💡 The 90-Day Window That Predicts ML/AI Job Switches

In recruiting, timing isn’t everything—it’s the only thing.

Our analysis of 2,830 ML/AI engineers in the Bay Area revealed this: the 90 days before a job switch are packed with behavioral signals that often go unnoticed.

When you focus on this short window—not broad quarterly sweeps—you reach candidates when they’re most open to change, cutting through the noise and boosting your conversion rate.

 

🚨 Why 90 Days Is the Sweet Spot

• Open_to_Work (Inferred): Just 15.1% of engineers showed an “Open to Work” signal within 90 days of switching, but they accounted for over half of all job moves in that window.
• Recency Effect: Each 30-day decrease in “Days Since Last Job-Search Activity” boosts the odds of switching jobs within six months by 45% (OR = 1.45).

Bottom line? Recent job-search activity is one of the most powerful predictors of churn—and gives you a narrow but actionable outreach window.

 

🔍 The Signals to Watch

In the 90-day lead-up to a switch, track these behaviors:

• “Open to Work” Flags: Public toggles or subtle shifts in recruiter-response patterns.
• Profile Edits: New certifications, featured projects, or added skills.
• Endorsement Spikes: Sudden surges in peer endorsements (think Python or TensorFlow).
• Promotion Lag: Internal promotions older than 6 months, hinting at restlessness.

Used together, these signals help you turn a noisy market into a precise, high-intent funnel.

 

📋 The 90-Day Outreach Playbook

 

✉️ Tailored Message Templates

• Open_to_Work Lead (Score ≥ 3):
  “Saw you toggled ‘Open to Work’—any interest in leading a new ML platform team?”
• Profile Edit (Score = 2):
  “Congrats on adding that GAN project—would love to discuss applying that at scale.”
• Endorsement Spike (Score = 1):
  “Noticed a spike in Python endorsements—want to hear how we’re using it in our AI research lab?”

 

📊 Measuring & Iterating

• Key Metrics: Track reply rates, interviews, and offer acceptance in your 90-day group.
• A/B Testing: Compare outreach themes (e.g. Open_to_Work vs. Profile_Edit) with a 50-candidate test set.
• Dashboards: Visualize “Reply Rate by Days Since Signal” to find your sweet spot—often around Days 40–20.

 

🧠 Final Thought: The Signals Are There—Are You Listening?

The three months before an ML/AI engineer moves are full of quiet indicators. By tuning into those signals and syncing your outreach to them, you’ll shift from reactive recruiting to proactive placement—boosting engagement and slashing time-to-fill.

 

🔄 Try This Next

• Pilot & Compare: Run a 90-day campaign on 100 engineers—compare against your quarterly approach.
• Refine Scoring: Re-weight signals based on what drives responses, then re-test every 4 weeks.
• Expand Signals: Layer in data like funding round alerts or promotion lags to boost precision.
• 👀 Start watching that 90-day countdown—and watch your ML/AI pipeline come to life.