Econometrics Specialist

Job Description

About the Role:Marketing Science Specialist Econometrics (Grade 1C) ResponsibilitiesUnilever invests huge amounts of money every year in building different elements of the marketing mix - media, distribution, pricing, promotion, brands. It is imperative to unlock the growth opportunities by optimizing the investments in these elements of the marketing mix.Marketing Science specialist in Econometrics will be working in multi-functional project teams comprising of members from external analytics partners, internal advanced analytics and business leadership teams. The role of the project teams will be to design, build and embed large scale business relevant tools / capabilities in the area of media and mix optimization.Responsibilities of Marketing Science specialist in Econometrics in the project teams will be to help

• Build Capabilities and Tools: in the area of marketing mix and media mix modelling to optimize the investments in different elements of the marketing mix to drive long-term and short-term growth. This will include designing how best to source and transform data for each element of the mix, conduct experiments (working together with analytics partners) to better predict and optimize investments such as harnessing LSTMs / Causal Forest / Transfer learning, Bayesian models for segment / group level models, harness unstructured data to develop early warning indicators of sales movements, apply better optimization methodologies such as MCMC, integrate marketing mix models with A/B tests, evolve how best to model lead-lag indicators, look at new distributions for better estimation of decay parameters, explore best ways of integrating impact on short-term sales and long-term brand metrics.
• Embed and Evolve Capabilities: Support PMRA managers in embedding the outputs / outcomes from the tools with business stakeholders, train business users and CMI community on the use of the tools, help evaluate the capabilities and outputs of analytics partners, work with analytics partners to continuously challenge and improve the current models.

The key to success in the role is the ability to build and democratize at scale sophistication in predictive analytics which delivers better business outcomes.Skills and ExperienceThe skill-sets required for effective implementation of the vision for this role are:

• Expert in Python, R, R Shiny
• Deep Expertise in Marketing Mix Modelling, A/B test or experimental tests
• Familiar with data and data transformations required for integrating media, price/promo, distribution, short and long-term brand metrics in mix models
• Have experience / familiar with some of the recent evolutions in marketing mix models such as
  • Use of Bayesian models for segment level marketing mix models
  • Integration of unstructured data to identify early warning indicators of sales movements
  • Use of machine learning for decomposition
  • Harness ensemble machine learning models to drive prediction accuracy of marketing mix models
  • Calibration of Attribution using Marketing Mix outcomes and vice-versa
  • Pros and Cons of Top-down marketing mix models from aggregated metrics vs. bottom-up propensity models
  • Comparison of fixed vs. variable baseline for integrating long-term brand equity in mix models
  • Opportunities and Challenges of using agent-based models
  • Good hang of statistical and optimization techniques such as MCMC, Bayesian models, Agent Based Models, Genetic Algorithms
  • Ability to design and program simulators for experiments (need not be production ready at scale)
  • Familiar with technology provider tools and capabilities such as Facebook Prophet model
• Strong expertise in working with external analytics partners in designing and evaluating experiments to address business objectives for example, how to account for potential of new product launches in driving future growth via-a-vis other marketing mix levers without biasing the model
• Ability to partner with Data Architecture and Engineering to help productize Innovations
• Have worked hands-on in a commercial environment in the area of analytics for marketing mix or media optimization
• Have managed stakeholders both commercial and technical stakeholders
• Possess business acumen and ability to work in ambiguous ever-changing environment.
• Enthusiastic about driving Innovation, working in connected networks and have demonstrated evidence in this regard
• Ability to communicate technical findings in a business-friendly language
• Hungry for creating something new and make a mark

Good Project Management skills in prioritizing multiple tasks,

Keyskills :
printer fleet managementbrand equity marketing mixearly warning project teamsbusiness acumen machine learningretail analytics