Nov
17
posted at: 3:03 PM
Today is an exciting day.
It's really wild to think it really all started with some research I was doing one snowy day in late 2008 on automated sentiment analysis. I didn't even know that was the technical term for it back then, and there was no way I could have anticipated it would be considered one of the hottest tech trends of 2009.
Not long later, I took my research and a prototype of the technology and started showing it to people at work. I hooked it up to the Twitter API and made a compelling UI demonstrating a stream, and how you could view how emotional perception for topics could change in real-time. twendz was born, and once we released it as a free product, people started using it every day to track the emotional reaction to their product, brand and searches that interested them.
Today it is used every day by people in Fortune 50 companies and Ivy League Universities. It is still considered amongst the best applications of it's kind by top publications like The New York Times, and high-profile blogs like WebWorker Daily and Mashable. It even has imitators. But as we know, competition drives innovation, so after launch I went back to work.
Inspired by concepts and conversations shared with coworkers, the emerging trends of the social media industry, and Waggener Edstrom's focus on the nature of influence (we have an internal innovation initiative which helped me get motivated), I began mining Twitter user data on a massive scale. Millions and millions of users.
I created a basic system. People were already leading in the space of understanding and measuring influence on a detailed level on Twitter. I wanted to be able to identify a magnitude of influence for as many people as possible on Twitter. Comparing and ranking important people instills doubt and dissent, creating a High-School psychological complex. There was more value in being able to view a conversation and see who were the most important people spreading a message.
In order to do that, I knew I would need to be able to look at an entire conversation and tell you who was influential and who wasn't. Who are the most important people contributing, and are they sending a positive message or a negative one? I needed to know as close to everyone in a conversation as I could.
Once I had enough influencers to make this effective, my coworker Marc and I put our heads together and hashed out ideas to assess the impact of the message. We took inspiration from the work others were doing, and came up with many of our own ideas to create some sketches of what we considered a dream Twitter analytics tool.
Well, I went to work after that, spending my nights and weekends for a month until I spit out an alpha of what I was calling Project Seagull. Our team kept it very quiet internally, gradually refining it, getting feedback, until before we knew it we were teaming up with our Measurement team to refine our formulas, and our Marketing team to polish the app up and make a new product.
Before long, we got to today. Today WE are launching twendz pro, reinforcing our position as one of the leaders in full-scale Twitter analytics.
I'll spare you the, "What it does," because you can see that here. Or you can jump right in and take it for a spin yourself.
I'm incredibly excited about this product, as it has been a ton of firsts for me, and I'm looking forward to working with our development team to continue to bring more world-class social media products in the months to come.