【FPT Japan】Senior Data Scientist（リテール・マーケティング分析）/ 東京勤務 / 年収1,000万〜1,600万円

**Company:** FPT Japan

**Location:** Tokyo, Japan --- Onsite with hybrid flexibility

**Employment Type:** Full-time / Permanent

**Openings:** 3--4 position

**Salary:** **¥10,000,000 -- ¥16,000,000 JPY / Year** (commensurate with experience)

**Languages:** Japanese (JLPT N2 or higher) · English (Conversational / Business level)

About the Role

We are seeking **3--4 Senior Data Scientists** to join a high-impact analytics program supporting one of the world's leading global apparel retailers. This is a hands-on, client-facing role where you will own end-to-end analytical workstreams --- transforming complex retail data into trusted decisions that shape product supply, pricing strategy, store network expansion, and marketing investment across Japan and international markets.

You will not just build models --- you will explain their findings, validate their reliability, and guide business stakeholders toward evidence-based choices. Each successful candidate will anchor one or two core domains while contributing flexibly across the broader program.

Core Domains \& Responsibilities

1. Demand Forecasting \& Supply Planning

* Diagnose and correct systematic bias in existing ML forecasting models (LightGBM, time-series approaches)
* Improve model features, evaluation frameworks, and production pipelines --- including migration to Databricks
* Analyze forecast accuracy across product categories, channels, and seasonal periods
* **Business Impact:** Optimize production volumes, inventory levels, replenishment timing, and product availability

2. Pricing, Promotion \& Marketing Measurement

* Measure price elasticity, markdown response, and promotion lift using causal inference and hierarchical modeling
* Review, strengthen, and recalibrate Marketing Mix Models (MMM) --- including adstock, saturation, ROI calculation, and budget guidance
* Reconcile multiple data sources to build consistent product and promotion histories
* **Business Impact:** Sharper discount strategies, more effective campaigns, smarter marketing budget allocation

3. Store Network \& Location Analytics

* Evaluate store performance, format efficiency, and customer catchment areas using geospatial and statistical methods
* Assess new-site potential, trade-area overlap, and cannibalization impact on existing locations
* Build scenario models to support expansion, relocation, and network investment decisions
* **Business Impact:** Lower-risk store openings, optimized network layout, targeted market coverage

4. Cross-Domain Analytics \& Data Foundations

* Build and maintain reliable data pipelines and reusable datasets on **Google Cloud / BigQuery / Databricks**
* Standardize business definitions, automate data-quality checks, and ensure reproducibility across all workstreams
* Document data lineage, assumptions, and limitations for full transparency and auditability
* **Business Impact:** One trusted source of truth, reduced reconciliation effort, faster adoption by client teams

Key Responsibilities --- Across All Domains

* Translate ambiguous business questions from merchandising, supply chain, store development, and marketing teams into structured analytical plans and testable hypotheses
* Apply rigorous validation: chronological splits, backtesting, holdout samples, sensitivity analysis, and benchmarking against existing models
* Clearly distinguish correlation from causation; communicate uncertainty, limitations, and confidence levels honestly
* Write maintainable, well-documented SQL (BigQuery Standard SQL) and Python workflows; use Git for version control and code review
* Present findings, trade-offs, and recommendations concisely to technical and non-technical stakeholders --- in both Japanese and English
* Take full ownership of analysis quality; verify all results before sharing; never submit unchecked output
* Collaborate with offshore and global colleagues; share knowledge openly; support junior team members as needed
* Formal people management responsibilities are not required

Qualifications

Required

* **5 years of hands-on professional experience** in Data Science, Applied Statistics, Econometrics, or Advanced Analytics
* Proven track record of independently owning full analytical or modeling deliverables from design through to stakeholder presentation
* Strong statistical foundation: regression, time-series analysis, hypothesis testing, controlled comparison design, out-of-sample validation
* Practical experience building and evaluating ML or statistical models --- including gradient boosting (LightGBM), forecasting, and response models --- with clear understanding of model mechanics, limitations, and failure modes
* **Advanced SQL skills** for querying and joining large multi-source datasets in a cloud data warehouse
* **Strong Python skills** (pandas, scikit-learn, equivalent) for data preparation, modeling, validation, and workflow automation
* Working knowledge of data modeling principles: data grain, temporal joins, slowly changing dimensions, lineage, and validation at scale
* **Japanese:** JLPT N2 or equivalent --- confident participating in meetings, discussing analysis, and presenting findings with business stakeholders
* **English:** Conversational / Business level --- comfortable reading documentation, attending team meetings, and coordinating with international colleagues
* Bachelor's or higher degree in Data Science, Statistics, Computer Science, Mathematics, Economics, Engineering, or a related field

Strongly Preferred

* Domain experience in **retail, apparel, e-commerce, supply chain, merchandising, or marketing analytics**
* Deep expertise in one core domain: demand forecasting, pricing/promotion analytics, store/location analytics, or marketing mix modeling
* Experience with **causal inference or quasi-experimental methods** (A/B testing, geo-lift designs, matched comparisons)
* Hands-on experience with **Google Cloud Platform, BigQuery, Databricks** or similar cloud analytics platforms
* Experience reviewing existing models or third-party solutions and communicating findings to non-technical stakeholders

Nice to Have

* Bayesian or hierarchical modeling experience (PyMC, Stan)
* Geospatial analytics tools and methods
* MMM frameworks (Robyn, Meridian)
* Orchestration tools (Airflow, dbt/Dataform)
* Consulting or cross-border project delivery experience

Candidate Profile

* **Self-directed:** You move work forward independently while knowing when to align and ask clarifying questions
* **Quality-minded:** You verify every result; you distinguish evidence from assumption and communicate limitations honestly
* **Collaborative:** You share methods, document clearly, and support the broader team
* **Client-focused:** You listen deeply, understand the real business need behind the question, and explain complex topics simply
* **Flexible:** You adapt as priorities shift across workstreams

First 90 Days --- Success Milestones

**Days 1--30**

Complete structured knowledge transfer; map key data sources, business definitions, and existing model pipelines; reproduce and verify established baseline results

**Days 31--60**

Independently trace a business question from raw source data through to final output; deliver a reviewed improvement to analysis, validation, or data quality

**Days 61--90**

Own a full analysis or model enhancement end-to-end; present business impact and limitations; leave behind reproducible code, documentation, and a prioritized roadmap

What We Offer

* Opportunity to directly influence major commercial decisions at a globally recognized brand
* Professional growth in applied data science, econometrics, cloud engineering, and client consulting
* International, collaborative team environment
* Comprehensive onboarding and knowledge transfer program