About Analysis Overview Downloads Authors

About our project

Kickstarter is a funding platform which allows for creators to help bring their ideas to reality. It supports a wide variety of projects, many of which have enjoyed widespread commercial success, including the card game Exploding Kittens and the Pebble Smart Watch. Long before they were even commercially sold, both of these projects had campaigns not only met their funding goals, but far exceeded their stretch goals. This is the dream for any entrepreneur who wants to use Kickstarter, but it isn’t easily attainable. In the shadow of every high profile project, there are thousands of utter failures that don’t even make their funding goals. For some projects like The Dual Shower Heads for Two, or the "completely true" Animated Adventures of Samurai Mary vs. Ninja Jews (The Birthing of Christ) it can seem obvious as to why some of these ridiculous projects didn’t get any funding. But often it isn’t as cut and dry. For example, an annual calendar that helps support cat animal rescues didn’t make its funding, while a project to help someone make a potato salad for themselves made over $55,000. This begs the question, what makes a successful Kickstarter? It has to be more than just having a killer idea at the right moment. For our project, Kickstats, we wanted to uncover any factors of success behind Kickstarter projects and see if there are any specific measures a project can take to ensure its goals can be met.

Overview of our Analysis

While most of our work is described in our paper, we have a page with an overview of the higher-level parts of our analysis, and some of the visualizations we produced.

Go to overview

Try your Kickstarter idea!

Got an idea for a project? We created a linear model based on the dollar amount, category, and whether your project is a staff pick or not. You can enter your hypothetical project in the form below to see what our data says about projects with those properties.

The application was made with R Shiny. You can also view it directly on The source code for the model and Shiny app is available in the downloads section.

What is this?

This is our class project for CSC/STA 223: "Introduction to Data Science" at Cornell College, taught by Ann Cannon and Tony deLaubenfels in January 14 through February 6, 2019. In the first week of the class, each student looked at the web and other resources for data collections, and made a proposal to the class. Students voted for their preferred projects, and were divided into groups of four for the remainder of the class, working on their preferred project. The initial idea for this project was presented by Ben.

Data Sources

We obtained our original data from a post on Kaggle by Mickaël Mouillé. We got some additional data for many of the records in that dataset by scraping Kickstarter's advanced search API as described in our report.