Clearspring is all about data. Big data. Lots of data. Data coming out of our ears data. Clearspring processes 10 terabytes of data per day. The AddThis button is on over nine million websites and has a billion unique users worldwide. Data is generated by users of the platform sharing content across social and news networks as well as when people visit sits that AddThis is embedded on. Clearspring will use this round of funding to accelerate its next-generation publishing products and continue growth of its advertising offerings.

"People are pretty excited. It makes you dizzy," Hooman Radfar, CEO and co-founder of Clearspring, said to us. "You think, I have the entire Internet measured and I can have it in real-time. You look at comScore, for instance, every month they give you a flash report. I can give it to you on the minute of who is on your site. Right now, across every site. What could you do with that?"

This is Internet data done right. AddThis gets data from everywhere, all the time. Clearspring can make this data available to publishers with high-end analytics to be able to take action on what users are reading and sharing on their sites and optimize the Web experience to the user.

"We can now bring the power of that reach back to the individual publisher via our next generation tools and services, which will focus on providing publishers actionable data, not just from the social web, but the entire web," said Radfar. "Imagine the challenges we can help a publisher solve- -it's a staggering opportunity."

What To Do With All That Data?

Clearspring has a few primary focuses. Foremost is the publisher platform that shares content around the web as well as producing data for Clearspring content creators. Second is the data analytics. Essentially, Clearspring creates an entire Library of Congress worth of data every week (around 70 terabytes). They are building data centers in Ashburn, Va. and Los Angeles. Third is the social media optimization and advertisers network. The biggest part for Clearspring is that all of the data analysis in done in real-time. They showed me a screen board in their offices in McLean, Va. with screen tracking the Internet that was counting how many shares and views and the like were coming across on a huge portion of the Internet. It is truly mind blowing how much data Clearspring can process.

"That is what everyone is excited about, that next generation of tools," Radfar said. "It is cool that we have always had that vision around data and analytics and being this big data company and now it is like everyone gets big data and it is all about big data because the timing was really good."

The hope for the publishing industry is that firms like Clearspring and other will be able to provide enough rich data that advertisers will up the ante on rates for digital content as opposed to print content. The amount of data that Clearspring brings to the table should go a long way in helping bridge that divide.

"Just imagine, you take the power of that entire network, a billion uniques and nine million sites and run it through that processing," Radfar said. "What are all the products that we can offer a publisher and bring that back to an individual."

That's a whole lot of information. ClearSpring sees not just what you're sharing, but nearly everything you're doing on the pages its products are on. (AddThis is on ReadWriteWeb, for example.) So what on earth is it going to do with all that data? Like they said in Spiderman, "with great power, comes great responsibility." We asked ClearSpring's CEO about these super hero-like responsibilities and his thoughts are below. You can decide for yourself whether he can be trusted, but the work the company is doing is very cool.

In February we wrote about ClearSpring competitor ShareThis and its venture backed plans to become a major development platform with all the data it sees by facilitating content sharing. What are ClearSpring's development plans and how will it treat user privacy? We asked ClearSpring CEO, Hooman Radfar, and this is what he said.

To summarize: Radfar says ClearSpring is analyzing the content on all the web pages it sees, as well as user intent in sharing content, and categorizing that information. And will soon make APIs available that allow developers to build trend analysis and recommendation services on top of at least some of the data. The company is also taking privacy seriously, Radfar says. As nerds about aggregate data analysis and APIs, we think this all hits the sweet spot. It's kind of creepy too, though.

Here's our interview.

ReadWriteWeb: So, you see half the internet now. What's that old phrase? "With great power comes great responsibility?"

Hooman: Yea - weird right? It's honestly a bit surreal.

ReadWriteWeb: What are you going to do with all that info?

Hooman: Well, there are two parts to leveraging our reach. First, now that we have so many folks using AddThis, we want to bring other service providers into the mix. The next twitter, etc. [Support for more sharing services.]

The next part - which you alluded to - is USING the data. Using the data is a longer term project. We are starting to classify it into categories and apply it to problems to understand how it can enhance the AddThis experience for users via new products, or enhancements to existing products.

ReadWriteWeb: What kind of categories?

Hooman: So we are looking at intent-based information, as well as behavioral. We profile users of the AddThis/clearspring platforms and attribute intent or behavioral info to the profiles.
Within each of those categories, we are contextualizing them vertically. So, for example, within intent, we are looking at subsections like auto, shopping, etc. It's an insane amount of information.

Here are some use cases. First, we have a social distribution business where we distribute branded widgets for advertisers. We can leverage the data to optimize distribution by targeting users with a higher proclivity for sharing and the content type.
That's a low-hanging fruit opportunity.

Some of the other opportunities that we are planning on pursuing are more product focused. For instance, we see a large sample of shares into Facebook and Twitter, in real-time. Every second we are seeing shares.

We have built a platform to expose that information and built recommendation applications a la Digg, that show real-time recommendations based on user activities. The idea is that we could leverage the data to launch user applications that will enhance the 'sharing' ecosystem by providing more recommendations, we are helping publishers get more traffic and users get better content.

This summer we plan to expose our internal platform and capabilites. To put it in perspective we see orders of magnitude more than some of the most popular recommendation sites.

ReadWriteWeb: What will that look like? Will it be limited to data concerning sharing activities?

Hooman: Still working that out to be honest. We probably will expose more - but we have to be careful, per your spiderman quote. We are a small company with a whole lot of info. To put it in context, if we were a publisher only Google, Yahoo and Microsoft have more reach - and we are getting closer to Microsoft. We keep it simple - as long as we can sleep at night then we are probably on the right track. We have already started talking to privacy experts.

So products we have built on our data internally include trends (think google), recommendations (think digg).

ReadWriteWeb: Can I get a raw data dump to play Hadoop or whatever with?

Hooman: We have not done anything like that yet. We want to be sensitive to Personally Identifiable Information. That being said, we are going to expose rather large data sets soon. We want to be sensitive, but we can see trending topics a la twitter.

ReadWriteWeb: What large data sets will you expose?

Hooman: Sets we are looking at are...trends across the ecosystem from a topic perspective, so you can find if a 'term' is rising and falling. Exposing the actual sharing data so someone could create a 'shares around the world' mashup with google maps. And we might do some categorization and have an API to let people actually see trends within categories like auto, etc.

ReadWriteWeb: Are you analyzing the content on the pages being shared? Or just the messages that get posted along with the sharing?

Hooman: Both. That's where it gets powerful.

ReadWriteWeb: How are you analyzing the content?

Hooman: Contextualization. We have access to pages via javascript so we can categorize, not unlike Lijit or Google. We are partnering to do that.

ReadWriteWeb: With whom?

Hooman: Can't say yet.

ReadWriteWeb: Oh come on now!

Hooman: I wish I could. I don't know if I am allowed.

ReadWriteWeb: If you don't know if you're allowed then you have not been told to stay quiet about it. That's a key detail.

Hooman: Ha ha ha. I know.