22
DATA ANALYTICS Let’s Break it Down Talk at General Assembly , Boston on October 19, 2015 Twitter: @ArpitGupta https://www.linkedin.com/in/TheArpitGupta

Data analytics - Let's break it down

Embed Size (px)

Citation preview

Page 1: Data analytics - Let's break it down

DATA ANALYTICSLet’s Break it Down

Talk at General Assembly, Boston on October 19, 2015

Twitter: @ArpitGuptahttps://www.linkedin.com/in/TheArpitGupta

Page 2: Data analytics - Let's break it down

Hi! I am Arpit Gupta▸ Senior Product Manager, Analytics @Fiksu, Mobile

applications advertising▸ Instructional team at GA’s Data Analytics Course▸ Past: Healthcare consulting for 5 years and non-

profit▸ Analytics for a long, long, time!

Twitter: @ArpitGuptahttps://www.linkedin.com/in/TheArpitGupta

Page 3: Data analytics - Let's break it down

What’s on your mind? Why are you here?

Page 4: Data analytics - Let's break it down

Goals▸ Define data analytics ▸ Why it’s so important▸ The stages of analyzing data▸ What tools are used▸ Recommended next steps for learning to analyze

data yourself

Page 5: Data analytics - Let's break it down

What is Data Analytics?▸ Learn to make sense of data; tell a story; defend your

proposal ▸ We can store data points, but learning from them is an

entirely different skill.▸ Drive business value.▸ Other terms

●Business Analytics●Web Analytics ●Social Media Analytics ●Real Time Analytics●Data Science / Predictive Analytics

Page 6: Data analytics - Let's break it down

How is Data Analytics used?

▸ Transportation▸ Fashion▸ Healthcare ▸ Non-profit | Social Good | Fundraising ▸ Marketing | Advertising▸ Content Strategy | Buzzfeed? ▸ Finance▸ Education▸ Food

Page 7: Data analytics - Let's break it down
Page 8: Data analytics - Let's break it down
Page 9: Data analytics - Let's break it down
Page 10: Data analytics - Let's break it down

What data does Uber have?

What questions does Uber want to answer?

Page 11: Data analytics - Let's break it down

▸ User Acquisition● How many new users are signing up on the platform?● What’s the breakdown by platform, OS● Which sources are most effective in driving new users?

▸ User Retention● What’s the average time before users abandon your product? ● What’s the lifetime value of my users?

▸ Revenue● Which city generated maximum revenue in last 7 days, 30 days, etc.● What % of revenue is from recurring customers?

▸ Product● How are users using your product’s features? are people recommending? ● Has a new feature resulted in bad customer experience and a drop in

usage/revenue?

Type of questions

Page 12: Data analytics - Let's break it down

Analytics Workflow

1. Identify the problem

2. Obtain the data

3. Understand the Data

4. Prepare the Data

5. Analyze the Data

6. Present the Results

Page 13: Data analytics - Let's break it down

Data Transformation

TransactionalData

AggregatedData

Page 14: Data analytics - Let's break it down

Tools for Data Analytics▸ Excel / Google Spreadsheet▸ Database - SQL ▸ R ▸ Python▸ ETL Tools - Extract, Transform, and Load▸ Data Visualization/Dashboards

● Powerpoint/Excel● Industry-specific dashboard (Healthcare, E-commerce, etc.)● Role-specific dashboard (Marketing, Finance, Sales, etc.)● Tableau ● GraphiQ https://www.graphiq.com/ , D3.Js● Create your own Dashboard

Page 15: Data analytics - Let's break it down

Data Types▸ Categorical (also Qualitative)

●Categorical variables represent types of data which may be divided into groups. Ex: race, sex, age group, and educational level

▸ Numerical (also Quantitative)

●Values of a quantitative variable can be ordered and measured. Ex: age, height, sales, volume

●Numbers are not always numerical data. Ex: Gender (0=Male, 1=Female)

Page 16: Data analytics - Let's break it down

Typical challenges▸ Data is stored in too many places▸ Stored in different formats.

●How many ways can you use store date?

▸ Requires engineering effort to pull or transform data

▸ Quality of data is not good▸ Data is there but need to jump hoops to get

access ▸ Delay in answering questions▸ How to interpret data

Page 17: Data analytics - Let's break it down

Source: http://mbtaviz.github.io/

Page 18: Data analytics - Let's break it down

Source: http://mwpdigitalmedia.com/blog/without-a-video-your-kickstarter-project-will-probably-fail/

Page 19: Data analytics - Let's break it down

Resources - datasets, competitions▸ Datasets

● City of Boston https://data.cityofboston.gov/ ● https://www.quora.com/Where-can-I-find-large-datasets-open-to-the-public● https://github.com/thearpitgupta/data_science_resources#data-sets ● http://www.gapminder.org/ ●Your own data: Uber, Runkeeper, Mint, Fitbit, Social media,

sleep, etc.● https://www.facebook.com/help/405183566203254 ● https://www.linkedin.com/settings/data-export-page ● https://riders.uber.com

▸ Data Competitions●Social Good http://www.drivendata.org/competitions/ ●Kaggle https://www.kaggle.com/ ●Baseball hack http://www.baseballhackday.com/ ●MIT Sloan Analytics Hackathon

Page 20: Data analytics - Let's break it down

$1M Grand Prize - Inactive

Source: http://www.netflixprize.com/

Page 21: Data analytics - Let's break it down

Resources - jobs, continued learning▸ Inspiration

●MBTA http://mbtaviz.github.io/●TED Talk

https://www.ted.com/talks/hans_rosling_shows_the_best_stats_you_ve_ever_seen

●Quantified Self http://quantifiedself.com/

●538 political blog http://fivethirtyeight.com/

●Crazy Egg http://blog.crazyegg.com/category/analytics/

● Ocam Razor http://www.kaushik.net/avinash/

▸ Learn - Codeacademy, W3schools▸ Jobs - Angel.co, Venturefizz, StartupJobsBos.com ▸ General Assembly - Data Analytics & Data

Science

Page 22: Data analytics - Let's break it down

What’s on your mind?

Twitter: @ArpitGupta

https://www.linkedin.com/in/TheArpitGupta