Analytical Worlds Blog – Predictive Analytics and Text Analytics – by Eric Siegel, Ph.D.

By: Eric Siegel, Founder, Predictive Analytics World 

In anticipation of his upcoming conference co-presentation, Building a Digital Marketing Data Science Discipline at Cisco: Experiments to Excellence, at Predictive Analytics World Chicago, June 8-11, 2015, we asked Viswanath Srikanth, Senior Program Manager, Analytics at Cisco, a few questions about his work in predictive analytics.

Q: In your work with predictive analytics, what behavior do your models predict?

A: Our models predict for our digital visitors, topics of interest, lead potential, likelihood to consume a type of offer, and probable path through the site.

Q: How does predictive analytics deliver value at your organization? What is one specific way in which it actively drives decisions?

A: Predictive Analytics is used for understanding propensity to buy, amount of potential purchase as well as identifying leads. Identified leads are actively pursued by our outbound marketing teams (email, contact centers) and by our Sales teams. The lead potential is a very useful filter in prioritizing which leads to pursue, significantly cutting down on unproductive calls.

Q: Can you describe a successful result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: I cannot disclose the actual lift percentage here, but suffice it to say attaching scores to leads allowed for a double digit improvement in number of deeper conversations we had with potential customers (that is, took the next step along the journey).

Q: What surprising discovery have you unearthed in your data?

A: Perhaps not entirely surprising – but our digital visitors want better information, timely information and want to figure out solutions to their problems – and the more intuitively we can make our digital property in helping them get there, the more likely they are to engage in deeper conversations with us.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A: The focus of the talk will be on the journey that Cisco's Digital team undertook in moving from an environment of experimenting in data science to making it an entrenched part of the business – and we hope the audience will be able to apply some of the successful aspects (and avoid the unsuccessful aspects) in their own business – both from a business process and modeling aspect.

———————

Don't miss Viswanath Srikanth’s conference co-presentation, Building a Digital Marketing Data Science Discipline at Cisco: Experiments to Excellence, and workshops at Predictive Analytics World Chicago on Tuesday, June 9, 2015, from 3:05-3:25 pm.  Click here to register for attendance.

By: Eric Siegel, Founder, Predictive Analytics World

No Comments yet »

By: Eric Siegel, Founder, Predictive Analytics World

In anticipation of his upcoming conference keynote presentation, UPS Analytics – The Road to Optimization, at Predictive Analytics World for Business Chicago, we asked Jack Levis – Senior Director, Process Management at UPS, a few questions about his work in predictive analytics.

Q: In your work with predictive analytics, what behavior do your models predict?

A: We use a tremendous number of predictive and prescriptive models at UPS. They are used to help make decisions, which range from where to build a facility and what type of aircraft to purchase to which packages go in each trailer and how to maintain our delivery fleet.

We currently have 700 dedicated resources working on a system called ORION, which has been called "arguably the world's largest Operations Research Project." With ORION, we are using analytics to determine the best way for a driver to serve our customers at the lowest cost.

Q: How does predictive analytics deliver value at your organization? What is one specific way in which it actively drives decisions?

A: We do not do anything by the "seat of our pants." Analytics is engrained so deep in our culture, it is difficult to separate analytics-driven decisions from normal business processes.

In 1954, our CEO said, "If we did not have operations research, our rate of growth might have been affected. As we grow in size, our problems increase geometrically. Without Operations Research, we would be analyzing our problems intuitively only, and we would miss many opportunities to get maximum efficiency out of our operations."

Analytics has helped UPS make better decisions in all parts of our business.

Q: Can you describe a successful result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: In 2003, UPS began using predictive models to better plan our delivery operations. This suite of tools called Package Flow Technologies along with Telematics has been responsible for a yearly reduction of 85 million miles driven per year. This reduced our fuel needs by over 8 million gallons and reduced carbon emissions by 8,500 metric tons.

In addition, because the analytics and business processes are fully aligned, we have been able to deploy new products for customers. UPS' MyChoice is a prime example of that.

Q: What surprising discovery have you unearthed in your data?

A: Two surprises come to mind. The first has to do with the maintenance of our delivery fleet. Through the use of Telematics data with diagnostic and predictive models, we can predict vehicles and parts that are getting ready to fail. This has reduced our failure rate and maintenance cost at the same time.

The second surprising discovery is from our ORION optimizations. We have seen significant additional improvements beyond predictive analytics. ORION has been able to determine new ways to run a delivery route that we did not think of through previous methods.

ORION and Telematics are examples of how analytics can take an already efficient organization and make additional gains through predictive and prescriptive modeling.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A: Conventional wisdom says that as an organization moves from descriptive analytics to predictive to prescriptive, the skills need to grow. However, the benefits grow tremendously during this process as well. This matches UPS's experience exactly.

We will show our journey through these stages of analytics, as well as the lessons learned along the way. The presentation will culminate with examples of "non-intuitive" insights that can only be seen through the use of advanced analytics.

Don't miss Jack Levis' keynote presentation, UPS Analytics – The Road to Optimization, at Predictive Analytics World for Business Chicago on Tuesday, June 9, 2015, from 8:50-9:40 am. Click here to register for attendance.

By: Eric Siegel, Founder, Predictive Analytics World

No Comments yet »

By: Eric Siegel, Founder, Predictive Analytics World 

In anticipation of his upcoming conference presentation, Applied Predictive Analytics to Workload Automation, at Predictive Analytics World Chicago, June 8-11, 2015, we asked Arcangelo Di Balsamo, IBM Workload Automation Chief Architect at IBM, a few questions about his work in predictive analytics.

Q:  In your work with predictive analytics, what behavior do your models predict?

A:  As the Workload Automation architect, I'm applying predictive analytics mainly to IT Automation-focused use cases. Those use cases are about:

                1) Predict job duration using history
                2) Resolve issues quickly by learning from the past
                3) Improved insight by detecting relationships across resources
                4) Optimize performances

Q:  How does predictive analytics deliver value at your organization? What is one specific way in which it actively drives decisions?

A:  In my case, predictive analytics is delivering value to the product I develop (it's a workload automation platform which includes also a job scheduler). The predictive analytics is adding to the product a reliable Operational Level agreement monitoring capability (such as being sure that a critical workload completes within a given deadline). Moreover it provides the knowledge required to reduce failures by discovering hidden relationships between jobs and resources, so that if a resource is failing, it's possible to also predict the failure of the job bound to it.

Q:  Can you describe a successful result, such as the predictive lift of your model or the ROI of an analytics initiative?

A:  When you run a million jobs per day, a failure rate of 1% brings to 10.000 failures per day. Imagine how expensive it is to find the root cause of all of these and take the remediation action!  We have customers that with the enhanced predictive capabilities in the product have reduced the failure rate to 0.01%, which is a huge operational cost savings for them.

Q:  What surprising discovery have you unearthed in your data?

A: Frequently, we discover recurring patterns in job duration (i.e., a job taking longer at the end of each month) that are not obvious. We also discover correlations between job failures and resource unavailability (like job1, job2 and job3 fail when the RDBMS is not available).

Q:  Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A:  Predictive analytics applied to workload automation is a tremendous opportunity to better know what's happening within your data center in order to reduce costs and prevent failures and delays in your critical business applications.

———————

Don't miss Arcangelo Di Balsamo’s conference presentation, Applied Predictive Analytics to Workload Automation, and workshops at Predictive Analytics World Chicago on Tuesday, June 9, 2015 from 3:55-4:40 pm.  Click here to register for attendance.

By: Eric Siegel, Founder, Predictive Analytics World

Eric Siegel is the founder of Predictive Analytics World (www.pawcon.com) — the leading cross-vendor conference series consisting of 10 annual events in Boston, Chicago, San Francisco, Toronto, Washington D.C., London, and Berlin — and the author of the bestselling, award-winning book, Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die.

No Comments yet »

By: Eric Siegel, Founder, Predictive Analytics World

In anticipation of his upcoming conference presentation, The Revolution in Retail Customer Intelligence, at Predictive Analytics World Chicago, June 8-11, 2015, we asked Dean Abbott, Co-Founder and Chief Data Scientist of Smarter Remarketer, a few questions about his work in predictive analytics.

Q: In your work with predictive analytics, what behavior do your models predict?

A: I’ve built models that predict a wide variety of behaviors and patterns. A short list is provided here:

• Customer behavior: response, churn, product up-sell and cross-sell, best marketing creative, days to next purchase, days to next visit;

• Signals (radar): tank, truck column of tanks; (sonar): man-made vs. biologic;

• Financial: fraud or suspicion of fraud, debt repayment period, debt repayment amount, insurance claim repayment likelihood, claim amount of repayment.

Q: How does predictive analytics deliver value at your organization? What is one specific way in which it actively drives decisions?

A:  I’ll speak to Smarter Remarketer, Inc., the company I’m Co-Founder of and Chief Data Scientist. There is no one specific way our predictive models drive decisions, but they are involved in the decision-making process in several ways, all related to selecting customers to promote to, whether that be selecting customers to send an email to, show a display ad, or present content on a page that is of greater interest to the customer.

Consider our models that predict the likelihood that someone will purchase a product during a visit to the company’s web site within 3 days. Each visitor is scored while they browse on the website and at the end of their session. The company now wants to create a new campaign to increase sales of a particular product by emailing them a promotion code with a 20% discount. If the customer is likely to purchase a product on the web site within 3 days, the models will exclude these customers from the email list; why take away margin from sales that are likely to occur anyway. Or what if a customer was very likely to purchase within 7 days last week but is no longer likely this week? This is a form of churn (but based on expected behavior, not actual behavior), and these customers could be given incentives to visit again.

Q: Can you describe a successful result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: It is difficult to describe most of the results my models have generated because they are considered sensitive information for the company or government agency. I’ve had models in use by organizations for 10 years before they were refreshed. I’ve had another model so successful that it was put on the “do not tell” list by the organization because it became a strategic initiative for the organization. I’ve had fraud models identify multi-million dollar cases to investigate that were clearly fraud but had previously eluded detection.

Q: What surprising discovery have you unearthed in your data?

A: Most surprising? There have been many surprises over the years, usually related to the data itself and patterns of behavior that we may overlook, but are important nevertheless. For example, with the days to next purchase models, one expects that visitors on a website who look at lots of hot products are more likely to purchase soon; these are engaged visitors. However, it turns out that some of the most likely purchasers are those who visit just one item. The vast majority of the time, one-item visitors are not engaged and therefore are unlikely to purchase. But, if these one-item visitors were previously highly engaged, it’s a different story; they are focused like a laser beam on one product only. So the surprise was that there is this subset of visitors who look awful but are actually fantastic!

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A: The most important take-away in my talk is this: When you prepare data for modeling, think about how the algorithms interpret the data. Each algorithm has weaknesses that can result in strange or misleading behavior. It’s our job as predictive modelers to help the algorithms do the best job they can.

Q: In addition to keynoting, you will be teaching two one-day workshops at PAW Chicago, Supercharging Prediction with Ensemble Models and Advanced Methods Hands-on: Predictive Modeling Techniques. How would you advise attendees to choose between these workshops and would it even make sense to attend both?

A: There are many workshop options, and all of them are worthy of attending. I think of the Supercharging and Advanced Methods workshops as complementary to your Online Introduction to predictive and John Elder’s Modeling Methods, with the sequence being (1) Intro, (2) Modeling Methods, (3) Advanced Methods, and (4) Supercharging. The Modeling Methods can be taken the day before Advanced Methods in the same conference; Modeling Methods provides a framework for predictive modeling, and Advanced Methods lets you try it out on commercial software. Supercharging takes predictive modeling to the next level, introducing the methods that win modeling competitions and have provided me with extra accuracy has made the difference between successful models and very successful models in my consulting practice.

———————

Don't miss Dean Abbott’s conference presentation, The Revolution in Retail Customer Intelligence, and workshops at Predictive Analytics World Chicago on Tuesday, June 9, 2015 from 10:30-11:15 am.  Click here to register for attendance.

By: Eric Siegel, Founder, Predictive Analytics World

No Comments yet »