Ping, ping, ping! That’s the sound made day and night by the new social media technologies rapidly proliferating around the web… and the machines are getting tired. Polling for updates to user data streams, wishing they spoke the same language and dreaming they knew which accounts belonged to the same people across different services.
Sounds like a great opportunity for an infrastructure provider, doesn’t it? Enter the sexiest infrastructure provider we’ve seen in a long time: Gnip. Venture funded and built by exited MyBlogLog co-founder Eric Marcoullier, Gnip wants to serve as the grand central station and universal translation service for the new social web.
What Gnip Does Now
The primary service that Gnip offers at launch today is to capture user data updates from any web application and then serve up the very latest information to anyone else who requests it. Your application doesn’t have to ping Flickr, YouTube, etc. etc. every few minutes and ask “have any of our users done anything on your individual service?” Now with Gnip, Flickr (a launch partner in fact) can report user data updates to Gnip, which can then pass that data along to consuming parties, along with data from all the other social media services of interest.
It’s about scalability and decreasing latency to near zero. It sounds like a great idea.
What Gnip Says it Will Do Within 90 Days
The sexiest features are still in the works. Here are just a few plans that the company has disclosed so far.
Protocol switching You want to communicate instantly with an application using the IM protocol XMPP/Jabber but it only publishes and consumes RSS feeds? Gnip will stand in the middle and translate so each end of the transaction can send and receive data in the format it wants. Hot!
Standardized metadata Different services publish user data with different titles for the various fields sometimes. That makes it hard for the robots to know what exactly is being said. Gnip is working with visionary developer Chris Messina to create a meta data standardization process, for social bookmarking activities in particular. Gnip will consume feeds from user bookmarks on any service, then make publicly available those same feeds appended with another version of the same data but in a standardized format. That’s a really big deal because it makes interoperability possible.
Identity discovery Right now it’s hard for services and for users to tie multiple social media accounts together from around the web. Gnip will let users provide usernames or emails and then check to see where else those identifiers are being used. That’s a solid sounding idea.
All of this will be free, none of it will be public-facing. Application developers will tie into Gnip and there may be premium services available eventually, like translation of data into a particular vendor’s proprietary template.
Is This Too Centralized?
Gnip’s Eric Marcoullier acknowledges that the centralization here is worth questioning. Two primary concerns come up: scalability and privacy.
As far as scalability is concerned, that’s the name of the game here. If Gnip can’t scale with fantastic uptime then there’s no service. The company has been working with Pivotal Labs, the strike force hired to fix Twitter, since Gnip started work.
When we asked Marcoullier about privacy, he emphasized that Gnip is only working with publicly available data right now. The company might venture into user authentication and private data, but that’s “a whole other can of worms,” he said.
No doubt Google is already indexing most of the information that Gnip will be transmitting, but we can’t help but think that Gnip will be in a uniquely powerful position to do some mining of anonymized user data and social graphs. That could lead to very big money but it could also raise some concerns on the parts of users. Gnip says they aren’t worried about monetization right now, they just want to build out their service and add value to the market place.
That may be the case, but we think that Gnip has a whole lot of potential to deliver huge value to the applications leveraging it, to the backers financing it and ultimately to all the users of the emerging class of social web applications. We love this kind of stuff.