BackerTracker: Predictions & Trending Projects

Dan Goldenberg
BackerKit, Product Manager

TL; DR Today we released two products that will help crowdfunding project creators gain more exposure, prepare for managing their success, and perform research for future projects. We built a new crowdfunding projection model with help from our friends at Insight Data Science and a new trending page to help backers with project discovery. Check out BackerTracker here.

From Surveys to Shipped

Our mission at BackerKit is to help crowdfunded creators with the parts of the campaign fulfillment process that aren’t related to actually MAKING the thing people gave them money for. BackerKit helps creators do more of the things that they (and their backers) love like making mini museums and less of the tedious stuff like manually sending out tracking numbers.

BackerKit started largely out of the difficulties we experienced while fulfilling our very own record-setting crowdfunding campaign back in 2010. We realized then that getting funded was the easy part compared to the pain of fulfillment. Since we started in 2013, we’ve helped over 1,500 projects get through fulfillment and deliver something that their backers love.

One of the biggest challenges of working with creators is getting them to understand how complicated the post-campaign fulfillment process actually is. The #1 mistake we see creators make is underestimating the post-campaign challenge. That context switch from “Woohoo! I just raised a bunch of funding!” to “Oh man, I need to ship all this stuff!” doesn’t happen fast enough and failing to start on fulfillment early results in missed deadlines.

BackerTracker Projection Algorithm

That is why we created BackerTracker. We wanted to give creators a way to estimate their final funding amount raised sooner so that they can start thinking about how they can ramp up production and fulfillment even earlier.

We worked with Praveen Chowdary from Insight Data Science to help us build a crowdfunding prediction model. What’s cool is that the model uses the historical performance of past campaigns to predict the performance of a new campaign. Praveen explains that, “for any given campaign that is in progress, we find the past campaigns with a similar trend of funds using the nearest neighbor approach. The projected funding curve of the ongoing campaign is then estimated as a weighted average of these past campaigns.”

image

Each graph is the result of searching through thousands of past crowdfunding projects, finding the campaigns that have the most similar funding curve, and then averaging those campaigns to create a funding projection.

The projection adapts as we get more days of funding information which means that you’ll see an updated projection if a project gets a sudden influx of backers from a press mention or a new Facebook marketing campaign. In a way, it can help creators get early feedback on whether or not the activities they try out today will have an effect on their final funding amount.

Eva St. Clair, co-founder of Princess Awesome, explains that having projections “can change how the creators approach the media, their backers, and the scope of their project. It can help inform decisions such as what kind of stretch goals to set and how to achieve them, invaluable information that has the potential to push a project well past its original funding goal.”

Trending Projects

And we didn’t stop there. We are also releasing our BackerTracker trending page for project discovery. The trending page is not only a great way for crowdfunding fans to see what is new and cool across Indiegogo and Kickstarter (think of it as a front page for everything crowdfunding), but is also a great place for creators to start researching high-performing projects and study what they’re doing right.

image

We can’t wait to see how BackerTracker helps creators. If you have any feedback or great ideas, please let us know!