So… What is OneSpot… Exactly?
Thursday, February 5th, 2009by Mason Hale, CTO
For a city with a metro-area population of over 1 million, Austin is still very much a small town. If you’ve lived and worked here for a decade or more, it seems anywhere you go you are bound to meet someone who either knows you or knows someone who knows you. This is especially true in Austin’s technology industry sub-culture, where many people have at one time or another worked at one of the small handful of bedrock companies that serve as the hubs in Austin’s many-spoked high-tech ecosystem.
“OneSpot is fundamentally a recommendation engine for web content. If you’ve ever shopped on Amazon.com or browsed movies on Netflix you’ve seen something similar.”
And so, whether I’m grabbing a coffee at Genuine Joe’s or breakfast at Taco Deli or taking in a movie at the Alamo Drafthouse, it is not uncommon at all to run into someone I know but haven’t seen in years. The conversation invariably turns from “What are you up to these days?” to “So… what is OneSpot… exactly?”
In these situations a mouthful of marketing mumbo-jumbo might suffice but it certainly does not satisfy. These situations require a more precise, direct and unvarnished answer. No drone-speak thank you, just give me the straight skinny. Why is it cool? Why should I care?
My answer is that OneSpot is fundamentally a recommendation engine for web content. If you’ve ever shopped on Amazon.com or browsed movies on Netflix you’ve seen something similar. People who bought the item you just added to your cart, also bought these other items. Given that you rated this movie 5 stars, we think you’ll enjoy these other movies. You get the idea. The fundamental concept is the same, it sometimes goes by the name “collaborative filtering.”
“Thousands and thousands of new web pages are published every hour, and it’s our job to sift through that torrent of content and to quickly, efficiently and intelligently mine those rare gems that our customers and their audiences will find interesting.”
The difference in OneSpot’s case is scale. Now, with all due respect, Amazon and Netflix are both dealing with huge volumes of data. But where Netflix is recommending movies and Amazon is recommending products, OneSpot is recommending web pages, and there are many, many times more pages on the web than there are movies on Netflix or products in Amazon’s expansive catalog. Thousands and thousands of new web pages are published every hour, and it’s our job to sift through that torrent of content and to quickly, efficiently and intelligently mine those rare gems that our customers and their audiences will find interesting.
Of course, a recommendation engine alone does not make a business, and that’s not the whole story of OneSpot. Stay tuned for more on that topic. But seen through my CTO prism and targeted to the typically technically-inclined person I’m likely to be chatting with, this answer succinctly describes the technology running at the core of OneSpot, why it’s challenging, and why it’s cool.




February 9th, 2009 at 8:25 pm
what if users dont rate the content correctly? non-skilled users rate content well non skilled - correct?
is Martin Scorsese going to rate a movie the same as alice?
bobm
February 9th, 2009 at 10:34 pm
@bob - thanks for the comment.
You are absolutely correct. Some people are vastly better at selecting good content relative to others. This is why all content-raters cannot be weighted equally.
To clarify, the comparison to Netflix and Amazon.com recommendation engines is a bit of a simplification. I don’t know whether Netflix and Amazon weight all raters equally in their systems, but I can say emphatically that OneSpot does not.
I’m working on a follow-up post to round out the picture of what OneSpot does and how it works at a high level.
Suffice to say, we spend a lot of effort trying to find and identify the Martin Scorsese’s of the world and to pay attention to what they have to say and recommend.