clrkwng / kickstarter-success Goto Github PK

1. Deadline/Duration vs. SF (DL)
2. Staff pick vs. SF (Sean)
3. Goal (amount) vs. SF (Sean)
4. Spotlight vs. SF (Sean)
5. Category vs. SF (C)
6. Category vs. Location (Center of Category) (DL)
7. Map
8. Probability of Success in center vs P(success) out of center (DL)
9. Length of title vs. SF (FM)
10. Blurb Length vs. SF (FM)
11. Total pledged vs. category (S)
12. Avg. pledged by individual (pledge amount/backers count) vs. category (S)

• Finish Intermediate Python Programming (DataCamp)

• Python Data Science Toolbox Part 1 (DataCamp)

• Python Data Science Toolbox Part 2 (DataCamp)

• Comes up with suggestions for focus question for our project

• Complete assigned analysis with your partner (upload to GitHub)

• Description for your aspects done by next Monday's meeting (Nov. 26th)

• PowerPoint done for next Monday's meeting (Nov. 26th)

• Jupyter Notebook Report done next Monday's meeting (Nov. 26th)

• Which factors are correlated with Kickstarter success?

  • Compare about 10 features

• Graphs to show correlation (or no correlation) of:

  • Goal amount vs. success rate
  • Canceled, suspended = failure
  • Can elaborate on what this implies
  • Funding rate vs. success rate
  • Backer_count vs. success rate
  • [Duration of campaign vs success rate]
  • Title characters

• Research more data to determine what the focus question should be

• Suggestions and ideas for the focus question

• Learn how to use GitHub (if needed)

• Email Alex

Presentation
Background (objective of kickstarter) (40 min)
What is kickstarter?
How to define success?
Goal is to have project backed
What type of concerns do creators have regarding the project? (starting/backing)
Motivation (20 sec)
To analyze what decisions should creators make?
Some can be controlled, others are situational
Location: you have the network
Findings (40 sec per analysis) 120 - 160 words
Aspects of Projects (cannot control)
Distance from the center vs. SFR (does location matter?)
Staff pick/spotlight (does advertising on site help?)
Category (Are certain ones more competitive populated than others?)
SFT
Total pledged amount
Avg. pledge amount
Decisions to make (can control)
Title length
Blurb length
Project duration
Goal amount
Further Areas of Research (30 sec)
Collect data on network/friends
Does product get made
Follow up on products?
Predictive model on the success rate

Goals by the end of semester

• Predictive Model/Tool
• Enters features/factors
• Rough % or range (success rate?)
• What you can improve
• Report
• Data Visualization
• Algorithm/Model
• Methods of Manipulation
• Analysis
• Background
• Source of Errors

Factors to consider

• Funding goal
• Funding start
• Fund length
• Category
• Founder gender/age
• Video (y/n)
• Quality of video
• Social media
• News saturation
• Location
• Purpose
• Reward levels
• Quality of page
• Thumbnail
• Featured (Y/N)
• Founder fame / Reputation(well-known or not?) based on social media
• Launch time
• Campaign growth
• Successful in funding or not, % of goal
• Even if successful, did they deliver?
• “Uniqueness”
• Previous experience of founder
• Update frequency
• Early bird
• Locked (reach certain funding goal)
• Product/brand/campaign partnership

Scope for our own project

• Category/sub
• Time (2015 onwards)
• Factors

Focus Question

• What influences people to invest in a kickstarter campaign?
• What makes a kickstarter campaign successful?

• Data Camp Modules

• Try things out with one data set from the September 10th file for Web Robot's Kickstarter Data (Use Jupyter)