INTRODUCTION TO DATA ANALYSIS

A. INTRODUCTION
Draw insight from data to make decision
1. THINKING ABT ANALYTICAL PROBLEM

- Think backward: start w the decision you want to make

+ will an analysis influence the decision? ROI, time

+ understand the context: business, problem,…

- Consider the info. Needed to make that decision

+ what would the output look like?: have a good idea what need to accomplish, be as specific as
possible, can I actually see outcome x or y, any chart/graph to illustrate the different, draft the
final presentation

+ what method and tool? Most straight-forward, simple

+ What data is need?

+ where will the data come from? Internal Data warehouse, outside resource

+ enough time, budget?

2. CONCEPTUAL BUSINESS MODEL

- diagram show how the business functions, important elements, how

=> learn your business, problem, give the context - very important!

3. INFORMATION-ACTION VALUE CHAIN

- Up-stream:

+understand the CBM

+ where the data come from

+ what real phenomenon do they describe

- Mainstream: run analysis, find out the meaning of the data

- Downstream:

+ how the analysis result will be used to make decision

=> most of the time, will work on US and DS

- MT: data analysis, 3 types

+ Descriptive: what happens now or in the past

=> to better understand the business/how it works; apply that to make better decision; provide
input/exploration to the other 2

+predictive: use the past data to predict what may happen in the future

+ prescriptive: explicitly link analysis to decision making by recommendations on what we

+ can combine all 3

- DT: summarize vs interpret the result (simplicity, key points (narratives), graphs) – develop strategy
and plan (more senior, advanced position; evaluate plans and pick the one with meaningful impact; data
to measure the impact for further analysis) – pitch – take action in the marketplace

=> The more you understand about the way the business works and how information flows through
business systems, the better prepared you will be to both execute and interpret your analysis. Also,
the more skill you have in finding and accessing data, the more productive and valuable you will be as
an analyst!

Q: Why do we bring data together into a common location?

- We can establish relationships among data sources

- It’s more convenient for extraction to have data in one place

- Sometimes we can’t access source systems directly (particularly when they are critical to business)

- Source data may be unstructured or not formatted for analysis

4. REAL WORLD EVENTS/CHARATERISTICS

- tech provides enormous data (types +vol.), the data that can be useful and impactful to the business

- PEOPLE

+characteristics (gender, race, status,…)

+Preference: belief, attitude, motivations – harder to get

=> group them for analysis:

+ demographics

+ psychographics

+ technographics

=> may be other categories depending on the business;

+ Location: Physical or Virtual

=> information comes from transaction (purchase,…), consumption (vol, usage behavior);
consumer interaction (social media, customer service)

- OBJECTS

+ physical charac: size, color,…

+ Location vs movement: delivery, sale location

+ Function: machine data, intersection of machine & human physical data

- EVENT/ENVIRONMENT

+ natural event

+ Non-natural event: traffic, election,…

+ Both: outbreak of disease

=> everything starts in the real world

Awful lots of thing out there to potentially access and use

5. DATA CAPTURE BY SOURCE SYSTEMs

1. core enterprise systems: ties directly to the financial operations of a company:

+ billing, invoice system

+ enterprise resource planning (ERP)

+ supply chain management: the flow and storage of goods and services through a system
=> track raw material/product from original points to consumption; provide insight: throughput,
inventory

+ accounting

2. CUSTOMER & PEOPLE

- Customer relationship management (CRM): track and manage cus. interaction across all touch points
and cus.’s life cycle w the company – very broad!

- Customer care: direct contact w cus. (call), keep record (note, cmt)

- Lead Management / Sales Force Automation Systems: keep track of potential cus.; record sale
performance, link to other system to help manage orders, contracts, any sales related process

- Campaign Management Systems: track mkt campaign performance, can tie directly to CRM/sales

- Human Resource Systems: internal, employee related

- Electronic Health / Medical Records: detail of patients, sensitive data

- Product management

- Content mana.: store contents (web)

- Web mana. And analytics: issue w web, activity and performance

4. TECHINICAL OPERATION

+ process monitor

+ alarming & fault monitor

+ticketing and workflow

+ Telematics & Machine Data Processing Systems: data directly from machine, feed other system –
exciting when IOT and connected home

5. EXTERNAL

+Credit Agencies

+ Demographics & Segmentation Providers

+ Partners & Suppliers

=> a lot of available data out there, thinking backward from the problem -> where to look for the
certain data in the business
B. ANALYTICAL TECHNOLOGIES
Tools to store, manipulate and analyze data

The most important tool is your brain

1. DATA STORAGE

- data storage and databases