Twitter No More

This isn’t what you think. I’m not closing my Twitter account. No, this post is addressing something that happened over half a decade ago. In May 2008, I began a side project. While running SoberCircle, I built a Twitter aggregation service that worked solely upon local Twitter users and their networks of followers. The service was called NashMash and indexed the activity of over 30,000 Twitter users in the Nashville market.

Why did I build it? I wanted to play with the Twitter API while growing my local network. I had met Dave Delaney (for those of you who don’t know Dave, check out his book New Business Networking) at SXSW in March 2008 and it was Dave’s strong presence on Twitter that encouraged me to look at Twitter as a means to grow my local connections. Most of my work at that time was with people outside of Nashville, so I didn’t have a strong network in the community at the time.

So off I went. For the next six months, I spent some nights and weekends working on this project. Granted, when you’re running a social network, your nights and weekends are fairly scarce. By October of 2008, I launched the service. For myself, I allowed the service to rank and following the top rated Twitter users in the market. This was great, because I instantly had created new connections that previously did not exist. Having come from building an social network that was an inclusive environment, I quickly learned that the inclusiveness that people wanted in the social network, was not what people wanted when it came to open messaging services like Twitter.

Practically overnight, people were using the service and finding out that it could rapidly grow your followers, but this came at a price. Eventually, I realized that the mass following service I had built wasn’t as desirable as it would seem. It was abusive in some regards, but its real value was now the index that quite accurately ranked users based on their activity within the community.

The ranking algorithm took into account a users location, their followers locations, their retweeters locations and the location of those who mentioned them. What resulted was a fascinating data tracking service that could tell you not only who was most popular in a market, but it also was doing a better job of accurately predicting local audiences that users should follow that Twitter’s own recommendation engine could do at the time.

By late 2009, I realized that regardless of how cool the service was, it was adding much more value to another company’s offering than it would ever create on its own and I decided to sell it. By this point, seeds were being planted for the transformation of Twitter’s Developer API from a service open to all developers, to one that was built for strategic alliances that were beneficial to Twitter. As a company, Twitter reserves the right to do whatever they want with their service and I felt that the writing was on the wall.

It’s easy to create value when you’re a plugin to a much greater service, but you do so knowing that at anytime you could be cut off and the value you created is suddenly greatly diminished. That’s not where I wanted to be, so despite many people in the Nashville community encouraging me to build out NashMash, I took the off ramp.

Coincidentally, I took the off ramp out of Nashville completely in 2010 and relocated to Denver. People asked me at the time if I would start a Twitter aggregator in my new city, and my answer was always no. I’d reached a point where I’d rather create value than augment an existing product.

I know several people attempted similar Nashville related aggregation sites since NashMash. I am glad that nobody has made a serious run at it because as cool as it might be, there is very low value in it as a company.


Now read this

Why B2B Startups are Easy or Hard

I appreciate those who work on hard problems. Scratch that. I admire them. Too often though, I see startups that sell directly to businesses building upon concepts that are unnecessarily hard. In a business you have cost centers and... Continue →