StatSocial: ‘We’re The Nielsen Of Social’

Michael Hussey StatSocialTwitter analytics and targeting platform StatSocial isn’t CEO Michael Hussey’s first trip to the customer data rodeo.

Back in 2006, Hussey founded PeekYou, a people-based search engine that aggregates publicly available information – Twitter handles, Facebook URLs, blogs, physical address, phone number, interests, schools attended – into customer profiles that are then available for purchase.

In other words, Hussey and his team were sitting on a data gold mine.

“That’s when we started thinking about the next application of the information in this database,” Hussey told AdExchanger. “So, in 2012, we launched PeekAnalytics, which focuses on the demographics available in Twitter Audiences.

"Let’s say a brand has 2 million followers. That’s great – but who are these people? The analytics tool allows us to search in our database for meaningful information like geolocation, age, affinities, education level, etc., and build models around that.”

This past summer, Hussey and his team launched version 2.0 of PeekAnalytics, also taking the opportunity to rebrand the platform as StatSocial. The new version includes “tens of thousands of new affinity segments,” Hussey said, as well as 370 interest categories based on the IAB standards for social audiences and segmented breakdowns of 8,200 consumer brands, 22,000 celebrities and influencers and 1,100 TV shows.

Users can create follower reports or pull tracking information for keywords, hashtags and URL mentions. The tool needs at least 300 followers or tweets before it’s possible to create any kind of robust reporting. All of the data in StatSocial’s database is verified against public information pulled from outside the Twittersphere and compared to baseline stats for context. The company is working on enabling users to segment information by state, Nielsen market and ethnicity, as per public census data.

“We have about 500 million Twitter profiles in our database, of which we’ve identified just over 10%, so about 50 million, as consumers,” Hussey said. “We’re only measuring consumers. We’re setting aside business accounts or what we think are spam or anonymous accounts. We’re focused on real people, and one way to do that is to find them outside of Twitter.”

Basically, StatSocial is like Twitter’s Tailored Audiences solution on speed.

“We can tell a client who they’re reaching within any defined audience,” Hussey said. “When the product started out it was about searching for people, but we quickly realized that we had so much information that we could bring to life in an ad targeting solution.”

Hussey spoke with AdExchanger.

AdExchanger: StatSocial sometimes refers to itself as the “Nielsen of social measurement.” How so?

MICHAEL HUSSEY: With social audiences, particularly on Twitter, it’s pretty easy to find out how many people are talking about you, and there are also a lot of tools around sentiment. But there aren’t a lot of ways to find out who these people really are. When we say we’re the Nielsen of social, we mean that we’re not just reporting on how many people are watching a specific TV show. We’re reporting on who these people are and their specific demographics. We’re a company that brings Twitter users to life through demographic and lifestyle segments.

Can you give me a bit more on the StatSocial backstory?

We started a people search engine called PeekYou, which now generates about 11 million visits a month, making it one of the top 400 sites in the US. It’s basically a sort of phone book with home address, phone number and social profiles organized to specific individuals. That was the core idea around PeekYou: How do you unify all of this public information and blog data that people tell the world about themselves into a single consumer record. And then, more importantly, what kind of applications can you build from that index of the web around people.

Tell me about the pivot toward advertising.

We’re developing Tailored Audiences on Twitter to deliver the best people for specific brands or direct-response campaigns based on the 30,000 attributes in our database. Basically, we’re helping companies that advertise on Twitter reach better audiences.

On the analytics side, we work mostly with publishers, like Mashable, for example. Publishers have all of this native sponsored content, and they need to be able to go back to sponsors to tell them who’s engaging with it. We’re able to prove that the audience that saw the content is the right audience. In the case of Mashable, that means people with a high net worth who are technology enthusiasts.

We also work with television networks that are promoting hashtags during TV shows. All of them also have listening platforms like Radian6 and Simply Measured, etc., but those don’t offer information about the “who,” as in, who these people are. We’re not a replacement for a listening platform, we’re a complement.

How does it work in relation with other measurement solutions?

One of our customers/partners is Salesforce.com, the Radian6 platform. We have a lightweight version of our demographic data and audience interest segments that’s currently an upsell for them. You go onto Radian6, set up a query and you can get age data, income, education data and more structured geolocation information. One of our intentions is to work as a complement to all standard listening platforms, either as a separate subscription or, potentially, baked into the platforms themselves.

What about CRM data?

We’re not set up to work with traditional CRM data right now. We don’t have the email addresses that are tied to the social profiles. We’re crawling the web sort of like Google does. We’re looking at the public-facing information, not logged in data, and we make matches based on that to other social profiles that belong to the same person. We’re taking any structured and unstructured data that’s around – work, school, brand affinity – and pulling it all together in the same record.

To link CRM data, we would need a partnership with a company like FullContact, which can take email addresses and match them to social profiles using an API. Integrations with companies like Fliptop or Salesforce.com, which has social.com, could also help us get better data. There are ways, but it hasn’t been what we do.

The key in our database that everything starts from is a person’s social URL, not their home address or email address. That’s what we use to find someone outside of Twitter; it all starts with an array of Twitter IDs. We put them together and start aggregating.

Is StatSocial more about tracking or targeting?

Both, really. But the ad targeting is all ad hoc. It’s part of a managed service right now. We’re in the testing phase working directly with Twitter clients and API clients. We’re running as many campaigns as we can get our hands and building what I would call a Facebook style audience builder for Twitter Audiences. If you want to reach a specific demographic or a celebrity’s followers, we will allow you to define your audience in a more granular way than with a standard audience builder.

Do you work with brands?

We do, but I can’t disclose most of the names, although I can say that we recently signed on US Soccer as a client. We also work with a number of financial and insurance companies that are doing Twitter outreach, as well as various retailers and a number of agencies, including 10 different agencies housed with IPG, WPP and Omnicom. On the network side, we’re all project-based.

Do you have an official partnership with Twitter?

We’re not a certified Twitter partner, but we are a customer of their partner DataSift, and that’s where we get a lot of the Twitter data we analyze. But Twitter is well aware of who we are and there was a robust approval process to access the firehose and Twitter’s historical data. We’re also in regular contact with a number of people at Twitter both on the analytics side and on the targeting side. Right now it’s an informal relationship, but there is definitely a degree of awareness at different levels.

Is there any danger in building your tech on top of another company’s technology? What if the company tanks?

There’s no question about it. But the long-term solution for any company leveraging a third party, like a social network, is to bring value to those networks. In our case, if we’re driving significant amounts of advertising spend because of our analytics and targeting solution, that’s a good thing.

But while it’s true that we’re focused completely on Twitter today, we’re structured in a way that would allow us to build an analytics and targeting solution for any social network, be it Instagram, Tumblr, YouTube or whatever. We can match across 60 different social networks to create identities. Not that we think Twitter is going anywhere, but, in theory, StatSocial could easily recreate the same solution for other networks.

 

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>