Advancing the Analytics-Driven Organization

Upon completion of the Advancing the Analytics-Driven Organization course, students will be able to:

• Build critical analytic leadership traits required to establish a sustainable internal analytic practice
• Evaluate and address cultural, environmental and talent issues that frequently derail analytic deployments
• Unify project teams to establish a common strategy, implementation framework and monitoring process for greater coordination, efficiency, clarity and impact:

? For leadership to better understand and trust what the analyst delivers ? For analysts to obtain clear direction on analytic goals and delivery requirements ? For integrators to qualify and prepare for deployment in advance of modeling ? For subject matter experts to be leveraged for domain context and results translation without adding bias to otherwise objective models

• Make right-sized and properly-timed investments in analytic talent, software and supporting resources
• Identify analytic opportunities; then validate, organize and prioritize projects for lowest risk and highest gain
• Leave ad hoc and esoteric statistical exercises behind in lieu of targeted, insightful and understandable analytic outcomes that drive organizational decision making with residual benefit
• Engage in guided peer discussion driven by actual case examples to experience realistic analytic project leadership and implementation issues
• Leave with the resources, contacts and actionable plans to substantially increase targeted analytic outcomes while minimizing dead ends

Advancing the Analytics-Driven Organization is intended for:

• Organizational Leadership – who desire a greater understanding of analytics’ true capabilities, limitations, risks, rewards and high-level function from an unbiased, vendor-neutral perspective. Participants will learn to confidently set expectations, define goals, establish mindset and sell the vision.
• Functional Managers – who are seeking clarity on how to assemble a fully-formed and right-sized analytic team. Participants will learn how to confidently oversee the implementation of a low-risk / high-impact analytic function that provides measurable data-driven decisioning.
• Data Scientists and Analytic Practitioners – who are motivated to enhance their tactical quantitative experience with a pragmatic strategic layer. Participants will learn how to align with organizational project goals and will broaden their capability and value on the analytic team.
• Experienced Statisticians – who wish to increase their value to the organization by applying their classical training to a more goal-driven mindset for prescriptive analytics.
• IT Specialists, Data Enrichment Vendors and Integrators – who wish to gain a better appreciation of the overall analytic process to more effectively prepare resources for data analysis and integrate resulting decision models within today’s sophisticated data and deployment environments.

• While the density of material establishes an aggressive pace in the presentation of content, no preparation is required for this course. Content is presented in nontechnical terms. Exercises are conveyed through demonstrations and guided discussion.

Session I – Core Concepts

• Orientation to data science and organizational analytics
• Trends within the analytically competitive organization
• The advent of data science
  • The Arena: From business unit-based to IT department-based
  • The Professionals: From analyst to data scientist
  • The Analyses: From descriptive analyses / business intelligence to predictive analyses / data mining / machine learning
• What is predictive analytics’ role in Big Data?
  • Big data needs advanced analytics …but does analytics need big data?
  • You will never have a perfect model
  • Market perceptions of big data
• ROI of big data and associated analytics
  • Retail use case
  • Guerrilla marketing use case
  • Medical or government use case
• The future of big data and advanced analytics

Session II – How to Think Like a Data Scientist

• Stats 101 in ten minutes
• A / B testing and experiments
• BI vs predictive analytics
• IT’s role in predictive analytics
• Statistics and machine learning: complementary or competitive?
• Primary project types
  • Predicting a value given specific conditions
  • Identifying a category given specific conditions
  • Predicting the next step in a sequence
  • Identifying groups
• Common analytic algorithms
  • Regression
  • Decision Trees
  • Neural Networks
  • Genetic Algorithms
  • Ensemble Modeling
• Popular tools to manage large-scale analytics complexity
  • R and Python
  • Hadoop, MapReduce and Spark
  • Data Mining “workbenches”
• Performing a data reconnaissance
• Building the analytic sandbox
• Preparing train / test / validation data
• Defining data sufficiency and scope

Session III – The CAO’s ROADMAP

• The Modeling Practice Framework™
• The elements of an organizational analytics assessment
• Project Definition: the blueprint for prescriptive analytics
• The critical combination: predictive insights & strategy
• Establishing a supportive culture for goal-driven analytics
• Defining performance metrics to evaluate the decision process
• What is the behavior that impacts performance?
• Do resources support stated objectives?
• Leverage what you already have
• Developing and approving the Modeling Plan
• Selecting the most strategic option
• Planning for deployment
  • What will the operational environment be?
  • Who or what is the end consumer?
  • How do results need to be purposed or presented?
• Measuring finalist models against established benchmarks
• Preparing a final Rollout Plan
• Monitoring model performance for residual benefit

Session IV – Building The Goal – Centered Analytics Operation

• Attracting and hiring the right analytic talent
• The roles and functions of the fully-formed analytic project team
• Specialization in analytic project teams
• Analytic opportunity identification, qualification and prioritization
• Organizational resistance and developing a culture for change
• Project failure is not the worst outcome
• Staging the organizational mind shift to data-driven decisioning
• Motivating adoption by domain experts, end users and leadership
• Recording ongoing organizational changes
• Monitoring and advancing organizational analytic performance
• “Democratizing” analytics: Advantages and risks of “self-service”
  • Tableau
  • Watson Analytics
  • Establishing performance dashboards
• Standing up an agile analytic modeling factory
• Knowledge retention and skill reinforcement