- Principal architect of FICO® credit-scoring franchise during 15-year career with Fair Isaac Corporation, inventors of the FICO credit scoring system.
- EVP-Financial Services with global responsibility for Fair Isaac's dominant business line worldwide.
- 12 years in private consultancy in marketing and credit analytics, and provision of expert witness services in credit score-related litigation.
- All 7 Best Practices
- Pre-Call Discovery Process
- One-on-One Call with Expert
- Session Summary Report
- Post-Session Engagement
Predictive Analytics Strategy
- More and more of what we do is being recorded digitally.
With ever-increasing frequency, the actions we take and even our mere presence results in permanent digital records. Life events are accompanied by bar-coded licenses, and our movements are readily tracked. Our facial patterns are logged, and our heights and weights and detailed ailment histories are recorded on our medical records. The majority of our financial activities including our purchases, and the personal likes and dislikes we log on social media, form vast stockpiles of information.Now more than ever as we live life we are creating data. As a result, there's more data to be used and more applications for such use -– more messages and more actions that businesses take that relate to us. Predictive analytics sits right on top of this trend.
The increase in data allows us to deploy more analytics. And the reason it matters is that it can drive efficiency in business, and indeed completely spark new businesses and industries that would not exist without it.
- Mobile devices enable data creation, but also form a channel for deployment of actions.
The evolution of a strong mobile channel has created many additional data elements. But privacy issues and rules governing the use of data that could be mined from mobile devices are also still developing. Mobile devices are evolving as a key channel for most of the types of businesses that benefit from predictive analytics. The channel allows businesses to communicate readily with us, and to do so in the context of our physical location, movements and actions, all of which can change momentarily.
- "Smart" devices are becoming predictive.
More and more products are becoming smart. Your refrigerator may soon be creating your grocery shopping list and telling you when something in it is about to go bad. The next level of "smart" is technology that senses and records behavior, and then predicts. We are seeing this in thermostats that adjust their settings when the pattern of household activity changes – effectively predicting the comfort demand to meet.
- There's a growing recognition of the importance of predictive analytics.
Before 2007, there was no college degree program in analytics. There were degrees in statistics or mathematics or operations research, engineering, various related areas, but nothing that was focused on analytics and certainly nothing focused on predictive analytics. But today, there are 49 analytics degree programs in the United States.
- There's growing transparency – and growing consumer awareness of these technologies.
- Back when I was working at FICO, nobody knew what a FICO score was. FICO had to build that awareness, and it took years to do it. Partly thanks to that awareness, today there is an assumption in the financial world that most of the decisions are driven by some kind of predictive model.
Along with this trend come privacy concerns. Consumers have expectations regarding transparency that businesses using this technology cannot afford to ignore. They also have expectations about what information is fair and reasonable to use in a particular setting. Firms make explicit decisions about what they are going to be transparent about and what they're not.
- The rules governing consumer data use are evolving.
No matter what the sector, companies cannot ignore the evolving landscape of privacy and data use. There will be more attention paid to the flow of consumer data. As predictive analytics becomes more prevalent we can expect new rules regarding what consumer data can be used, and for what purpose. As rules take hold, they can have an impact on the development of new predictive models and may force changes in those that are already deployed.