Amazon Data Scientist Career Ladder
Every level of Amazon's data science ladder from L4 to L7 — typical timelines, what changes at each level, why data scientists get stuck, and how the promo doc process drives promotions.
Last updated: 2026-03-23
Aiming for your next level at Amazon?
Get a personalized promotion plan with AI coaching.
Level Overview
| Level | Title | Typical Years | Median TC | Terminal? |
|---|---|---|---|---|
| L4 | Data Scientist I | 1.5–3 yr | $184K | No |
| L5 | Data Scientist II | 2.5–4+ yr | $255K | Yes |
| L6 | Senior Data Scientist | 3–5+ yr | $380K | Yes |
| L7 | Principal Data Scientist | 4–7+ yr | $567K | Yes |
Promotion Cycle
Frequency
Twice yearly (aligned with Forte review cycles in Q1 and Q3)
Decision Maker
panel
Manager-driven with promotion panel review. The manager writes a formal promotion document (promo doc) using input from the data scientist — project outcomes, business metrics, experimentation results, Leadership Principle stories, and stakeholder names. The doc is reviewed adversarially by a promotion panel, then finalized at the Organization and Leadership Review (OLR) where senior leaders calibrate ratings across teams.
Key Details
- •Promo doc is the primary evidence — structured around Leadership Principles with emphasis on business impact and technical depth
- •Panel reviews adversarially — panelists actively challenge claims about analytical impact and methodological rigor
- •Forte review system provides the performance signal (twice yearly) — ratings range from Top Tier (TT) to Least Effective (LE)
- •OLR calibration uses forced distribution — Top Tier (~5%), High Value (HV3/HV2/HV1, bulk), Least Effective (~5%)
- •Data scientists are evaluated on business impact, technical depth, experimentation rigor, cross-functional influence, and LP demonstration
- •L5 to L6 requires influence beyond your immediate team — shaping how others work, not just doing excellent individual analyses
- •L6 to L7 is an inflection point — requires setting vision, defining standards, and influencing org-wide data science decisions
- •L7 promotions require executive sponsorship — your VP needs to champion the case
- •Amazon's back-loaded RSU vesting schedule (5/15/40/40%) makes promotion-driven refreshers particularly impactful
- •Least Effective (LE) rating triggers Focus/Pivot (Amazon's PIP program) — same process as engineering
L4 — Data Scientist I
Entry / New Grad DSEntry-level data science role. You execute analyses and build models under guidance, learn Amazon's experimentation frameworks, and begin demonstrating business judgment alongside technical skills. The focus is on delivering end-to-end projects — from data extraction to insight delivery — and learning to write crisp documents in Amazon's writing culture.
Typical Time at Level
1.5–3 years (typical: ~2.5 years)
Total Compensation (US)
$150K–$215K (median: $184K)
Source: Levels.fyi
Why Engineers Get Stuck Here
- •Not delivering end-to-end projects independently — staying in task-execution mode where someone else frames the problem
- •Technical competence without business judgment — running sophisticated models that don't connect to business outcomes
- •Not building Leadership Principle stories early — promotions require documented LP evidence from your first year
- •Not learning Amazon's experimentation and causal inference standards — the methodological bar matters at L5+
L5 — Data Scientist II
Mid-Level DSThe standard data scientist level at Amazon. You own analytical domains end-to-end — designing experiments, building models, defining metrics, and translating findings into product decisions. This is where most external DS hires land. You are expected to drive data-informed decisions independently, partner with product and engineering teams, and demonstrate Leadership Principles in daily work. L5 is effectively a terminal level — many data scientists stay here.
Typical Time at Level
2.5–4+ years (typical: ~4 years)
Total Compensation (US)
$220K–$300K (median: $255K)
Source: Levels.fyi
Why Engineers Get Stuck Here
- •Impact stays within a single project scope — L6 requires influence across multiple projects and teams
- •Not demonstrating technical depth in experimentation or causal inference methodology — doing standard analyses without pushing the methodological bar
- •Promo doc lacks specific business impact metrics — 'identified insights' doesn't cut it without dollar figures or user impact
- •Not mentoring L4 data scientists — L6 is the first level where developing others is explicitly expected
- •Moving from L5 to L6 within the same team is tough — common advice is to switch teams for visible projects and broader scope
- •Not proactively managing the promotion process — waiting for your manager to initiate rather than building the case yourself
- •Not socializing findings through internal channels — publishing results that nobody reads doesn't build influence
L6 — Senior Data Scientist
Senior DSTechnical leadership in data science. You own the analytical strategy for a product area, define measurement frameworks, shape experimentation practices, and influence product roadmaps through data. This is where most experienced data scientists end up at Amazon. The shift from L5 is not just doing excellent work — it is shaping how others work. You mentor L4/L5 data scientists and are expected to raise the methodological bar across your team.
Typical Time at Level
3–5+ years (typical: ~5 years)
Total Compensation (US)
$310K–$470K (median: $380K)
Source: Levels.fyi
Why Engineers Get Stuck Here
- •L7 requires setting vision and defining standards across the organization — most L6s stay within their product area scope
- •Not influencing organization-wide data science practices — L7 means shaping how the entire org uses data, not just your team
- •Not enough executive sponsorship — VP-level visibility is required for L7 consideration and most L6s don't have it
- •Impact narrative doesn't demonstrate next-level scope — your promo doc needs evidence of L7-level influence sustained over multiple review cycles
- •L7 promotions are rare and require proving you are already operating at that level for an extended period
L7 — Principal Data Scientist
Principal / Staff DSOrganization-wide data science leadership. You set the data science vision for a business unit, define standards and best practices across multiple teams, and influence strategic business decisions at the executive level. Very few data scientists reach this level. Impact is measured in terms of how you've shaped the organization's approach to data-driven decision making.
Typical Time at Level
4–7+ years (typical: ~7 years)
Total Compensation (US)
$480K–$700K (median: $567K)
Source: Levels.fyi
Why Engineers Get Stuck Here
- •Impact limited to a single product area rather than a business unit
- •Not driving data science strategy at the VP level
- •Extremely competitive — very few promotion slots available, and external hires compete for the same headcount
Additional Context
Amazon's data science organization sits across retail, AWS, Alexa, advertising, and other business units. Data Scientists use the same L4-L7 level numbering as engineers and PMs. Amazon has a separate Applied Scientist track focused on ML/AI research with different (often higher) pay bands — Data Scientists focus more on business analytics, experimentation, causal inference, and product metrics. At L5 or L6, data scientists can transition to the Data Science Manager track (L6/L7), leading teams of 4-8 data scientists. The same Forte review system, OLR calibration, Leadership Principles, and Focus/Pivot PIP system apply to data scientists as to all other roles at Amazon.
Data sourced from Levels.fyi (March 2026), Team Blind (verified Amazon employees), InterviewKickstart, InterviewQuery, TeamRora, StrataScratch, and Acciyo. Compensation figures from Levels.fyi and Team Blind. Last verified March 2026.