For those familiar with the Signal Patterns team, you’ll know we have a very diverse team of scientists, researchers and developers here. Most are recognized leaders in their respective disciplines but some are well known outside of science and academia and are famous with the general public.

I got the chance to see one of them last night discussing his new book…and then some. Best selling author (and Signal Patterns Science Team member) Dan Levitin was at the Lincoln Center Barnes & Noble discussing his latest book, The World in Six Songs: How the Musical Brain Created Human Nature. In it, Dan explores the evolution of music and the human brain over tens of thousands of years. And as the title suggests, Dan identifies six types of music that have developed over time and become a key part of our culture.

However, this wasn’t just your typical book store reading that you might catch on C-SPAN. Because in addition to leading the McGill University Laboratory for the Study of Music Cognition, Perception and Expertise, Dan is an accomplished musician, having played professionally with artists like Mel Tormé, Nancy Wilson, and members of the Steve Miller Band and Santana (as well as producing or consulting on albums by artists including Stevie Wonder, Steely Dan and Chris Isaak).

After discussing his book, Dan then played a well-received set with Nashville recording artist Rodney Crowell, highlighting how each of Rodney’s songs fell into one of his categories (like songs of knowledge or songs of comfort).

While pondering Dan’s “six songs” thesis, I kept thinking of our Music Patterns survey, Dan’s research on its underlying FUSES Model and how it all ties together. We’ve been doing a lot of work on this application based on all the feedback we’ve received and look forward to delivering some significant news on this front in the near term.

Just six songs? Guess I need a smaller iPod!

What’s your favorite style of music? Classical or Jazz? Rap or Dance? Salsa or Tango?

The music research we’ve been doing here at Signal Patterns indicates that people’s preferences for music genres tend to group together. In other words, people who really like a particular style of music, say classical, tend to also like certain other genres, like jazz and world music.

So far, it looks like preferences for a large assortment of music genres can be grouped into five music dimensions. Those dimensions spell the acronym FUSES:

• Forceful (described as loud, distorted, and energetic, and includes heavy metal, rock, alternative and punk),
• Urban (described as rhythmic, percussive, and funky, and includes rap, dance, techno and electronica),
• Sophisticated (described as intelligent and inventive, and includes classical, jazz, and world),
• Earthy (described as melancholic and real, and includes country, folk and bluegrass) and
• Smooth (described as relaxing and gentle, and includes R&B and soul).

One of the reasons we’re interested in people’s music preferences is because we believe (as we suspect do most of you) that the types of music people like reveals information about their personalities. To be sure, we’ve compared your responses to our Music Patterns preferences survey with your responses to our Personality Patterns survey. And the results are in…

Forceful music fans tend to be thrill-seeking, rebellious, haphazard, at ease, and self-indulgent.

Urban music fans tend to be outgoing, friendly, confident, warm, and generally happy.

Sophisticated music fans tend to be imaginative, reflective, creative, quick-witted, intellectual, and private.

Earthy music fans tend to be nice, understanding, honest, sympathetic, and nurturing.

Smooth music fans tend to be cooperative, friendly, conventional, rule following, hard working, and empathic.

So there you have it; there are connections between the styles of music people like and their personalities, many of which seem fairly intuitive and may be in line with any stereotypes you have about types of music fans. And that raises another interesting question: Do we listen to music that fits our personalities, or do we acquire personality traits that fit the music?

Personality assessment is based on the “study of individual differences“. But how different are people actually? Well - very different. Old school personality tests provide results in the form of a “type”: You may be a “gentlemen” or a “thoughtful leader”, or a “stressed out couch potato”. These old assessment instruments divide the human population into a small number of “buckets” and tells you what bucket you’re in. The Myers Briggs test, for example, uses 16 different possible personality types. That’s not very unique . Facebook currently has about 70 million active users and MySpace about 200 million. Assuming people are evenly distributed between the 16 types, you and almost 13 million MySpace users are the same type…

The scientific method Signal Patterns uses are “trait based” - they capture what characterizes people based on data collected from a large number of individuals. The Big Five personalty assessment framework captures the degree to which a person exhibits five main dimensions of personality. The Signal Patterns personality survey extends that level of detail to 45 traits, capturing subtle differences even between very similar individuals. It’s a long tail filter into the ocean of people that are online today.

Here is Signal Patterns scientist David Rosen’ take on this:

In my previous post, I identified three types of influencers that Malcom Gladwell discussed in The Tipping Point: Mavens, Connectors and Salesmen. Finding these alleged influencers has been the subject of many snake-oil salesmen and marketing companies since Gladwell’s book came out. Now, Pollara’s study has corroborated what some marketers have feared all along: that influencers aren’t necessary to reach millions of people, and that other factors, such as family suggestion and randomness, actually have far more of an impact on the propagation of trends.

These approaches are not necessarily irreconcilable. There is great variability in how we treat the opinions of family and friends. If I were interested in the best poker website, I’d ask Justin, who can tell me about the competition, the rake, and the interface of any site out there. He’s a Maven when it comes to online poker, and he’s also a friend that I trust—the latter doesn’t disqualify him from being the former. Another Maven might have a popular poker blog, but I wouldn’t trust it if I wanted the best information. So pitting “friends and family” against “Mavens or popular sources of information” sets up a false dichotomy. Pliny the Younger (pictured) was sought out by family members and strangers alike, but among his family members, he had a large influence.

Connectors, too, can be found among friends and family—in fact, if anyone’s likely to be counted among friends, it’s a connector. Finding that friends and family are more influential in purchasing decisions still doesn’t help delineate which friends and family have the most impact. Since I’d rather get advice from someone in my social network, I’d go to the Connector in my network if no one in my immediate circle could be trusted as an expert.

The same line of reasoning applies to salesmen, although the process would be less of a conscious one. Who are we persuaded by among our family members? The outgoing, enthusiastic ones who love what they do—Gladwell’s salesmen.

To sum up, it may be important to find Mavens, Connectors and Salesmen because they spread influence among their social groups, not because they have massive clout in the flat, online world of social networks. They have friends and families too, and may impact those groups accordingly. The influence that people have may vary in both quantity and quality, and both of those facets are important; imposing distinctions between friends and family as compared to wide, non-personal influences manages to ignore distinctions between the two facets. It’s certainly possible that online “influencers” have a greater quantity of influence (in the sheer number of people that they reach), but family members will have influence that’s of a higher quality—and more likely to influence behavior. It is still the influencers among that group, however, that may have the biggest impact.

Gavin O’Malley recently wrote about a study that found that people were more likely to try a product that was recommended by friends and family than they were to buy a product recommended by a well-known blogger. His conclusion is that “so-called ‘influencers’ might have less clout than some marketers think.” The authors of the study, the Pollara research firm, reach a similar conclusion, “Marketers might have to reconsider who the real influencers are out there,” according to the article. This flies in the face of over 50 years of thought on the issue. But if friends and family are most influential, should marketers give up on finding social influencers? I’ll explore this issue in a series of posts.

Before we get to the issue of how important these influencers are, let’s think about who they are thought to be. In Malcom Gladwell’s 2002 book The Tipping Point, he writes about Mavens, Salesmen, and Connectors, each of whom plays a role in spreading social phenomena and critical information.  Mavens collect information about the marketplace, often to the point of obsession. They remember specifics, down to specific prices. They also want to share the information with others, and to help others find good deals—helping is one of their primary objectives. They continually teach and learn, but don’t try to persuade.  Connectors are interested in connecting others—they are primarily responsible for introducing people to other people. They are action-oriented: They don’t sit around and think about things, but rather they do things. They are curious, self-confident, sociable, and energetic.  Salesmen are persuaders. Their enthusiasm allows others to trust them. They are energetic, likable, optimistic, and emotionally expressive. They love being the center of attention.

These people sound as if they might have a big impact when it comes to the spread of social phenomena, but O’Malley and Pollara seem to think that you’re more likely to seek advice from your loved ones than you are to seek it from experts, social butterflies, or persuaders. In my next post, I’ll discuss whether “so-called influencers” are still worth targeting.

The web started as the software equivalent of a library. At first, it was only you and the books. You could “browse” different pages of different books, look at the pictures on the pages, and read on. As the initial years went by you started noticing that there are other people in the library as well. They were sitting and reading - like you, but it was hard to know more about them. A bunch of faceless mannequins or crash-test-dummies…

We all know what happened next: the library started carrying videos, CDs and all sorts of other stuff - all for the browsing pleasure of the visitors. Then, visitors were allowed do donate their own stuff if they wanted. Image sharing, video sharing, blogs, and social networks. The mannequins got a face, clothes, and their chance of writing a book about themselves and publishing it in the library!

Use this analogy and you understand why most people use MySpace and Facebook to stay in touch with people they already know and rarely meet new people. Reading books and watching videos is fun, reading other people’s books and watching their videos is more fun. And - leaving them a note at the end of the book is even more fun. But it’s still not really social. There’s no party going on at the library yet.

So what’s missing? We call it “social fabric.” It’s the thing that gets a party going, the vast flow of information that people pick up instantly about other people. When you walk in to a real party you see people move, interact, respond to their environment (the music in the room), and choose their spot (standing next to the punch bowl). You can immediately tell who’s a geek, a party animal, a sports fan, or a metal head. And more importantly - you know who is like you and who reminds you of people you know.

And while it could be pretty darn difficult to get a party with 100 people going, getting it going with 100 million people is much harder. This is what Signal Patterns is trying to do.

Party on

Last October, Google launched an attack against the much vaunted Facebook platform by promising to be even more open with its own OpenSocial. It had also managed to assemble an impressive line-up of launch partners including MySpace. For a while it seemed as if the world had split into two camps — Facebook and the rest of the world. It seemed as if yet again, Google had pulled the rug under someone with its stellar execution and superior technology. Or had it?

Fast forward just a few months later to Spring 2008. Pretty much every one of the launch partners is announcing its own proprietary extensions to the standard. It seems OpenSocial is fast becoming the lowest common denominator — a spot it shares with many such poorly executed standardization efforts. Clearly there is huge consumer benefit if there are open standards in this space. A walled garden is not the right model. But, as OpenSocial partners and application developers (including us) quickly discovered, each of these networks caters to a different user base and serves different needs — from the attention-challenged MySpace to the staid LinkedIn. So, in practice, the promise of write-once, run-anywhere will not be possible with OpenSocial. As an application developer, I am forced to design my MySpace application differently from my LinkedIn application.

A better model that is emerging is exemplified by applications such as LivingSocial, Chirp and ours (See a sneak peek below). Each of these applications is offering a valuable service and does not require you to be logged into the various Social Networks. You get to share your information and get updates from friends no matter where you are — just like you IM with friends in AOL, Yahoo, Google Talk, Jabber or MSN using either a multi-protocol chat client or through gateways that allow you communicate with friends elsewhere. Users will not remain stuck in a walled garden and will cruise wherever they want to, however they want to. Then, and only then, should Open and Social be mentioned together.

“We barely had TV when I was growing up.”
“Really? So you must have been on the internet all the time then…”

This was just one of the highlights from our West Coast trip last week. Though we’re deep in developing our social web applications (coming real soon!), a colleague and I took a quick detour last week to go to the O’Reilly Graphing Social Patterns conference in San Diego.

It was a great opportunity to meet and learn from a number of the leading players in the social networking space: Facebook, MySpace, Google (Open Social), application builders and users, and social ad networks, among others.

Forrester’s Charlene Li had one of the more interesting talks. She started things off with a simple concept: “Social networks will be like air.”

They’re becoming so commonplace (if you have kids, is Webkinz not your top bookmark?) that they’re just part of the environment…nothing special. Heat, food, Facebook?

Coming down the road in the not too distant future she sees data portability and then ubiquitous social networks emerging. But what do people do on social networks today? Certainly a big chunk is collecting ‘friends’ and wasting time playing lots of basic, simplistic games. The novelty of the plethora of available applications is still new, and there’s certainly lots of interest based simply on these applications’ existence, rather than any compelling value or true insight being shared. But I thing we can all agree we’re in the early innings of this game.

One of Charlene’s recommendations was to “compete on creating the most compelling social media experience, not social graph lock-in.” We couldn’t agree more. Our social web applications will leverage your existing social graph, wherever it is, while adding value to your time online and presenting opportunities for you off-line. Walled gardens, silos, whatever you call it - that just doesn’t fly anymore.

She added that you need to “develop social applications that have meaning.” Exactly. The voracious appetite for the volume of simplistic Facebook and other networks’ applications will likely wear thin and users will instead focus on a few solid applications that provide a real value-add to their social networking (whether on-line or off-).

This was good to hear. Compelling experience and apps with meaning. Sounds like a plan.

Don’t we all remember that Friday when OpenSocial was announced. People sat and watched the hour-long campfire 1 video, held meetings, made calls, sought further input… When the smoke cleared it was certain that this is going to be big and important - some day…

The business practice of engaging with a small set of partners and announcing an alliance is a legitimate one. However, an alliance is not a consortium, and it has little to do with standards and with being open. In fact, a standard is agreement among competitors.

For OpenSocial to be open, it should be open with “write permissions” not “read only”. Back in the day when I was still a speech-processing research guy at IBM labs I participated in IETF and W3C standardization activities. What I’ve experienced is companies opening up their internal APIs or technology, and then having to go through revisions of it for their competition before a widely accepted (and then hopefully adopted) outcome is finally reached.

We all love Google, and for good reasons. They changed great and fundamental things about the web, and they introduced a new standard of web ethics and practices - Google is king. OpenSocial is clearly a significant and important development towards a more open web, but the way the OpenSocial move was carried out is not in line with Google’s do good spirit.