Section4
Section4 strives to make elite business education accessible to all.
This immersive 3-week course, aptly called a Sprint was my favorite training to date.
Structured on pre-recorded videos from Tom Davenport , interactive livestream practicums with Industry practitioners ( Alexa Garrison, Director of Data and Corporate Strategy at Convene & Jon Francis, the former head of Global Analytics at Paypal) along with Teaching Assistant support and peer learning through Slack.
§ Provided a high-level view of Data Analytics as a practice & business strategy with real-world case studies of Dominos Pizza, FICO, Moderna, JetBlue, CapitalOne, UPS.
§ Offered applied learning via a final project which was reviewed by peers (390 + from 28 countries representing corporations such as Google, CapitalOne, L’Oreal, Disney, Amazon & Nike).
COURSE OUTLINE & CONCEPTS COVERED
The Power & Process of Analytics
o Types of Analytics – Descriptive, Predictive & Prescriptive
o Steps Involved – Collect Raw Data, Run a Model, Quality Assurance & Deployment
o Traditional Analytics vs. Machine Learning
Determining what problems are suited for Analytics
o Pneumonic = (U-DATA-I) “Put Data between U & I”
Untested, Defined, Acute, Testable, Actionable & Impactful
o Determine UAI first before assessing the other criteria
o Analytics is an iterative process
o Resist the urge to run into the analytics and first spend time fully understanding the problem & involve others (stakeholders) in framing the problem.
Exploring and Evaluating Data
o Quantity does not mean Quality. Evaluate the quality of the data in terms of:
Size, Reliability, Uniqueness & Accessibility
Designing Models:
o Once you have collected the data and framed the problem, the first step is creating a hypothesis to test. Understanding the business model is critical in this first step. Businesses can be oriented toward Value Creation, Value Delivery, Value Capture or a combination of all three.
o In order to have high confidence in your model, strive to give it context by incorporating various types of data (eg. Binary, Numeric, Ordinal, Categorical), assigning weights to various independent variables and monitor over time.
Interpreting Results:
o Is your hypothesis supported or refuted by the analysis?
o Determine your confidence in the results
o Explore correlation and causation (A/B testing)
Competing in Analytics as a Business (DELTA)
o Data driven companies relentlessly gather and leverage high-quality data (eg, the Boston Red Sox example, Airbnb, Amazon & Zillow)
o Enterprise wide approach of setting an analytics strategy and roadmap, manage a unified data analysis platform & improve data literacy throughout the organization (eg Capital One and its “information based strategies)
o Leadership – company leaders appreciate the importance of data and analytics and tout it to the whole organization. (UPS example)
o Identify clear business priorities and a feasible roadmap. Targets are lofty but attainable goals.
o Recruit and retain talented Analysts by providing challenging problems, support continuous learning, encourage analytical decision-making and offering competitive compensation.