A very early version of Horsebot3k’s website showing off a “web-scale” version of the team’s mascot. Photo by Danny Schreiber.
Horsebot3k, an Amazon review analyzer, took home the top prize at the fourth Startup Weekend Kansas City, a 54-hour event in which more than 100 participants formed teams and attempted to take an idea from concept stage to a prototype with a business model and then pitch it on Sunday evening. Bold Predictions, an app that timestamps users’ predictions (think sports scores) to gain points and share with friends, placed second. Dawg Bnb, the Airbnb for dog sitting, and Merry Plate, an app that encourages users to track and adjust their meals for a healthier diet, tied for third.
I had the privilege of covering the event and serving on the panel of five judges. Kudos to organizer Adam Coomes and Steven Chau for assembling four fantastic judges (and including me) as well as leading the charge to put on the event. (Stay tuned for a video interview with the two later this week.)
The teams were judged on three equally-weighed criteria: business model, customer validation and execution. Speaking as one of the judges, Horsebot3k came out on top because of their initial business model around Amazon’s referral program (although small, the team saw a couple dollars in revenue by Sunday), its potential to capitalize on an established and growing userbase (the team reported Amazon’s userbase to be 81 million regular users) and its execution during the weekend (the team had a working prototype to demo in its pitch).
If Horsebot3k can gain traction by proving the core of its concept – analyzing user-generated content and presenting the most “intelligent” pieces in a value-added manner – it could expand its offering to other review sites as well as comment sections. Think of a Huffington Post article with more than 1,000 comments or a YouTube video with more than 100,000. These comment sections contain nuggets of insightful and interesting information, it’s just a matter of extracting them in a quick and user-friendly manner.
The Horsebot3k team was made up of four Kansas Citians: brothers Justin and Jeff Graves, far left and third from left, Josh Bohde, second from left, and Aaron Marz, far right. Photo by Danny Schreiber.
“I for example, I use Amazon a lot. Actually, quite a lot. Like, seriously a lot. I’m actually credible for three of Jeff Bezos‘ monthly car payments.”
Justin said that user ratings and reviews are helpful, but when viewing a product with a high number of reviews, the effectiveness of the reviews begins to decline.
“There’s sometimes so many of them it’s hard to decide which reviews are good, which are bad, what are people actually liking and disliking about a product.”
Justin’s explained his team’s solution: “Horsebot3k analyzes reviews on Amazon using natural language processing – it’s fancy stuff,” Graves said. “We rate the intelligence level of people who write reviews to gauge how good they are and use this to help you in your decision making.”
He summed it up: “We help tell you what sucks and what does not, intelligently.”
Here’s a look at how Horsebot3k works:
1. Visit horsebot3k.com
2. Drag the bookmarklet to your browser’s bookmarks bar
3. Find a product on Amazon you’re thinking about buying and click your Horsebot3k bookmarklet
4. Watch your browser redirect to Horsebot3k’s website and show you an animated .gif of a horse running while it runs its algorithm
5. Get the results – the grade represents the average “intelligence” of the user reviews for each rating level – and if you’re compelled to purchase the product hit the “Buy this on Amazon!” button
6. Watch your browser redirect you to Amazon, where if you look closely you’ll see Horsebot3k in the URL, meaning that if the product is purchased Horsebot3k will recieve up to 15% in referrals
Disclosure: Danny Schreiber, the author of this post, both covered and judged Startup Weekend Kansas City.