

Out of 100+ profiles, having 10-15 survive temporarily felt like progress. But they all eventually died. Temporary survival isn't success.

5. Better tools don't fix a broken approach.

I kept buying more sophisticated anti-detection tools, better proxies, more advanced fingerprinting. I thought better technology would solve the problem.

It didn't. The approach itself was flawed.

What Actually Works

After this disaster, I finally did what I should have done from the beginning: rented LinkedIn accounts from real people.

The difference:

Fake profiles I created: All banned within months, $1,800 wasted, 120+ hours lost

Real profiles from real people: Actually work, stay active, generate results

The monthly cost seemed high until I calculated what my "cheap" approach actually cost.

If You're Considering Creating Fake Profiles

I understand the temptation. I was there.

But here's what will happen:

You'll spend months creating, troubleshooting, testing theories. Each failure will lead to a new theory. Each new theory will fail.

You'll waste money on infrastructure, tools, services trying to find the "right" setup.

You'll lose count of how many profiles you've created because so many die so fast.

And you'll end up where I ended up: needing to work with legitimate profiles anyway.

Learn from my mistake. Skip this entirely.

Frequently Asked Questions

Can fake LinkedIn accounts still work in 2025?

No. A few might survive temporarily, but they'll get banned eventually. LinkedIn's AI detection catches fake profiles within days or weeks. The time and money investment isn't worth the near-zero success rate.

What's the biggest mistake when creating fake LinkedIn profiles?

Thinking there's a "right way" to do it. I spent months testing theories—better IPs, better photos, better warm-up sequences. None of it mattered. The fundamental problem is they're fake, and LinkedIn's AI detects that regardless of how carefully you create them.

How many fake profiles actually survive long-term?

Out of 100+ I created, maybe 10-15 lasted longer than 60 days. All of those eventually got banned once I started using them for actual outreach. Long-term survival rate is effectively zero.

Do better anti-detection tools help fake profiles survive?

No. I invested in sophisticated anti-detection browsers, residential proxies, advanced fingerprinting tools. The profiles still got banned. Better tools can't fix the fundamental problem—the profiles are fake and LinkedIn's AI detects that through behavioral patterns, not just technical fingerprints.

What should I do instead of creating fake profiles?

Work with legitimate LinkedIn profile rental services that provide real people using their own accounts. Monthly cost is higher, but accounts actually work and stay active because they're real people, not fake identities.

Bottom Line

I created 100+ fake LinkedIn profiles. Lost count of the exact number. Spent $1,800 and 120+ hours.

Tested every theory I could think of. Changed variables constantly. Invested in better tools. Troubleshot endlessly.

None of it mattered. They all got banned.

If you're considering this, learn from my expensive mistake. LinkedIn's detection is too sophisticated. Save your time and money. Start with what actually works.

‍