Brad Donovan, Manager, Agile Analytics and Innovation, GlaxoSmithKline

SuperNova Award Category

Data to Decisions

The Company

GSK employees serve US communities by discovering, developing, and delivering new medicines, vaccines, and other healthcare products to help people do more, feel better, and live longer.

The Problem

The IT & Analytic communities could not respond when Marketing and Sales needed new insight. There was also a disconnect between IT & Analytics. The biggest problem was that the Analytics community was working outside the main EDW, using different data models and naming conventions as pushed out by IT in specific data marts.

The Solution 

Teradata, IBM & SAS came on as an advisory council to verify or refute IT's objections to Analytics using data directly from the EDW, using Teradata Data Labs. This is a move from where Analytics was seen as a bunch of PhD stats using SAS and pushing out models.

The Results

Data science became the overseer of this continuum:

  • data access
  • tech as enabler
  • process flow

Providing success in Analytics by working with the above 3 in an ecosystem as an Analytics framework.

The Technology:

  • Teradata, Data Warehouse and Data Labs
  • QlikTech, QlikView
  • SAS
  • Microstrategy

Metrics

Took data aggregation from 30 hours to 3 minutes, model execution from 40 hours to 1 hour, QC from 40 hours to 5 hours. Most importantly took the predictives and models of the Analytics team from a "bright shiny object" to an actionable solution in production, accessible by thousands of business analysts throughout the company.

Disruptive Factor

Most importantly took the predictives and models of the Analytics team from a "bright shiny object" to an actionable solution in production, accessible by thousands of business analysts throughout the company.

Shining Moment 

GSK moving away from the traditional hire to younger Enterprise Miner, R, open source… resulting in flexibility.