How to Get Promoted from Data Scientist IC4 to IC5 at Meta
Building your promotion case?
Get AI coaching on what to document and how to present it.
You've been at IC4 for a while now. You own your analyses end-to-end, your product partners trust your work, and your manager tells you you're doing well. But the last PSC cycle came back Meets All (MA), which means you're performing at level, and performing at level doesn't get you promoted.
IC4 to IC5 is the promotion that separates data scientists who stay mid-level from those who reach senior. It's also the last promotion with an up-or-out clock behind it. Meta gives you roughly 33 months from reaching IC4 to make IC5. After IC5, the pressure lifts since it's a terminal level. But getting there requires a different kind of evidence than the IC3-to-IC4 jump did.
What Changes from IC4 to IC5
IC4 is a fully independent data scientist. IC5 is where Meta expects you to lead analytically across a product area and influence decisions that go beyond your immediate team. The title changes from Data Scientist to Senior Data Scientist, and the expectations shift meaningfully.
| Dimension | IC4 (Data Scientist) | IC5 (Senior Data Scientist) |
|---|---|---|
| Scope | Owns analyses for a product surface | Leads analytical strategy across a product area or multiple surfaces |
| Impact | Analyses influence team-level decisions | Analyses shape product direction and resource allocation |
| Methodology | Designs experiments and monitors KPIs | Builds experimentation frameworks others use; sets analytical standards |
| Stakeholders | Works with immediate product team | Influences cross-functional partners (PMs, eng leads, design) at senior level |
| Direction | Executes on team priorities | Identifies high-value analytical questions the team wasn't asking |
| People | Helps onboard teammates | Mentors IC3/IC4 data scientists; raises the analytical quality bar |
The core shift: at IC4, you answer the questions your product team brings you. At IC5, you identify the questions that matter most, design the analytical approach, and make sure the product team acts on your findings. You go from executing on the analytical agenda to shaping it.
How Meta Promotions Work at IC4-IC5
The process is identical to IC3 to IC4 in mechanics: you write a self-review, collect peer feedback from 3-5 nominators, your manager drafts their assessment, and the packet goes to calibration. But calibration evaluates IC4-to-IC5 candidates differently.
At this transition, all four performance dimensions carry real weight:
- Analytical Impact: not just what you shipped, but what changed in the product because of your analysis. The committee looks for product decisions that were directly shaped by your work.
- Methodology: the rigor and scalability of your approach. Did you build something others now use? Did you improve how the team runs experiments?
- Direction: did you identify high-value problems the team wasn't focused on? IC5 data scientists don't wait for requests.
- People: are you raising the analytical bar for the team? Mentoring junior DS, reviewing experiment designs, sharing best practices.
The rating bar is the same: Exceeds (EE) signals promotion readiness. But at IC4, the committee is looking for sustained EE across multiple dimensions, not just strong analytical output. An IC4 who ships great analyses but doesn't influence product direction or mentor anyone has a weaker case than one who does both, even if the analyses are slightly less impressive.
The up-or-out context: Meta gives roughly 33 months from IC4 to reach IC5. Two consecutive Meets Most (MM) ratings trigger an automatic PIP. The clock is real, and it creates genuine pressure for IC4 data scientists who are growing but not yet demonstrating IC5 scope.
How Long IC4 to IC5 Should Take
| Pace | Timeline | What's happening |
|---|---|---|
| Fast | 18-24 months | Clear IC5 scope from the start; strong project fit; visible cross-team impact early |
| Standard | 24-33 months | Steady growth, building scope and influence over 2-3 PSC cycles |
| Slow (danger zone) | 33+ months | Hitting the up-or-out boundary; requires immediate scope expansion or team change |
The compensation jump here is significant. Based on Levels.fyi data, median total comp jumps from roughly $273K at IC4 to approximately $443K at IC5. That's a $170K increase, driven almost entirely by stock grants that roughly triple from IC4 to IC5.
What Actually Gets You Promoted
Shift from answering questions to asking them
The single biggest difference between IC4 and IC5 data scientists is who sets the analytical agenda. IC4s respond to product team requests. IC5s identify the questions worth asking.
This looks like: noticing a metric trend before anyone asks about it, proposing an analysis that reframes how the team thinks about a problem, or flagging that the current experiment design won't answer the question the team actually needs answered. When your product partners start coming to you not because they have a specific request, but because they want your read on what's going on, you're operating at IC5 scope.
Build cross-team influence
IC5 scope means your work affects decisions beyond your immediate team. This could be building an experimentation framework that other DS teams adopt, presenting findings at a cross-team review that shifts a product group's roadmap, or identifying a data quality issue that affects multiple surfaces.
You don't need to force cross-team projects. Look for natural connection points: a metric your team owns that another team depends on, a methodology problem multiple teams share, or an insight from your product area that has implications elsewhere.
Create reusable analytical infrastructure
IC4s run analyses. IC5s build things that make future analyses faster and more rigorous. This might be a reusable experimentation template, a metric definition framework that standardizes how the team measures success, or a monitoring system that catches regressions before someone files a ticket.
This is where Methodology dimension scores come from. Calibration committees notice when a data scientist's work scales beyond a single analysis.
Document and communicate at a higher level
Your self-review needs to tell a story about product impact, not analytical output. Instead of listing experiments you ran, describe the product decisions that changed because of your work. Instead of describing your A/B test methodology, explain why you chose that approach and what the team would have gotten wrong without it.
"I identified that our retention metric was masking a cohort-specific drop in new user activation. Restructured the team's measurement approach to separate first-week and returning-user retention. Product team shifted Q3 roadmap priority based on the gap analysis. The new activation flow launched to 100% of new users and improved 7-day retention by 4.2%."
That's IC5 self-review language. It connects analytical work to product outcomes at a level IC4 descriptions rarely reach.
Invest in the People dimension
IC4 data scientists sometimes treat mentorship as optional. At IC5, it's expected. The committee looks for evidence that you're raising the team's analytical quality, not just your own output. This means reviewing experiment designs from junior DS, helping IC3s learn the product area, sharing methodology insights in team forums, and giving feedback that makes others' work better.
Mistakes That Keep Data Scientists at IC4
Staying reactive. If every analysis you run was requested by a PM or your manager, you're operating at IC4 scope regardless of how good the analysis is. IC5 means proactively identifying what the team should be looking at. The shift from reactive to proactive is what calibration evaluates.
High-quality work with narrow impact. You can run the best A/B test analysis in the org and still get Meets All if the work only affects your immediate team's decisions. IC5 requires demonstrating that your analytical work has implications beyond your product surface.
Treating the People dimension as a checkbox. Doing one mentoring session per quarter doesn't satisfy this dimension. Calibration looks for sustained, visible investment in other data scientists' growth. If nobody on the team is citing your feedback or using frameworks you created, the evidence isn't there.
Writing task-log self-reviews. "Ran 12 experiments and built 6 dashboards" is IC4 language. IC5 self-reviews connect each piece of work to a product decision and explain why the analytical approach mattered. If your self-review reads like a list of activities rather than a narrative of impact, your manager has less material to argue with in calibration.
Not managing the manager relationship. Your manager is the person who presents your promotion case. If they don't know you're aiming for IC5, don't have clear examples of IC5-scope work, or haven't heard you articulate your impact, the calibration packet will be weaker than your actual contributions deserve.
Frequently Asked Questions
How long does it take to get promoted from IC4 to IC5 at Meta?
Most data scientists who get promoted spend 24-33 months at IC4. Strong performers with good project fit and visible cross-team impact can do it in 18-24 months. Meta's up-or-out policy sets a deadline of roughly 33 months. Data scientists who haven't been promoted by then face exit pressure.
What's the pay difference between IC4 and IC5 at Meta?
Based on Levels.fyi, median total compensation jumps from roughly $273K at IC4 to approximately $443K at IC5. That's a $170K increase, driven primarily by stock grants that roughly triple.
Is IC5 a terminal level?
Yes. IC5 (Senior Data Scientist) is a terminal level at Meta. There's no up-or-out pressure to advance beyond IC5. You can stay at IC5 indefinitely. The next promotion to IC6 (Staff Data Scientist) is entirely optional, extremely competitive, and based on demonstrating org-level impact.
What PSC rating do I need for promotion from IC4 to IC5?
Sustained Exceeds (EE) ratings across multiple dimensions are the standard promotion signal. One cycle of EE focused primarily on Analytical Impact isn't enough. The committee looks for EE evidence across Impact, Methodology, Direction, and People. Greatly Exceeds (GE) makes the case very strong but isn't required.
Should I switch teams if I'm stuck at IC4?
If your current team doesn't have the scope for IC5-level analytical work, switching can unlock the promotion. Some teams have limited experimentation volume or product decisions that don't require senior DS input. A team with more product surface area, higher experimentation velocity, or closer integration with senior product leaders gives you more opportunities to demonstrate IC5 scope. But weigh this against the 33-month clock, since a team switch resets your context.
CareerClimb maps your analytical wins to Meta's performance dimensions and shows you where your IC5 evidence is strong and where it's thin. When the PSC window opens, your self-review is backed by documented impact across all four dimensions, not a scrambled task log. Download CareerClimb