Guidance and results for the data rich, yet information poor.

Data Mining Escort Series

Ensure Project Success With Experienced Guidance

Table of Contents

  1. Escort℠ Overview
  2. DM Project Assessment
  3. Data Exploration Package
  4. Knowledge Discovery Package
  5. Information Modeling Package

Escort℠Overview

Successful data mining requires a rare blend of science and art. In the 1990s, pioneering companies realized the potential advantages of employing data mining technology as early as possible. They chose to undertake this initiative in-house. Too often, these innovators would prematurely dismiss the technology when they fell victim to common, yet elusive data mining project pitfalls. This left the door open for their competitors to enjoy the advantages of ‘getting there first.’

Other progressive organizations decided to maintain focus on their core competencies andoutsource data mining. They too rarely met with success. Although their consultants may have had adequate modeling and industry expertise, critical project management issues specific to data mining as well as the integration of internal domain expertise were overlooked.

Data Mining Escort℠ Product Model

The Modeling Agency (TMA) holds a philosophy that successful data mining projects involve a structured collaborative effort between the enterprise and the data mining expert. To that end, TMA has developed a progression designed to:

  • Establish an immediate starting point with a clear yet flexible process for data mining Reduce start up costs via a compartmentalized progression
  • Reduce risk through early reporting and actionable results at the conclusion of each stage
  • Distribute knowledge between the collaborative team through a workshop-oriented progression to ensure redundant support capabilities
  • Ensure project success based on effective management of expectations and highly effective data mining project management principles
  • Provide “consultative products” with fixed, firm prices; making a cooperative effort easier to visualize, justify, budget and purchase
  • Data mining projects are highly susceptible to failure without a thorough assessment and resulting plan

A Sample Case: Customer Relationship Management

With The Modeling Agency’s experience spanning most major application areas, Data Mining Escort may be readily adapted for virtually any business or industrial modeling challenge. It will be far more substantive to describe TMA’s Data Mining Escort using a tangible business application.

Customer Relationship Management (CRM) has become a key business initiative for organizations of all sizes – making it an excellent business case, as it is both a vertical and horizontal market application.

“Successful customer relationship management is the competitive advantage of the Information Age and is the bottom line for market leadership,” write Peter Fingar, Harsha Kumar, and Tarun Sharma in their book Enterprise E-Commerce.


Data Mining Project Assessment

Data mining projects are highly susceptible to failure without a thorough assessment and resulting plan. TMA developed guidelines to provide accurate estimates of works and cost, define responsibilities of the client-vendor team, and yet provide reasonable flexibility required through the knowledge discovery process. One of the most critical components of this process is the Data Mining Project Assessment (DMPA).

The DMPA creates an essential foundation for a successful data mining initiative thorough pre-project assessment, understanding key stakeholder motivations and vision, leveling expectations by educating on strengths and limitations, gaining an appreciation of each unit’s business processes, assuring end user adoption in advance, defining and valuating benchmarks for project performance, and requesting domain expert contribution for results interpretation.

Data Mining Project Assessment Deliverables:

  • On-site interviews (two days) with stakeholders, functional and database managers, domain experts and user group representatives to understand needs, objectives and resources
  • Assessment Briefing presented within 10 days of the on-site interviews, typically via PowerPoint and teleconference although on-site briefings may be arranged as well
  • Assessment Notes will be a collection of written findings, plans, and other artifacts amassed during the course of the DMPA
  • Supporting Proposal will present a customized plan of actionable recommendations using the subsequent Escort℠ series as a framework

Recommended Training

In order to provide a stronger platform for collaboration, TMA recommends that technical leaders and project managers attend TMA’s public Data Mining: Level I course, while analysts and IT staff participate in the Data Mining: Level II course or both levels depending on previous exposure to data mining-related technologies. The best effect will be achieved if the training is completed prior to the project assessment or data exploration stage.


Data Exploration Package

This initial implementation package begins with a data exploration workshop. Valuable insights into your customers’ buying behavior, spending patterns, geographical disbursements and other distributions will be uncovered almost immediately, graphically displayed and interpreted. This new information is only the first step in a process that will lead to sophisticated customer profiling and highly effective personalized campaigns.

Within two weeks of the conclusion of the workshop, a summary report of findings and recommendations specific to your data will be prepared by a database marketing expert. The report will assess your present data collection processes, data quality and modeling potential. The recommendations report will provide a customized actionable plan for more advanced analysis and strategic direction based on your database marketing goals.

Data Exploration Package Deliverables:

      • Fifteen hours to remotely conduct TMA’s Data Certification Process and workshop preparation by the workshop consultant in order to apply statistics, generate reports, prepare visualizations, and assess initial results interpretations
      • One hour pre-workshop phone meeting to brief on preparation results, confirm agenda and coordinate resources
      • Data Exploration Application Workshop:
        Two days at your site
        • Review Topical Data Analysis Results [1 day]
          • Recency
          • Frequency
          • Monetary Value
          • Location
          • Visualization Exercises
          • Strategic Consulting [1 day]
            • Measure capabilities of additional modules and data mining in the context of your data
            • Data mining technology valuation exercises
            • Tools survey and recommendations
            • Expert recommendations based on analysis results, assessment findings and your goals
            • Within two weeks of the workshop, a findings and recommendations report will be submitted, completing this package


Knowledge Discovery Package

The Knowledge Discovery package will provide a bridge from the analysis performed in the data discovery session into advanced data mining capabilities.

The introduction of data mining technology will expose a new level of intelligence to apply to your campaign management strategy. Beyond the capabilities of standard statistics and database queries, data mining can reveal additional insights through prediction and advanced segmentation. Armed with this technology, you may increase your competitive lead and marketing efficiency by understanding and anticipating your customer’s behaviors and preferences. You will be able to create customer-focused marketing strategies such as:

            • More Effective Cross-Selling
              Based on previous buying patterns and profile characteristics, you will be able to predict what each customer is most likely to purchase next
            • Predictive Customer Valuation
              Giving you the ability to predict each customer’s value over time, allowing you to target the most desirable customers for your business – or forecast when your most valuable customers are about to leave
            • Automated Market Segmentation and Profiling
              You will be able to cluster customers into various groupings and differentiate them in order to tailor campaigns to each based on their characteristics
            • Propensity and Response Modeling
              Scoring or identifying the most likely candidates to respond to a given promotion based on their previous buying patterns, psychographics and demographics, location and other attributes

By applying these methods to your marketing strategies, you will gain a comprehensive view of your customers, allowing a more targeted allocation of your marketing resources.

Knowledge Discovery Package Deliverables:

Knowledge Discovery Application Workshop I
Two Days at Your Site

            • Introduction to Data Mining
                • Definition
                • What data mining can and cannot do
                • Data Mining Process
            • Working Introduction to Modeling
                • Loading and Accessing Data
                • Creating and Using Data Dictionaries
                • Generating Models
                • Viewing Reports
                • Scoring Models
            • Begin the development of a cross-sell model
                • Drill-Down I: Data Manipulation
                • Accessing Your Data
                • Defining Features
            • Drill-Down II: Generating Models
                • Supervised Learning algorithm options
                • Unsupervised Learning algorithm options
                • Advanced Algorithms
            • Drill-Down III: Viewing Reports
                • Using reports to assess model performance
                • Using reports to answer business questions
                • Customizing Reports
            • Drill-Down IV: Scoring Records
                • Scoring options
                • Deployment options
                • Understanding reports and customizing

Knowledge Discovery Application Workshop II
Two Days at Your Site within 4 Weeks of KD Workshop I

    • The development of the Cross-Sell model continues by iterating through the knowledge discovery process introduced in Workshop I, day 2. Additional iterations will serve to refine the model and reinforce the instruction of knowledge discovery.
    • The model’s results will be translated into marketing and monetary impact specifically for your business. As well, the benefits of data modeling and automating the newly discovered knowledge will be described.

Information Modeling Package

Now that we have exposed valuable knowledge about the dynamics of your market and customers through data mining, it is time to assimilate that advanced information into your marketing strategy.

Your modeling consultant will develop a migration plan for the refined Knowledge Discovery model toward an intelligent database marketing process. Imagine setting up a system whereby the value measured from the single model can be invoked with ease and reused for multiplicative effect.

Beyond building a customized roadmap for an intelligent marketing process, two more prototype models will be developed in separate workshops. These prototypes will offer a minimal-case projection of the potential marketing impact of other data mining model types such as response modeling, segmentation or customer valuation. The potential will be described in terms of better understanding your customers, increasing your competitive lead, and leveraging a much greater return on your marketing dollar.

Information Modeling Package Deliverables:

  • Information Modeling Application Workshop I
    • Two days at your site within 4 weeks of KD Workshop II
    • We begin with the model created in the Knowledge Discovery package. A customized conceptual roadmap will be developed. The roadmap will chart the steps required to develop a comprehensive marketing program driven by the model’s results. The latter part of the workshop will be focused on the selection of two new prototype models to develop, and the collection of corresponding data.
      • Information Modeling Application Workshop II
        • Two days at your site within 8 weeks of IM Workshop I
        • Development of a second model by stepping through the knowledge discovery and information modeling process. A model may be selected from the following menu:
        • An additional Cross-sell model
        • Customer valuation (contingent on descriptive data)
        • Response model (contingent on historical data)
        • Market Segmentation
        • The model’s results will be translated into marketing and monetary impact specifically for your business.
      • Information Modeling Application Workshop III
        • Two days at your site within 8 weeks of IM Workshop II
        • Development of a third model
        • By stepping through the knowledge discovery and information modeling process. A model may be selected from the following menu:
            • An additional Cross-sell model
            • Customer valuation (contingent on descriptive data)
            • Response model (contingent on historical data)
            • Market Segmentation
            • The model’s results will be translated into marketing and monetary impact specifically for your business.
            • 40 Hours of “wire support.” Wire support is technical support or consultative guidance offered by a modeling consultant via telephone, Email, fax, on-line chat, whiteboard or video-conference. All requests will be answered within three business days.

Experience & Qualifications

TMA has provided on-site data mining workshops for the following:

                  • Alltel
                  • eBay
                  • Acxiom
                  • Quaero
                  • PC Mall
                  • GM Card
                  • Ararmark
                  • LA Times
                  • Circuit City
                  • US Cellular
                  • Day-Timers
                  • Dollar Bank
                  • HCIA-Sachs
                  • KSL Resorts
                  • Radio Shack
                  • TN Marketing
                  • Rural Cellular
                  • Choice Hotels
                  • Manhattan East
                  • Bank of America
                  • CMS Enterprises
                  • AmeriMark Direct
                  • Wearguard-Crest
                  • Corporate Express
                  • Kaiser Permanente
                  • United Retail Group
                  • Jockey International
                  • Chrysler Corporation
                  • Maritz Loyalty Marketing
                  • Nissan Motor Company
                  • Baltimore Sun Company
                  • Kohl’s Department Stores
                  • USAA Insurance Company
                  • Starwood Hotels & Resorts
                  • Milwaukee Journal Sentinel
                  • Ingalls Healthcare Systems
                  • AAA Life Insurance Company
                  • American Society of Civil Engineers
                  • Kerzner International North America
                  • Fund-Raising Division of a Research Hospital
                  • Information Services International – Dentsu (Japan)

About The Modeling Agency

Although The Modeling Agency’s name is tongue-in-cheek, it’s by no means a misnomer. The Modeling Agency (TMA) provides predictive modeling support for those who are data-rich, yet information-poor through a structured network of independent consultants, or agents. To borrow a very fitting phrase from the popular Star Trek® series, each consultant adds their own uniqueness and industry background to TMA’s collective experience across numerous vertical markets.

To view TMA as simply a contracting and assignment company would fall well short of TMA’s true function. The structure shared by TMA’s consultants is provided by a standardized data mining process and project management guidelines specific to data mining which is inherently a discovery process. TMA’s business model provides a flexible framework, redundancy and continuity across its network, as solid as any company with a full-time staff. TMA has been privileged to provide services and solutions for dozens of highly recognizable organizations.


Senior Consultant Qualifications

Each consultant possesses three main qualities: First and most obviously, they are well steeped in data mining methods and techniques. Secondly, they’re business consultants who will assess each business situation and recommend an appropriate strategy to avoid building excellent models that answer the wrong questions. Thirdly, they are “multi-lingual” — able to engage and interact effectively whether with analysts, IT staff, functional managers or executive management.

All senior-level modelers hold a post-graduate degree in mathematics, computer science or related field, and a minimum of five years experience in:

                  • Developing and deploying highly effective predictive models
                  • Retail applications, propensity / response modeling, database marketing, churn analysis, segmentation and campaign management
                  • Leading highly visible data mining projects
                  • Training and translating complex technology into clear business language

References

Upon a successful discovery call, TMA will proudly provide direct contacts to customers who have already completed workshops. We trust that you too will honor us with the same courtesy in the future as a highly satisfied customer.


Make The First Move!

There are numerous ways to jump off the starting block:

                  • Call 281.667.4200 for an exploratory phone call
                  • Extend an invitation via our web form

1Based on unknown data conditions, model results will vary. A “best effort” model will be produced within the workshop time available.
2 A completed prototype model deliverable is one which has created an initial scored output file. Accuracy of results will vary, based on previously unknown data conditions.
3 All unused services expire 6 months after package initiation. We encourage full consumption of services to attain maximum effect.