Kickstarter_1 is an example of taking a large data set and extracting trends based on certain characteristics, such as success based on funding goals.
What are three conclusions we can make about Kickstarter campaigns given the provided data?
- Kickstarters of different categories were successful based on different variables, such as country (location), whether they were spotlighted, and the number of sponsors.
- From the general trendlines, the success of the kickstarter seemed to weakly and inversely correlate to the goal for lower range amounts. In contrast, for higher goal rates, the percentage successful dropped.
- More kickstarters were successful around the month of May for most years.
What are some of the limitations of this dataset? One limitation is We do not know what drove the staff picks and spotlights, or the definition of some of the data set labels. Spotlight seems so strongly to correlate with kickstarter success that it could be caused by kickstarter success rather than the cause of success.
What are some other possible tables/graphs that we could create? We can display the relationship between staff picks, spotlight, and success. We can also display the success of certain parent and subcategories by international regions rather than individual countries.
Green highlighted were more successful fundraising groups.
Red highlighted failed.
We can drill down on the failed ones to look for characteristics to avoid in future endeavours.
In this data set, analyzed with Excel (version 2019): conditional formatting, data sort, pivot tables, and pivot charts.