Mar 18-21, 2013
San Francisco
April 14-18, 2013
June 10-13, 2013
Sept 29-Oct 3, 2013
Oct 14-15, 2013
Oct 23-24, 2013
Nov 4.-5, 2013
For the latest on predictive analytics, see eMetrics’ sister conference, Predictive Analytics World

Predictive Analytics Training

Compute each customer’s future response.

(eMetrics no longer offers this workshop, but its sister event PAW offers a variety of workshops on the topic—see for more).

“At Intuit we're already using data as an asset on the web, but this course makes it very concrete how we can take it to the next level.”

Jared Waxman
Web Analytics Leader, Intuit

100% of October 2008 attendees rated this program
Excellent or Very Good.

What is Predictive Analytics?

Business metrics do a great job summarizing the past. But if you want to predict how customers will respond in the future, there is one place to turn: Predictive analytics. By learning from your abundant historical data, predictive analytics provides the marketer something beyond standard business reports and sales forecasts: actionable predictions for each customer. These predictions encompass all channels, both online and off, foreseeing which customers will buy, click, respond, convert or cancel.

If you predict it, you own it.

The customer predictions generated by predictive analytics deliver more relevant content to each customer, improving response rates, click rates, buying behavior, retention and overall profit. For online marketing, predictive analytics acts in real-time, dynamically selecting the ad, web content or email each visitor is most likely to click on or respond to, according to that visitor’s profile. This is AB selection, rather than just AB testing.

Predictive Analytics for Business, Marketing and Web is a concentrated training program that includes interactive breakout sessions and a brief hands-on exercise. In two days we cover:

  • The techniques, tips and pointers you need in order to run a successful predictive analytics and data mining initiative
  • How to strategically position and tactically deploy predictive analytics and data mining at your company
  • How to bridge the prevalent gap between technical understanding and practical use
  • How a predictive model works, how it's created and how much revenue it generates
  • Several detailed case studies that demonstrate predictive analytics in action and make the concepts concrete

No background in statistics or modeling is required. The only specific knowledge assumed for this training program is moderate experience with Excel.

Who this seminar is for:

Managers. Project leaders, directors, CXOs, vice presidents, investors and decision makers of any kind involved with analytics, direct marketing or online marketing activities.

Marketers. Personnel running or supporting direct marketing, response modeling, or online marketing who wish to improve response rates and increase campaign ROI for retention, upsell and cross-sell.

Technology experts. Analysts, BI directors, developers, DBAs, data warehousers, web analysts, and consultants who wish to extend their expertise to predictive analytics.

In order to meet the unique training needs of Emetrics Summit attendees, this training program is:

  • Business-focused. Unlike other training programs that also cover scientific, engineering and medical applications of data mining and analytics, this seminar focuses squarely on solving business and marketing problems with these methods.
  • Comprehensive across business needs. Within this realm, however, we step beyond the standard application of response modeling for direct marketing to solve the wider range of business problems listed below.
  • Vendor-neutral and method-neutral. This training program, which is not affiliated with any analytics software vendor, provides a balanced view across analytics tools and methods.

“An excellent overview on how to start using predictive analytics in any organization! In just two weeks I already have buy-in from upper management to explore ways to use predictive analytics to improve up-sells, cross sells and to determine what lifestyle imagery to display to our users. I'm super excited about these projects.”

Jennifer Boland
Onsite Marketing Analyst
Sierra Trading Post

Topics covered:

  1. Solving business problems with predictive analytics

    Predictive analytics solves many business problems, offering solutions such as:

    • Increased customer retention by predicting defection
    • Increased online conversions and ad takes by predicting clicks
    • Increased sales and acquisition rates by predicting cross-sell opportunities
    • Personalized web and email content by predicting online response
    • Greater relevancy by predicting customer needs
    • Increased direct marketing response with response modeling
    • Decreased campaign spending by predicting non-responders
    • Increased fundraising profit by predicting donations
    • Higher-valued acquisitions by predicting customer lifetime value

    In other words, customer prediction drives business actions, which deliver business results. We cover case studies across this range of applications, with detailed examples running through both days of the training program.

  2. Creating predictive models

    Data is your most valuable asset. It represents the entire history of your organization and its interactions with customers. Predictive analytics taps this rich vein of experience, mining it to produce predictive models. Where multi-channel data is available, predictive analytics discovers interactions across customer touch points, such as key online behavior that may predict which customers will respond to direct mail.

    Whatever the application, the core methodology of predictive modeling is the same. We will uncover, in concrete terms, how modeling transforms your data into actionable customer predictions. To this end, we will see exactly what a model is, taking a look inside to see how it works and how it is created. Then we will:

    • Explore several example models in action
    • Turn the knobs that tweak and control modeling
    • Compare and contrast modeling methods intuitively, including:
      • Decision trees
      • Business rules
      • Naive Bayes
      • Linear regression
      • Logistic regression
      • Neural networks
      • Other more recent advanced modeling techniques

    Live demos of predictive analytics software. We will include detailed demonstrations of a general-purpose tool that implements multiple predictive modeling methods, as well as CART (Salford Systems), a tool specialized for decision trees. Its friendly GUI-based capabilities make the predictive model transparent so we can drill down and really see the inner workings of specific examples.

    In addition to the products demonstrated, we will discuss the full spectrum of today's predictive analytics software, including free tools, cheap tools, and complete software suites.

  3. Measuring how well predictive models work

    Once you've got a predictive model, how do you know how good it is? We cover methods to evaluate models, which fall into two groups:

    Forecasting: How large a boost in revenue, sales or profit will the model produce?

    Accuracy: How well does it predict, how often is it correct, and how much better is it than standard segmentation such as RFM?

    Deploying a predictive model is playing a numbers game that puts the odds in your favor and improves the effectiveness of campaigns, operations and web behavior. We create profit curves, ROI calculations and bottom-line analyses and talk through exactly what they're telling us. And we prepare for performance gotchas that sneak up on you.

  4. Management and project leadership for predictive analytics

    Although predictive analytics is technical at its core, it must be run as a business activity in order to generate customer predictions that have a business impact. This requires a wholly collaborative process driven by business needs and marketing expertise. This ensures that customer predictions are actionable within your company's operational framework, and that they have the greatest impact within your company's business model.

    Referencing the industry standard data mining process model (called CRISP-DM), we break down the requirements of a predictive analytics business initiative. We explore this process, by which analysts and managers collaborate to strategically position predictive analytics, sustain universal buy-in and understanding, and avoid common roadblocks and unforeseen hazards.

Interactive breakout sessions

Like sky-diving and SCUBA diving, after a few hours of learning predictive analytics, it's a good time to dive right in.  To this end, the training program includes breakout sessions, which are integrated with the conceptual flow of topics covered.  You will join a small team and actively collaborate to design deployment strategies for predictive analytics. Working together to solve specific business problems, you will design strategic processes that avert organizational challenges, and you will design a broad technical approach, including the data discovery, data preparation and evaluatory metrics needed to direct a predictive analytics initiative.

These engaging breakout sessions are conducive to exercising the concepts you've learned, making them more intuitive and ingrained, and also provide an opportunity to learn from colleagues.

You will also "get your hands dirty" by digging through some data with a hands-on exercise during the second day. Optionally working with a buddy for this short exercise of about 20 minutes, you will bring a predictive model to life and see it improve before your eyes.

Course information and registration:

Attendees receive a course materials book and an official Prediction Impact certificate of completion at the conclusion of the Predictive Analytics for Business, Marketing and Web training program.

Feel free to contact Prediction Impact with any questions about this training program.

Start Learning Right Now

The following short, published articles, written by the instructor, are a great place to get started. Note that these articles are not required reading; the material therein will be covered during the training program.

Predictive Analytics with Data Mining: How It Works
Get a handle on the functional value of predictive analytics for marketing, sales and product direction. DM Review's DM Direct.

Driven with Business Expertise, Analytics Produces Actionable Predictions
Run data mining as a business activity to generate customer predictions that will have a business impact. CRM Magazine's DestinationCRM.

Predictive Analytics' Killer App: Retaining New Customers
Predictively targeted discounts convert new customers who would otherwise never return to become loyal customers. DM Review's Extended Edition.

About the instructor

Eric Siegel, Predictive Analytics, Customer Intelligence, Predictive ModelingEric Siegel, Ph.D., is a seasoned consultant in data mining and analytics, an acclaimed industry instructor, and an award-winning teacher of graduate-level courses in these areas. Eric served as a computer science professor at Columbia University, where he developed data mining technology in the realms of machine learning performance optimization, integrating historical databases, text mining, and data visualization. The conference chair of Predictive Analytics World, Eric produced 11 peer-reviewed research publications and ran an MIT-hosted symposium on data mining. He also co-founded two New York City-based software companies for customer/user profiling and data mining. With data mining, Eric has solved problems in CRM analytics, computer security, fraud detection, text mining and information retrieval.

Eric has taught industry programs through Prediction Impact, The Modeling Agency and Salford Systems. In addition, he taught many semesters of university courses, including data mining-related graduate courses as well as introductory lecture series for non-technical audiences. Two of these courses have been in syndication through the Columbia University Video Network. Eric also published three peer-reviewed papers on computer science education.

Expo Hall Hours

May 4th 10:30am – 7:30pm
May 5th 10:30am – 3:30pm

Conversion Conference

Conversion Conferenceis co-located with eMetrics


Digital Analytics Association
WAA Base Camp™ Workshops
Toronto, April 23-24, 2012
Chicago, June 24-27, 2012


Foresee Results Autonomy Optimist




iPerceptions OnResearch Unica Google Analytics


atomics labs TeaLeaf OpinionLab SAS Exact Target Analytics Pros


ClickTale SiteSpect Stangeloop Networks Sawmill Stratigent Truviso Unilytics The University of British Columbia Continuing The University of California | Irvine Brooks Bell Mongoose Metrics CodeBaby

Newcomers Pavillion

Quantivo Loop11 UserTesting Futurenow Metronome Crowd Science Conversion Voodoo

Lanyard Sponsor


Sponsor This Summit