Exam 1 Prep - Behind Every Good Decision
Chapter 3: Provide a description and application for aggregate analysis.
+ Aggregate Analysis Description - Used to describe a population or a segment or to compare two segments + Aggregate Analysis Application - Descriptive analysis, profiling, campaign analysis, winner-lower analysis
Chapter 2: What is the difference between analytics and testing?
+ Analytics looks at a business event and analyzes the historical data to come up with insights. + Testing on the other hand, is a controlled experiment conducted when you do not have historical data on which to base a decision.
Chapter 4 - What is the BADIR framework?
+ BADIR framework takes us from data to decisions using a set of five lean, streamlined steps that can address 80% of business problems using simple, yet powerful, analytics. + BADIR stands for Business question, Analysis plan, Data collection, Insights and Recommendations
Chapter 7 - business professionals rely on access to data that is? Data analysts require access to data that is?
+ Business professionals should have limited access to aggregated data through a BI tool like Tableau + Data analysts require direct access to an enterprise data warehouse (EDW) where detailed underlying user and event level data is stored like SQL.
Chapter 4 - Describe Step 3 of BADIR framework (the Data Collection phase)
+ Collecting data is step 3 in the framework. Before we look at the data, you need to understand the problem and have an agreed upon analysis plan so you can strategize the data collection effort. Additionally, there are two phases in data collection (1. Data pull and 2. data cleansing & validation) --> Step 1: Data pull- collect the data per the data specification in the analysis plan, which is based on the analysis goal, hypotheses and chosen methodology. Always pull and eyeball a small sample to make sure you are getting what you expected to find. --> Step 2: Data Cleaning and validation - Validate your data as you go about collecting it. Start with a small data sample and compare it with what you expected.
Chapter 3: Provide a description and application for correlation analysis.
+ Correlation Analysis Description - Looks for the relationship between two or more things with the prospect of being able to explain or drive one with the other. + Correlation Analysis Application - Pre and post, test control, drivers, dashboard.
Chapter 3: Provide a description and application for Customer Life Cycle Analysis.
+ Customer life cycle analysis description - Looks at the different stages of the purchase process to determine what stage a group of customers is at and decide how to move them up to the next stage in the purchase process. + Customer life cycle analysis application - Customer progress stages from consideration through purchase to use, sales funnel
Chapter 7 - For component 2, analytics talent, Data scientists and analyst teams need a combination of what four skills?
+ Hands-On Business Analytics and Testing + SQL Skills + Hands-On Advanced Analytics + Stat Tools
Chapter 6 - What is data processing?
+ Most of the data generated by the front-end or received by third-party systems are unstructured and unorganized. They need to be processed, cleaned, combined, and conformed in a logical fashion to support analysis in the data warehouse. + The translation and correlations of raw data into base metric and business entities is a key initial step in the application of business rules and logic to support measuring and reporting in a structured fashion. These operations are called ETL (extraction, transformation, and loading) which are done in regular intervals to keep the database updated.
Chapter 3: Provide a description and application for Predictive Analytics/Time Series analysis.
+ Predictive Analytics/Time Series Description - Looks at both current and historical data to make predications about future events. + Predictive Analytics/Time Series Application - Drivers of conversion or consumer engagement, forecasting.
Chapter 3: Provide a description and application for segmentation analysis.
+ Segmentation analysis description - Groups customers or products into meaningful segments usually to enable better targeting for the purpose of driving higher value through customization. + Segmentation analysis application - Grouping customers or products for targeting and customization.
Chapter 3: Provide a description and application for sizing/estimation analysis.
+ Sizing Estimation Analysis Description - Structured approach to make a near accurate guesstimate in the absence of historical data. + Sizing/estimation Analysis Application - Business case with limited internal data or one that is dependent on external data and assumptions
Chapter 4 - Describe Step 2 in BADIR framework (the analysis plan)
+ The analysis plan has five building blocks: Analysis goals, hypotheses, methodology, data specification, and project plan --> Step 1 of Analysis plan (the Analysis phase) - Start by creating a SMART (Specific, Measurable, Attainable, Relevant and Time bound) analysis goal to answer the business question. --> Step 2 Analysis Plan (the hypotheses phase) - Before determining which data should be collected, come up with hypotheses and criteria to prove or disprove each hypothesis. Hypotheses are best generated through brainstorming session with all key stakeholders. Prioritize and rank them on how plausible or testable they are and their likely impact. --> Step 3 of Analysis Plan (methodology phase) - The type of business question will dictate the analysis methodology or approach that you will choose --> Step 4 of Analysis Plan (data specification phase) - Based on the hypotheses, the criteria to prove or disprove and the chosen methodology, collect the necessary data for analysis. --> Step 5 of Analysis Plan (project plan phase) - The project plan ties all the building blocks together and is formulated only after there is a clear understanding of the analysis plan outlined above. The key elements in a project plan are: the resources; RASCI (who is Responsible for driving this project, who is Accountable, who plays a Supportive role, who needs to be Consulted and who needs to be Informed); Timeline and milestones; Risks; phasing and Prioritization
Chapter 4 - Describe Step 1 in BADIR framework (the business question)
+ The first step is to ask questions at the beginning to identify the problem in its context. Ask relevant questions that enable you to understand the present factors, past events or future strategy driving the request for analytics in the first place. + Do not try to propose a solution to the problem at this stage and also do not ask leading questions
Chapter 7 - What are the main takeaways of chapter 7?
+ To best leverage analytics, first assess your organization's analytics maturity + The four components of analytics maturity are leadership, analytics talent, decision making and data maturity.
Chapter 3: Provide a description and application for trends analysis.
+ Trends Analysis Description - Aggregate or correlation analysis over time, that is trends over a period of time + Trends Analysis Application - Trends of sales, revenues, breaks in trend and segments or drivers over a period of time
Chapter 4 - What is in the executive summary which is a part of phase 5 (Recommendations) of BADIR?
+ We need to write an executive summary which consists of the following: 1) Objective 2) Background 3) Scope 4) Approach 5) Recommendations 6) Key insights with impact 7) Next steps
Chapter 7 - For component 4, data maturity, what are the four primary components?
+ infrastructure + access + usability + instrumentation process
Chapter 2: Analytics and testing is?
+Analytics is analysis of past data to get insights and show relationships. +Testing is creation of new sample data through controlled experiments to derive insights and prove a causal relationship.
Chapter 2: What is the difference between business intelligence and analytics?
+Business intelligence is inclusive of all operations from when data is collected to when data is accessed. +Analytics, on the other hand, is performed on data delivered by business intelligence. Analytics then convert the data to insights, decisions, actions, and, eventually revenue or other impact.
Chapter 1: What is hypothesis-driven analytics?
+Hypotheses are generated by human intuition based on the collective intelligence and experience of stakeholders and their understanding of the business and their environment. +Data validates the hypotheses to come up with a convergent solution. The strength of the solution then will lie in the best of both-data and hypotheses
Chapter 4 - Describe Step 4 of BADIR framework (Insights phase)
+Insights phase (phase 4 out of 5 of BADIR framework) - Once you have the data you can review patterns, prove or disprove hypotheses and present findings.
Chapter 7 - What does it mean to be a data-driven leader?
+Is committed to learning about his or her customer and all dynamics that affect the customer to drive growth and innovation +Inherently believes that data will help drive superior business decisions and outcomes
Chapter 7 - The qualities of an effective data infrastructure are?
+Its design and architecture are open and flexible yet secure + It can scale to handle more data and users + It can expand to handle more than different types of data + It delivers the performance needed to handle complex queries and large volumes of data. + It can interact with a large and changing number of systems, technologies and tools + It is easily and securely accessible for authorized users, whether data analyst or business user.
Chapter 3: What are the top 4 most common analytics methods used?
1) Aggregate analysis 2) Correlation analysis 3) Trends analysis 4) Sizing/estimation (3 less popular analytics methods are include below): 5) Predictive Analytics/Time Series 6) Segmentation 7) Customer life cycle
Chapter 3 - A CMO expects what three key outcomes for business initiatives?
1) Bring more future customers to the door in the most cost-effective manner. 2) Convert more of those who come to the door into customers. 3) Keep the current customers buying
Chapter 1: What are the four steps of hypotheses-driven analytics?
1) Intent - What are you trying to do? What is happening that is not supposed to happen? - This is the intent of your initiative. 2) Hypotheses - What could drive it? Why is it happening? - Answer this question and you now know your potential catalysts and key drivers. Additionally, the answer to the question can help identify key stakeholders and initiatives. 3) Analysis - We could do A or B or C or D to solve this. Based on analysis, doing A and B will likely increase margin. - Here you do analysis to prove or disprove hypotheses with data and find actionable insights. 4) Testing - Analysis says B can increase our margins by percent and A by Y percent. Testing with consumers tells us, they simply love B. - To reinforce the actionability of the insights, it is important to test those insights for early learning and before final execution.
Chapter 4 - What are the five parts to address any sizing problem?
1) Stratification - Dice the problem into smaller pieces and identify segments that behave differently in relation to what you are estimating 2) Correlations and drivers - Determine what metrics and factors could have influence on the metric you are sizing. 3) Assumptions - What do we know about the various factors that make up the equation? 4) Computation - This basically involves doing the math to get the estimates for each segment, while running high, medium and low scenarios to help set boundaries. 5) Triangulation/orthogonal method - This means approaching the same estimation problem with a very different set of drivers.
Chapter 6 - What are the four broad categories of data and analytics tools?
1. Data storage and processing 2. Business intelligence or reporting 3. Business analytics 4, Advanced analytics
Chapter 5 - What is a decision tree?
A decision tree is a top down classification structure based on data generated through recursive partitioning, which simply means repeated evaluation and partitioning at each note until the model is able to deliver the desired results.
Chapter 5 - What is a regression model?
A regression model is a continuous target (the variable we are trying to predict), typically using linear regression, such as modeling the lifetime value of a customer.
Chapter 6 - What is a data warehouse?
After the ETL is done, the structured data flows into tables in the enterprise data warehouse (EDW). EDW stores all information from any number of systems in one place, does not interfere with transactional data and enables a single version of truth as opposed to having numberous independent files and tables.
Chapter 3 - We use what type of analysis for increasing the marketable universe by identifying new channels based on the existing customer profile
Aggregate analysis and sizing/estimation analysis
Chapter 4 - Describe Step 5 of BADIR (Aggregate Analysis)
Aggregate analysis short circuits the action to prove or disprove hypotheses as it is used most frequently when doing descriptive analysis such as profiling or comparative analysis, such as campaign analysis and winner-loser analysis.
Chapter 7 - What are the four areas a company needs to be strong in in order for the organization to be reach analytical maturity?
An organization needs to be strong our areas to successfully leverage analytics: leadership, analytics, talent, decision making and data maturity.
Chapter 7 - For component 3, decision making, an organization should setup what in order for everybody in the organization to understand?
An organization should setup a transparent decision making process
Chapter 2: Equation for analytics for impact is?
Analytics for impact = Data Science + Decision Science
Chapter 2: Analytics is only useful when it drives?
Analytics is only useful when it drives impact. Depending on your business, impact can be revenue growth, process efficiency and improved offerings
Chapter 1: What is analytics?
Analytics is the science of applying a structured method to solve a business problem using data and analysis to drive impact.
Chapter 3 - List the type of analysis used for the scenarios below: 1) Why has conversion dropped post-launch of a product? 2) How many elementary schools exist in New York State? 3) Determine if and why revenue growth for "Toys and All" has slowed down over the last few weeks? 4) Can you tell me which offer worked best in the last marketing campaign? 5) Are our London office employees younger than our Singapore office employees/ 6) What are the time cycles for our customers to go from hearing about us to downloading the free game and then paying for the premium features? 7) Of our one million customers, to which 200K should I send the next marketing campaign to get the best ROI? 8) What are the different use cases for which our customer is using our printers? What does that mean for us?
Answers Below to Questions: 1) Correlation 2) Size 3) Trend 4) Aggregate 5) Aggregate 6) Customer life cycle 7) Correlation, predictive or segmentation 8) Segmentation
Chapter 2: Big Data is a business intelligence issue or an analytics issue?
Big Data is a business intelligence issue not an analytics issue. Big Data is hard to store and render and thus requires special tools and technology.
Chapter 2: Big Data is often explained using the three?
Big Data is often explained using three V's where a very high volume of data with a lot of variety is flowing at a high velocity.
Chapter 5 - What does classification look at?
Classification looks at discrete targets through a decision tree or logistic regression.
Chapter 3 - Understanding engagement drivers (like certain offers, discounts, bundling, loyalty memberships and such) for each of the customer segments is what type of analysis?
Correlation analysis
Chapter 3 - What analysis do we use for understanding drivers of churn?
Correlation analysis
Chapter 3 - What type of analysis is needed for optimizing channels to increase ROI and decrease cost of customer acquisition?
Correlation analysis
Chapter 5 - What is correlation?
Correlation is the statistical measure of the linear relationship between two or more random variables, as represented by the correlation coefficient (r) with a value at or between + 1 and -1. Predictive modeling exploits the inherent correlation between the dependent variable as of T2 and independent variables as of T1.
Chapter 7 - For component, 1 (leadership) the next step after the assessment is?
Creating a vision or mission statement and committing to consistently use data as an enabler to make key-decisions.
Chapter 2: What is data science?
Data Science: This is the technical track, designed to derive insights from data.
Chapter 7 - What does data maturity rely mostly on?
Data maturity relies foremost on a nimble data infrastructure that sales appropriately and supports the kins of access the organization needs.
Chapter 1: Analytics is not just about data but is also about?
Decisions
Chapter 5 - What is a dependent variable?
Dependent Variable - A variable that is the object of the particular predictive analysis. It is determined by the business question that the model is designed to solve.
Chapter 5 - True or false? Predictive analytics is not a powerful tool that does not generate significant business outcomes and it is not resource and time intensive?
FALSE - Predictive analytics is a powerful tool that can generate significant business outcome. Additionally, it is resource and time intensive. Use it judiciously and only when the ROI can be justified.
Chapter 2: Analytics is constrained by Big Data challenges? True or False?
False, analytics is NOT constrained by Big Data challenges. When done right, analytics always deals with subsets of data.
Chapter 2: Is analytics constrained by Big Data challenges? True or False?
False, analytics is not constrained by Big Data challenges. When done right, analytics always deals with subsets of data
Chapter 2: Is analytics reporting? True or False?
False, analytics is not reporting
Chapter 2: The solutions for growth hacking are not created custom to the organization, its goals and its products? True or False?
False, growth hacking are solutions created custom to the organization, its goals and its products.
Chapter 6 - If the data at your disposal is not accurate what should you do?
If the data at your disposal is not accurate it is an urgent call to look at the information architecture and data too.s
Chapter 5 - What is an independent variable
Independent variables are unknowns that may have a relationship with the dependent variable and no relationship with each other. These are determined by the hypotheses developed to solve the business question
Chapter 1: What leads to good decisions
Intuition + Data = Powerful insights --> Good decisions
Chapter 2: What is growth hacking?
It is a process within an organization that has the singular focus of driving scalable growth of a growth metric. Fb is an example where they used this concept to grow from 45 million consumers to 1 billion consumers.
Chapter 4 - When is sizing and estimation analysis utilized?
It is most frequently used for developing a business case with limited internal data.
Chapter 1: How does hypothesis-driven analytics strategy work?
It starts with answering a set of plain sounding questions, "What is happening?" and "Why is it happening?"
Chapter 5 - What is linear regression?
Linear regression is a statistical technique that quantifies the relationship between a continuous dependent variable (y) and one or more continuous independent variables.
Chapter 5 - What are some of the most commonly used predictive techniques in business?
Linear regression, logistic regression, decision trees, k-means clustering, times series forecasting, survival analysis, and neural networks.
Chapter 5 - What is logistic regression?
Logistic regression - a special case of regression in which the dependent variable is not continuous. Instead, it is discrete, or categorical, and mostly binary (0/1).
Chapter 5 - What is misclassification?
Misclassification: One of the ways to evaluate model performance. It is the ratio between the total number of incorrect predications and the total number of predictions. The lower the misclassification, the better the model.
Chapter 6 - What are some driving forces behind Big Data?
New computing platforms, such as Hadoop (an open source, scales through inexpensive commodity hardware, and supports both structured (tables) and unstructured data is the new driving force behind big data.
Chapter 5 - Where can we see predictive analytics in use?
Predictive analytics has seen wide application in improving the customer experience through retention and churn models, customer service operations and customer experience optimization models.
Chapter 5 - What is predictive analytics?
Predictive analytics uses statistical techniques to analyze current and historical facts to make predictions about future events or behavior.
Chapter 5 - What are predictors?
Predictors are unknowns that have a relationship with the dependent variable. The terms predictors and independent variables are sometimes used interchangeably.
Chapter 3 - What type of analysis is needed for segmenting the base to drive engagement?
Simple segmentation-RFM
Chapter 7 - For component 2, analytics talent, business professionals need to develop what three skills?
Skill 1: Hands-on business analytics and testing Skill2: Working effectively with analysts Skill 3: Introductory advanced analytics
Chapter 2: Most useful analytics techniques in business are contained in a small set of simpler techniques that can be learned by most professionals. True or False?
TRUE
Chapter 2: True or False? 20 to 30% of decisions really require the use of advanced techniques like predictive analytics. 70 to 80% of business decisions can be addressed with business analytics or simple analytics techniques?
TRUE
Chapter 2: True or False? Growth hacking is not formulaic, and that is its strength?
TRUE
Chapter 5 - True or false? BADIR can maximize the potential of predictive analytics that otherwise can become a statistical exercise?
TRUE
Chapter 3 - Better targeting of messages and offers based on past marketing campaigns to increase response is what type of analysis?
Testing & correlation analysis
Chapter 3 - What type of analysis is needed for identifying conversion drivers and inquiring about certain fulfillment options, user experiences, review options, cart options, payment options, offers and whether promotions drive incremental conersion?
Testing and correlation analysis
Chapter 3 - What analysis do we use for understanding campaign effectiveness?
Testing, aggregate as well as correlation analysis
Chapter 6 - What is the data tier?
The data tier is the layer where data operations happens. The data tier stores data for serving the logic tier.
Chapter 7 - What does the decision making process require?
The decision making process requires: + A process for collecting all the proposed initiatives, along with information about their expected investment and returns + A team of stakeholders to review new initiatives at a regular interval, such as a monthly or quarterly formal planning an review process + A set of criteria to prioritize one project over another. + A clear description how projects will be executed including process, criteria and metrics
Chapter 7 - For component, 1 (leadership) to enable your company to analyze all its data what is the first step it should take?
The first step should be a detailed assessment of your organization, involving interviews with key stakeholders and a company-wide survey to determine the extent of hands-on experience with data analytics.
Chapter 6 - What are the five components to data storage and processing?
The five components are: Presentation tier, logic tier, data tier, data processing and data ware house
Chapter 7 - For component, 1 (leadership) the greatest gaps in leadership in terms of creating analytics maturity are?
The greatest gaps in leadership in terms of creating analytics maturity are vision, commitment and accountability.
Chapter 6 - What is the logic tier?
The logic tier occurs in the back-end and this is where the logic tier stores all the logic required to perform commands, mathematical calculations, analytical decision-making structure and other operations.
Chapter 6 - What is the most critical aspect of data storage and processing?
The most critical aspect of data storage and processing is its overall information architecture. Information architecture is the design of how information should flow, what gets stores and where.
Chapter 6 - When does the presentation tier occur?
The presentation tier occurs when a customer interacts with the business and generates data.
Chapter 6 - The success of the business intelligence tool is in large part measured by what?
The success of the business intelligence tool is in large part measured by adoption.
Chapter 6 - What are the two parts to aggregation?
The two parts to aggregation is: 1) Create relevant subsets of data: Analyze only the last three years of data. This analysis would be by week and month rather than by day to reduce size. 2) Roll up to appropriate granularity: Granular data can have thousands of metrics. For the user to easily access appropriate info, this needs to be culled down to a manageable number (100-300 by prioritization and aggregation.
Chapter 5 - What is a time window?
There are two time windows of importance as you think about building a model: the observation window and the prediction window.
Chapter 2: What is decision science?
This is the business track, designed to align stakeholders so that the valuable insights produced using the data science track can be inserted into the decision-making process and converted into action
Chapter 2: True or False? Analytics can only prove a relationship (AÑÒB), where testing can prove causation (AÒB).
True
Chapter 6 - True or false? Once you have access to data Microsoft Excel is your best friend to quickly get insights from data that will help solve 80 percent of business problems?
True
Chapter 6 - What are key considerations in choosing the right tool for advanced analytics
Visualization capability, type of data it can handle, cost, integration with other tools, data and user limitation, ease of learning and operational efficiency.
Chapter 6 - If as a business professional you don't have easy access to data through a visual BI tool what should you do?
Work with your head of analytics to quickly enable that.