Using Kickstarter data to predict Ubuntu Edge’s success

August 1, 2013
By

(This article was first published on Open Analytics - Blog, and kindly contributed to R-bloggers)

Earlier this week we published an article with an analysis of the Ubuntu Edge Indiegogo crowdfunding campaign, and the outlook was bleak. Donations had slowed to a crawl, and Canonical's $32 million goal appeared to be little more than a pipe dream.

Yes, readers replied, donations had died off, but that's par the course for crowdfunding efforts.

Some digging revealed a Kickstarter blog post documenting this trend.

According to the article:

"When Craig examined his project, he noticed that it dragged during the middle of the campaign. Craig called this period the “Dead Zone”; we call it the trough. When a project first launches, there’s an immediate spike of activity as the creator starts spreading the word. And as a project nears its deadline, there’s another spike as two things happen: 1. the creator and backers promote the project to make sure it’s funded or that their friends get in before the end, and 2. people who’ve been sitting on the sidelines (but meaning to pledge) finally jump in before the bell rings."

Armed with new data and information, we were able to explore some thrilling new hypotheses. Had the Ubuntu Edge campaign flatlined harder than the average crowdfunding effort in the “Dead Zone”? Given the end of campaign bump in donations, was there renewed hope in reaching the $32 million goal?

To start, we scraped Kickstarter's plot for data in order to compare the Edge's month-long campaign to the average 30-day Kickstarter.

Next, we transformed the rate chart (percent backings per day) to a plot showing the cumulative backing over time.

In this plot the diagonal line represents what the cumulative backing would look like if the rate of backings were constant over time.

Interesting to note, even with the “Dead Zone”, the cumulative backing mostly stays above the diagonal line, with a momentary dip below the linear trend about 80% of the way through the campaign. By contrast, the Ubuntu Edge dipped below the linear line to $32M after just a week, suggesting that their total backings will be below the $32M mark.

How far below $32M? To form an estimate, we combined the Kickstarter backing data with the Ubuntu Edge’s crowdfunding campaign data. The Kickstarter curve above is scale independent, meaning that its shape looks the same regardless of whether your campaign raises ten-thousand dollars or ten-million. All we had to do is find the end-total that creates a curve that most closely matches the Edge’s data.

Using a numerical optimizer, we found the total backing amount that minimized the squared error between the characteristic Kickstarter curve and the Ubuntu Edge data.

About $22 million.

Compared to the average Kickstarter, we can see that the Ubuntu Edge rose more sharply in the beginning and flatlined harder when entering the “Dead Zone”. This is further evidence that would-be backers are turned-off by Canonical’s incremental pricing structure. As the price of securing a phone has ticked from $600 to the present value of $775, backing has slowed to a snail’s pace.

Further, the $22 million estimate may be generous, as the raw data currently sits below the Kickstarter projection line.

As a last look at the data, we fit a Kickstarter curve that intersects with our most recent data point.

This projection puts the Ubuntu Edge campaign on pace for a $18 million total.

Even when incorporating Kickstarter’s trend of an end-of-campaign bump, the Edge is projected to fall well short of its goal.

Jason Waddell (@jwaddell_) and Willem Ligtenberg (@wligtenberg)

This post is about:
Edge, kickstarter, r, Ubuntu

To leave a comment for the author, please follow the link and comment on his blog: Open Analytics - Blog.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...



If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Comments are closed.