高级数据科学家
Senior Data Scientist
你将负责
与数据策略、媒体和客户团队合作，将业务问题转化为清晰、可测试的测量方案
设计和分析增量测试，包括基于地理的实验、保留组、匹配市场测试和其他因果推断方法
构建、验证和解释媒体组合模型，以评估渠道贡献、效率、饱和度和边际收益递减
开发上层漏斗和品牌媒体的测量方法，包括其对下层漏斗结果的直接影响和影响
进行功效分析、测试可行性评估、敏感性分析和模型诊断，以确保结果具有统计可信度
处理大型多源营销数据集；识别数据质量问题、测量缺口以及对分析的影响
将分析结果转化为媒体规划、优化和未来测试的实际建议
应用互补的高级分析方法——包括预测建模、倾向得分建模、细分和预测——以解决更广泛的客户和媒体策略问题
指导和培养初级数据科学家，并为共享的测量标准、代码和最佳实践做出贡献
你需要具备
教育背景：统计学、经济学、数据科学、计算机科学、工程或其他定量学科的硕士学历优先，或本科学历加5年相关经验
熟练掌握Python、R和SQL编程语言
有设计和分析增量测试的实际经验，如地理保留组、匹配市场测试、合成控制组或保留组，或随机实验
有构建、验证和解释媒体组合模型的实际经验
深入理解统计建模、因果推断、实验设计和时间序列方法
能够评估方法论的权衡，挑战薄弱的假设，并根据可用数据和业务决策选择合适的方法
能够在模糊的问题上独立工作，同时与跨职能团队紧密协作
加分项
有使用贝叶斯建模框架（如PyMC）或其他类似工具的经验
有通过增量测试校准或验证MMM结果的经验，或有将多种测量方法整合到统一建议中的经验
有品牌测量、意识研究、零售或线下销售数据，或多结果/漏斗模型的经验
查看英文原文
You Will Be
Partnering with Data Strategy, media, and client teams to translate business questions into clear, testable measurement plans
Designing and analyzing incrementality tests, including geo-based experiments, holdouts, matched-market tests, and other causal inference approaches
Building, validating, and interpreting media mix models to estimate channel contribution, efficiency, saturation, and diminishing returns
Developing measurement approaches for upper-funnel and brand media, including its direct impact and influence on lower-funnel outcomes
Conducting power analyses, test feasibility assessments, sensitivity analyses, and model diagnostics to ensure findings are statistically credible
Working with large, multi-source marketing datasets; identifying data quality issues, measurement gaps, and implications for analysis
Turning analytical findings into practical recommendations for media planning, optimization, and future testing
Applying complementary advanced analytics methods - including predictive modeling, propensity modeling, segmentation, and forecasting - to solve broader client and media strategy questions
Guiding and mentoring junior data scientists and contributing to shared measurement standards, code, and best practices
You Must Have
Education: Master’s degree in Statistics, Economics, Data Science, Computer Science, Engineering, or another quantitative discipline preferred or B.S. + 5 years of relevant experience
Strong programming skills in Python, R, & SQL
Hands-on experience designing and analyzing incrementality tests, such as geo holdouts, matched-market tests, synthetic controls or holdouts, or randomized experiments
Hands-on experience building, validating, and interpreting media mix models
A deep understanding of statistical modeling, causal inference, experimental design, and time-series methods
Ability to evaluate methodological tradeoffs, challenge weak assumptions, and select approaches appropriate to the available data and business decision
Ability to work independently on ambiguous problems while collaborating closely with cross-functional teams
Nice to Have
Experience with Bayesian modeling frameworks such as PyMC or similar tools
Experience calibrating or validating MMM results with incrementality tests, or integrating multiple measurement methods into a unified recommendation
Experience with brand measurement, awareness studies, retail or offline sales data, or multi-outcome/funnel modeling