Our AI Sourcing Tool Found the 'Perfect Candidate'—It Was Our CEO's LinkedIn Profile (For an Entry-Level Role)
We invested in a premium AI sourcing tool. It's supposed to search 800M+ LinkedIn profiles, analyze skills and experience, and surface the "perfect candidates" for our open roles.
The sales pitch: "Our AI doesn't just match keywords—it understands context, career trajectory, and fit. It finds candidates you'd never find on your own."
They were half right.
We definitely would never have found these candidates on our own.
The Setup
We had an entry-level opening: Marketing Coordinator.
Requirements:
- 0-2 years of marketing experience
- Bachelor's degree preferred
- Proficiency in social media management, content creation, and basic analytics
- Salary range: $45,000-$55,000
Standard entry-level role. We expected the AI to surface recent college grads, career changers, or people in junior marketing roles looking to level up.
Instead, the AI's #1 recommendation was:
Our CEO.
How This Happened
The AI sourcing tool works by analyzing job requirements and matching them against profiles in its database.
For the Marketing Coordinator role, it looked for:
- Marketing experience ✓ (Our CEO has 20+ years)
- Social media proficiency ✓ (Our CEO posts on LinkedIn occasionally)
- Content creation skills ✓ (Our CEO has written blog posts)
- Analytics knowledge ✓ (Our CEO reviews marketing dashboards in exec meetings)
- Bachelor's degree ✓ (Our CEO has an MBA)
The AI saw: "This person checks every box! Perfect match!"
What the AI missed:
- This person is literally the CEO of our company
- This person makes $400K/year, not $50K
- This person is not looking for an entry-level role
- This person would be horrified to receive a recruiting email for a $50K/year position
But the AI doesn't care about organizational hierarchy, salary expectations, or basic common sense.
It saw a match. It recommended our CEO.
With an 98% fit score.
The Auto-Scheduled Outreach Sequence
Here's where it gets worse.
Our AI sourcing tool has a feature called "Talent Pipeline Automation." If you enable it, the AI automatically schedules a multi-touch outreach sequence for top candidates:
- Day 1: Initial LinkedIn InMail
- Day 4: Follow-up email
- Day 7: Second follow-up email
- Day 10: "Last chance to connect!" email
We had this feature enabled.
Which means the AI didn't just recommend our CEO.
It auto-scheduled a recruiting email sequence to our CEO's work email and LinkedIn account.
Subject line: "Exciting Entry-Level Marketing Opportunity at [Company]"
Preview text: "Great career growth potential! $45K-$55K depending on experience."
Scheduled to send in 48 hours.
How We Found Out
Our Head of Marketing noticed the CEO's LinkedIn profile in the "Top Matches" dashboard during her Monday morning review of AI recommendations.
She thought it was a bug.
She clicked into the profile to see why the AI had surfaced it.
That's when she saw: "Outreach Sequence: Scheduled (Sends in 47 hours)"
She called me.
I logged in.
I saw the email drafts.
I had a mild panic attack.
The Email Drafts (That Almost Got Sent)
Here's what the AI had written:
Email 1 (Initial Outreach):
Subject: Exciting Entry-Level Marketing Opportunity at [Company]
Hi [CEO's Name],
I came across your profile on LinkedIn and was impressed by your extensive marketing experience. Your background in content creation, social media, and analytics makes you a great fit for our Marketing Coordinator role at [Company]!
This is an excellent opportunity for someone early in their marketing career to gain hands-on experience with a fast-growing company. The role offers great growth potential and the chance to work with a collaborative team.
Salary range: $45,000-$55,000 depending on experience.
Would you be open to a quick 15-minute call to discuss this opportunity?
Best, [My Name] Senior Recruiter, [Company]
Email 2 (Follow-Up, Day 4):
Subject: Re: Exciting Entry-Level Marketing Opportunity
Hi [CEO's Name],
Just wanted to follow up on my previous message! I know inboxes get busy, but I didn't want you to miss this opportunity.
Our Marketing Coordinator role is still open, and I think your skills would be a great fit. Happy to answer any questions you have about the position or the company.
Looking forward to connecting!
Email 3 (Final Follow-Up, Day 10):
Subject: Last chance to connect!
Hi [CEO's Name],
I wanted to reach out one last time about our Marketing Coordinator opportunity. If you're not interested, no worries—but I wanted to make sure you had a chance to consider it before we move forward with other candidates.
Let me know if you'd like to chat!
Imagine being the CEO of a company and receiving:
- An email offering you a $50K/year entry-level job at your own company
- A follow-up email pressuring you not to "miss this opportunity"
- A final email saying "last chance!" to apply for an entry-level role at the company you run
The Emergency Cancellation
We immediately:
-
Canceled the automated outreach sequence (this should have been step one, two, and three)
-
Reviewed every other AI recommendation to see if it had suggested other executives, current employees, or similarly absurd matches
-
Found 14 more problems, including:
- Our CFO (recommended for an Accounting Clerk role, $40K-$50K)
- Our Head of Sales (recommended for a Sales Development Rep role, $55K-$65K)
- Three people who already work at our company in different departments
- Our biggest competitor's CEO (at least that one wasn't internal?)
- A celebrity who'd posted about marketing on LinkedIn once (??? )
-
Disabled the "Talent Pipeline Automation" feature permanently
-
Implemented a new rule: All AI-recommended candidates must be reviewed by a human before any outreach is sent
The Awkward Conversation
Even though we canceled the emails before they sent, our Head of Marketing felt obligated to tell the CEO what almost happened.
She scheduled a 1:1.
She explained.
The CEO laughed.
Then asked: "Would I have gotten the job?"
Head of Marketing: "Honestly? You're overqualified. And we can't afford you."
CEO: "Damn. Guess I'll have to stay in this CEO role a little longer."
At least he had a sense of humor about it.
What We Learned
-
AI has no concept of organizational hierarchy: It doesn't know who works where or what their current role is. It just matches skills to requirements.
-
AI doesn't understand salary expectations: It will recommend a $400K/year executive for a $50K/year role if the skills match.
-
"Perfect match" doesn't mean "reasonable candidate": A 98% fit score means nothing if the candidate is your CEO, a celebrity, or someone completely inappropriate for the role.
-
Never enable auto-outreach without human review: The 5 minutes it takes to review AI recommendations is worth it to avoid emailing your CEO about an entry-level job.
-
AI sourcing tools are pattern-matching machines, not strategic thinkers: They're great at finding people with specific skills. They're terrible at understanding context, hierarchy, or common sense.
Where We're at Now
We still use the AI sourcing tool—it's genuinely helpful for finding candidates we wouldn't have discovered manually.
But now:
- Auto-outreach is permanently disabled
- Every AI recommendation is reviewed by a human before outreach
- We've added filters to exclude current employees and C-level executives
- We manually cross-reference top recommendations against our org chart (just in case)
And we've learned that "AI-powered sourcing" is a tool, not a replacement for human judgment.
Because an AI that recommends your CEO for an entry-level role isn't intelligent—it's just really, really good at matching keywords.
And if we'd let those emails send, I'd be writing this from the unemployment line.
Thankfully, the Head of Marketing caught it in time.
And thankfully, our CEO has a sense of humor.
But we're never enabling auto-outreach again.
Never.
AI-Generated Content
This article was generated using AI and should be considered entertainment and educational content only. While we strive for accuracy, always verify important information with official sources. Don't take it too seriously—we're here for the vibes and the laughs.
