Modern marketing is characterized and challenged by the sheer multiplicity and complexity of data sources and types; the technological drivers and proliferation of channels and devices used by increasingly sophisticated and fragmented consumers all point to a need for firms to develop data capabilities and assets as a strategic corporate resource if competitiveness is to be maintained. Understanding and anticipating customers’ needs and the optimal allocation of marketing resources is predicated on appropriate attention to the acquisition, accessing, analyzing and actioning of data. This module will apply a data science lens towards extracting insight through rigorous analysis of data in well-defined marketing contexts and utilise tools available or indeed developed by the students themselves through their earlier modules. There is a well-recognized deficit of marketing and data science professionals and this module hopes in conjunction with the other modules taken to address this deficit. The pedagogy will involve the careful presentation of a range of marketing/data and decision contexts and the presentation and usage of potential analysis approaches that can deliver insight and allow for better marketing decisions.
(self directed reading elements) understanding the modern ‘datascape’-the volume, variety, veracity and velocity of data in modern marketing contexts, emerging roles in corporate structures regarding data- CIO, CDO and the data scientist.
lead generation metrics & KPI’s, ROI, customer lifetime value calculations CLTV, churn analysis
Clustering approaches, profiling and targeting, marketing implications
Predictive Analytics & data mining
• Campaign Analysis: classification approaches- tree based approaches such as CHAID, CRT etc, logistic regression and neural networks – data examples and analyses – confusion matrices and performance metrics such as sensitivity, specificity, AUC etc. comparing models and methods.• Customer Churn analysis – Survival analysis, Cox regression model, hazard and segmentation, marketing optimization of intervention budgets, software tools such as SPSS, BigMl machine learning, 11 ants etc. predicting and profiling churners, strategic marketing implications.• Customer value analysis: RFM & transactional data, Market Basket Analysis (rule support, confidence, lift etc.), Sequence detection, conducting analyses and marketing implications.
use of Tableau software tool for visualizing data such as Google analytics, Facebook & Twitter.
A review & extension of the features and usages of GA; measurement plans, customized dashboards, goal conversion funnels, ecommerce metrics etc. mobile and app analytics etc.
Miscellaneous techniques: Time permitting; a look at one/some of the following:
• Social network analysis: graphs, measures of centrality, eigen values, small worlds etc.• Recommender systems : content-based filtering, collaborative filtering etc.• Discrete Choice Modelling: Brand choice modelling, Multinomial logit model, and modelling probability of choice, choice-based segmentation, and Case study.
A lab based module is envisaged with usage of relevant software being to the fore – featuring SPSS, IBM modeller and possibly some online machine learning platforms.
Students will have access to a range of data sets and tutorial material relevant to the usage and application of the software.
A range of case studies will supplement and contextualize some of the modelling tools developed.
Significant autonomous learning will also be a feature with selected academic journal readings in the field being undertaken and used for discussion and identification & benchmarking of analyses.
Synergies with the other analytical themes will be encouraged and facilitated especially through group project proposals
Given the collaborative and interfacing features of data science functions; a group project with a team structure is envisaged and to involve generating actionable insights from data. Peer presentation, support and critique (and from staff) will also inform their learning in this project element and generally where possible.
Some practitioner inputs and guest lectures by analytics firms will be envisaged and allow for deep insights into practical and current analytics issues in industry.
A practical applied orientation will be the hallmark of the module and as such each topic will be embedded in a well-defined marketing decision context where meaningful sophisticated analysis techniques are to be identified, deployed and to result in normatively sound insights that a decision maker might apply.
An element of a ‘flipped classroom ‘ approach will feature as students develop competence & confidence in selected areas and it is intended that they develop a shared repository, hosted online, of exemplary learning resources and case material, analyses, code development etc. which can be developed as a resource on a year to year basis.
Students will be encouraged to be active networkers in the predictive analytics field (see links below).
|Module Content & Assessment