C-GIDD Analytic Services

  • Home
  • C-GIDD Analytic Services

EIU Canback provides predictive analytic services to many of the world’s leading companies. Our work often leverages C-GIDD, but there are other standalone efforts.

Our Definition of Analytic Services

Canback’s analytic services focus on medium- to long-term predictive models for strategic, financial, and market development purposes. We neither do short-term demand planning supply chain or other operational/tactical models, nor data mining (which we are skeptical of when models are made to fit the data).

Further, our predictive models typically incorporate several information sources and use them in an integrated fashion. An example of this is our use of the Golder-Tellis model for demand prediction which uses household income, price, market presence (distribution availability), consumer confidence, and marketing spend as independent variables. Each of these are pulled from different data sources.

In sum, for executives who want to create medium- to long-term views of market outlook and competitive dynamics, Canback usually has a solution.

Examples of Deliverables

The following are examples of analytic services projects we have recently delivered:

  1. Demand prediction models at the category or brand/pack level by city with a 5-10 year horizon
  2. Demand simulators to test the impact of price and promotional changes
  3. Elasticities datasets that feed into other models
  4. Macro variable predictions similar to C-GIDD but covering additional variables such as ethnic trends, age bracket dynamics, ownership of household products, and spending by major categories

Delivery Mechanisms

We increasingly deliver our solutions through customized web portals. These portals allow us to update algorithms and datasets and ensure that all users have the same data.

Our less preferred way to deliver the results is through Excel spreadsheets. We prefer Excel over other software packages because our users are familiar with it. We do not use SAS or similar packages because they require specialist user knowledge.