Performing analysis on Kickstarter (crowdfunding) data to uncover trends

The purpose of this analysis is to determine how crowdfunding campaigns performed based off of their campaign launch date as well as their campaign fundraising goals. We are interested in looking at campaigns categorized specifically as theater and/or play due to our stakeholder's interest.

In order to understand outcomes based on launch date I created a pivot table with the filters on "Parent Category" and "Years". The pivot table fields consisted of "Outcomes" in the columns, "Date Created Conversion" in the rows, and a count of the Outcomes as the values. I also filtered the column labels to show the campaigns that were "successful," "failed," and "canceled." The pivot table provided a count of campaigns per month per Outcome.

Note: the "Years" filter, was created by using the YEAR() function to extract the year from the “Date Created Conversion” column.

In order to analyze the Outcomes based on Goals, I calculated the number of successful, failed, and canceled US campaigns that were subcategorized as plays by using the COUNTIFS statement displayed above. The formula adapted to account for the respective outcomes and Goal ranges.

Next, I used the SUM() function to total the number of campaigns per each Goal range in each Outcome.

Then I calculated the percentages of successful, failed, and canceled projects by dividing "Number Successful," "Number Failed," and "Number Canceled" by the "Total Projects". I formatted the row to be Percents and finally, I visualized this data with a line chart.

A possible challenge that could have been encountered was specifying the goal ranges in the COUNTIFS statements for the analysis of Outcomes Based on Goals. Paying attention to detail was critical in getting the right numbers because a wrong or misplaced sign (=,<,>) could result in either an error or a wrong number. To address this challenge, it's important to double check the formulas and each condition within it.

• What are two conclusions you can draw about the Outcomes based on Launch Date?

The most successful theater campaigns (111) were launched in May so May would be an ideal time to launch. Also, May, June, July, August, and October had roughly the same number of failed theater campaigns (~50).

• What can you conclude about the Outcomes based on Goals?

From fundraising goals of $0-$29,999 the general trend of the successful campaigns is negative meaning that the the higher the fundraising goal is the lower the percentage of successful camppaigns. The highest percentage of successful campaigns were in the $0-$1000 campaign range.

• What are some limitations of this dataset?

Some limitations of the dataset are the size and the source of the data that match all of our parameters- there are only 671 US plays that Kickstarter has funded. We also don't have any information about fundraising campaigns on different platforms besides Kickstarter.

• What are some other possible tables and/or graphs that we could create?

I would suggest creating a table and/or line chart to view the trend between Number of Backers, Average Donations, and Goals. It would be interesting to answer the question: Do lower fundraising goals have fewer backers but higher average donations? I also would create a bar chart to display the average length of time campaigns ran based on Outcome.