The Value of Data Data and AI Technologies are Creating Huge Value for Businesses The Value of Data Deal Making Example: Pedestrian Accident Data Data is the world’s most valuable resource Conclusion Data is a new fundamental resource that can be converted into business value in various styles. Data can be used endlessly, meet business needs and can be analyzed for business decision making. The Value of Data: Data Driven Decisions Real Cases: 7-Eleven Thailand Issue: American Express needs a data model to predict customer churn. Introducing facial recognition and AI technology across 11,000 stores in Thailand to : Identify loyalty members Analyze in-store traffic Monitor product levels Suggest products to customers Measure the emotions of customers as they walk around How do we know which customers are important, and what will indicate which are likely to leave? In just one quarter, T-Mobile USA managed to reduce customer churn rates by 50%! How did they do it? What questions did they ask? Goals & Outcomes Preserve 100M global cards, $1T annual charges. Biz model: Targets affluent customers ($150/charge vs. Visa $50/charge). Outcomes: Flagged 24% of accounts to churn next qtr → retention marketing. Reduced personalization model from three days to 20 minutes. Targeted-ads increased online customer acquisition by 40%. Case Studies Data Framework Part 1: Identifying Objectives and Questions Part 2: Data preparation and analytics Real Cases: T-Mobile Billing Analysis Drop Call Analysis Sentiment Science Used data science and data analytics to help answer : Draft objective: Reduce the churn rate of credit card customers. SMART objective: Reduce the churn rate for Super Prime credit card customers by 15% compared to the same Q last year by looking for factors that affect churn and finding measures to reduce those factors by 1 quarter. EXAMPLE: SMART Objective Where do data come from? 4 V's of Big Data Shows how often, where, and how long a user calls with whom If a customer moves and data shows he/she gets limited coverage in the new area, a customer rep is alerted to offer a new phone to prevent the customer from switching. Predicts triggers and indicators of future customer actions and their perception of T-Mobile. This helps T-Mobile to proactively respond to any complaints. Everyone working towards and objectives understands the who, what, how, and the why of the objective. The objective is measurable with data collection available. It’s feasible based on historic data and budget. It supports overall business goals; ladders up to a goal above it Clear start and end date. Example: Bangkok Credit Service (BCS) would like to know why Super Prime credit card customers cancel their usage and switch to credit cards from competitors Reduce time spent on administrative tasks in service center Why do some administrative tasks require more time than others ? How are we going to use this information ? Always a good starting point to clarify the objective further Five why’s Find the root cause of a problem or objective and identify the full picture What else do you think I should know ? Analysts should always end on this question to surface unexpected insights Which ‘call to action’ is driving the hightest conversion? Which products have the highest profit margin? Increase reach of Facebook advertising, whilst maintaining conversion rate. Increase number of hight-profit products sold EXAMPLE: Objectives Guides Questions Let’s Get SMART with Our Objectives! When solving data problem: aim for Insights which leads to action Actionable ! Organized and analyzed Raw Insights Information Data Does the spending amount from the last 3 billing cycles and the card's validity affect the churn? Does the amount of customer tax payment in the past 3 years affect the churn? Question Specific Measurable Attainable Relevant Time-bound Conclusion From objective to question: Three Tactics for Getting to a Good Question Setting SMART objectives and Hypothesis-Driven questions is an important step to start a Data Analytics Project. Data Analytics Workflow Frame Develop hypothesis-driven questions for your analysis Select, import, & clean relevant data Structue, visualize & complete your analysis Create insights and business decisions from your analysis Present data-driven findings and recommendations to your audience Prepare Analyze Interpret Communicate Department Operations Marketing Sales Objective Question VOLUME VELOCITY VARIETY VERACITY 2021 © TRUE DIGITAL ACADEMY Data Analytics Intermediate I Module1: Intro to Analytics Data Analytics Intermediate 1 Intro to Analytics Module 1