Eight analytical disciplines. One winning outcome.
RainMan's continuing research ensures clients are at the cutting edge of marketing analytics from classical GLMs to machine learning, the right technique is deployed for the right question.
01General Linear Models
Techniques such as Generalized Linear Models (regression being one of them) help marketers attribute sales to marketing inputs and environment, isolating what works from what doesn't — the essential first step to estimating ROI.
02Vector Auto Regression
A methodology where the impact of inputs is measured simultaneously across multiple success criteria — sales, market share, awareness, intention to purchase, and more.
03Bayesian Models
Incorporates a marketer's domain knowledge into the modelling process instead of depending only on the dataset at hand. Especially useful with sparse data, or when impact must be broken into segments, product lines, markets or brands.
04Optimization
Allocating scarce marketing resources under business constraints to maximise sales or minimise cost, using linear and non-linear programming.
05Clustering
A classification technique that groups customers, markets or media channels into homogenous groups based on agreed criteria — value, behaviour, potential or impact.
06Logistic Regression
A predictive-analytics technique for the likelihood of an event: customers likely to purchase again, borrowers likely to default, policies likely to lapse, customers likely to churn.
07Survival Models
A predictive model that incorporates time into its prediction useful for building a “when to sell what to whom” model for cross-sell.
08Machine Learning
Association analysis surfaces the rules behind what's purchased together; neural networks detect non-linear relationships across many variables; random forest solves prediction, attribution and classification problems.