Noticed


11
Feb 10

Google Buzz as Experience Pattern

Although Google has made a number of ventures into the social space (Wave, Latitude, Profiles, etc.), they have yet to capture audience and mindshare like competitors Twitter, Facebook, or Myspace. Google Buzz, an interesting and troubling new foray by the company, is the most recent shot across the bow of the major social players. Having spent a day or two looking at the product, I’ll share some of my early feedback.

At its core, Buzz is a content sharing network that exists loosely within Gmail and Google profiles; content shared publicly within Buzz is populated within Gmail and to the profile. To participate in Buzz, one must agree to have a Google profile, a place “visible on the web so friends can find and recognize you.” Notably, the Google profile is at the center of Google’s social search efforts.

As I’ve written in the past, Google’s has had to walk a very fine line with how they “reveal” what they know about your social circle. Realistically, Google sits on behavioral social network data that is of equal value to what is created in Facebook or Myspace. Mining our web search patterns, our chat and email logs, and our travels across the web with analytics, Google knows who we connect with. The challenge Google has always faced is putting this information into play in a way that doesn’t freak everyone out. Google Profiles were a first step in that direction, asking people to list their sites and friends with a promise of better search positioning (sound familiar?).

As the profile and social network play an increasingly important component in search relevance judgments, it is in Google’s interest to leverage the vast social network data it already has. With Buzz, Google can pre-populate friends lists in a slightly less than creepy way, and then leverage that information in social search via the profile. The big win with Buzz isn’t Google’s competition into Twitter or Facebook’s space (Buzz isn’t vaguely a Facebook or Twitter killer), but rather the data value Google is going to reap through a massive profile-creation effort. Buzz just might be glue necessary to encourage people to articulate the extant network connections in Google. For Google, it is more important that you use Buzz once than if you use it on an ongoing basis.

What are the implications of a system like Buzz? It is a pretty interesting case of what might be thought of as data leveraging. As more of our patterns are analyzed in a range of systems, corporations are going to be challenged manage our confrontation with our patterns.  I recall a conference I attended where some senior developers at a messaging analytics firm were discussing the creepiness factor inherent in showing people their behavioral patterns.  Experience and design patterns are rare for this sort of confrontation, and Buzz is an interesting case.

If such data-leveraged confrontations are going to become more frequent, we are challenged by implications of ubiquitous recording.  For example, Buzz is pre-populated with the “people you email and chat with most.”  One doesn’t need to be a Goffman scholar to know that a public listing of the people we chat with most presents social hazard.  It is remarkable to observe how often companies get the defaults of sharing wrong.

While we’re talking about privacy implications of Buzz, here are a few other points I’ve noticed as I look over the terms of service.

  • As I’ve mentioned, to use Google Buzz you must agree to have a Google profile created in your name.  Doing so shares things like your contact network and your accounts on other Google sites.  These are the defaults, which can be changed with effort.
  • You can’t delete your Google Buzz account.  If you create a Google Buzz account and wish to delete it, you have to delete your entire Google profile (killing your search listing, etc. at the same time).
  • If you wish to remove Google Buzz items, you must find them and delete them (“You have the option to remove your comments on others’ posts individually if you’d like”)
  • Finally, you are required to use your first and last name in Google Buzz (“you need to have a public Google profile which at a minimum includes your first and last name”)

Google Buzz is a truly interesting attempt from Google to leverage the vast social network resources that exist within their systems.  The implications of such a move are profound – for the company and for our experience with large-scale behavioral information.

Update: Google has responded to widespread criticism of the privacy defaults issue.  Many are uncomfortable with the following/follower pre-population, so Google has offered some new ways to manage these lists.  Notably, Google’s post still asks that you create a public profile – something completely unnecessary if you don’t have one already (that is, no one can see your pre-assigned contacts until they’re shared with a Google Profile).

I know this is cynical, but I’ve seen this happen enough to know that these “mistakes” often aren’t accidental.  As danah boyd has written previously, “In other words, this is “slippery slope” software development. Given what I’ve learned from interviewing teens and college students over the years, they have *no* idea that these changes are taking place (until an incident occurs).”  It is disturbing to see Google going down this road, trading PR for personal information as a calculated trust violation.


23
Sep 09

New York Times mention

Pretty exciting – my work was featured in a New York Times article on productivity tools.   Farhad Manjoo writes:

“One Mac app that has found a way to solve this problem is called Freedom, which blocks all of your computer’s networking functions for a pre-determined number of minutes. In other words, once you set it, you’ve got no Web, no instant messaging, no e-mail — and the only way to undo Freedom’s block before the time runs out is to restart your machine.”

Freedom also garnered a mention on Mashable, in a post on 20 productivity tools for Mac users:

“Sometimes the Internet is more of a distraction than it is a productivity aid. For those times, the best solution is just to unplug. Freedom will turn off your Mac’s networking card for up to 8 hours, so you can get what you need to get done done, without the distraction of Facebook, Twitter and the latest viral videos.

Obviously, as a web writer, this doesn’t really work for me during the work week, but it’s great for when I’m on a deadline — like trying to finish an article of 20+ Mac productivity tools!”

This was cross-posted to the Freedom blog (http://macfreedom.com).


1
Sep 09

The trouble with Internet surveys

Gary Langer, the director of polling at ABC News, shares the bad news regarding Internet surveys.

In the most extensive such analysis to date, David Yeager and Prof. Jon Krosnick compared seven non-random internet surveys with two others based instead on random or so-called probability samples. The non-probability internet surveys were less accurate, and customary adjustments did not uniformly improve them.

While the random-sample surveys were “consistently highly accurate,” the internet surveys based on self-selected or “opt-in” panels “were always less accurate, on average, than probability sample surveys, and were less consistent in their level of accuracy,” the researchers said. Further, they said, adjusting these samples to known population values had no effect on accuracy (and in one case even worsened it) as often as that process, known as weighting, improved it.

Also noteworthy:

While this paper is the first to evaluate the subject in such detail, intimations of these problems were posted in a blog item this summer by Reg Baker, COO of the research firm Market Strategies International. Estimates of smoking prevalence were similar in three probability samples, he reported, but less similar – with variation of as many as 14 points – in 17 opt-in online panels. In such panels, he said, “the results we get for any given study are highly dependent (and mostly unpredictable) on the panel we use. This is not good news.”

Yeager and Krosnick, meanwhile, provide one more eye-opener: The average highest weight for any one respondent across the opt-in online samples was 30 – one respondent, that is, standing for the equivalent of 30 in the full dataset. (And one went as high as 70.) The highest weights in the two probability samples, by contrast, were 5 and 8.

Nothing new or groundbreaking here, and yes, a little inside baseball, but relevant in the light of all of these web surveys showing that “Teens don’t tweet.”  First, convenience-sampled web surveys can’t offer standard errors, and the weighting process that produces errors is highly susceptible to inflation in areas where data are sparse.  This sparseness commonly occurs when studying the behavior of a low-response population such as young people, and is multiplied when studying an early-adopting phenomenon like Tweeting.

Langer’s blog is a worthwhile resource if you’re interested in survey methods.  And I hope to resume blogging – updating my syllabus, posting some recent papers, etc. – when I get a spare moment.

via Study Finds Trouble for Internet Surveys – The Numbers.


23
Jun 09

Appearance on Morning Edition

I made a very, very brief appearance on Morning Edition this AM:

Fred Stutzman, who studies social networks at the University of North Carolina, thinks charging for services will turn out to be the best way for social networks to get profitable.

“People will pay for good technology,” he says. “People will pay for a responsive company.”

He points to the professional networking site LinkedIn. It offers some free services, but users pay for a premium level with more features. With only 40 million users, LinkedIn is significantly smaller than Facebook or MySpace, but it’s making a profit.

Facebook, though, may face a bit of a conundrum. There are two groups on the site called “We Will Not Pay To Use Facebook. If This Happens We Are Gone.” Their combined membership? Nearly 8 million.

Stutzman thinks that ultimately Facebook, MySpace and Twitter are going to be around for a long time. They just might not be the big cash cows that some people expect.

Unfortunately I missed the live broadcast.


18
Jun 09

Zimmer on the Facebook Dataset

Michael Zimmer has released a new critique of the “Facebook Dataset” – and it is well worth reading.

Recall that last fall, a group of researchers affiliated with the Berkman Center for Internet & Society at Harvard University released a dataset of Facebook profile information from an entire cohort (the class of 2009) of college students from “an anonymous, northeastern American university.” While the researchers took good faith steps to preserve the anonymity of the source of the data (and, presumably, the privacy of the subjects), I quickly narrowed it down to 7 possible universities, and then with only a little more effort, identified the source (with some confidence) as Harvard College. All this without ever even downloading or looking at the actual data.

Download the draft of Michael’s paper.


15
Jun 09

Facebook passes Myspace

Via Inside Facebook: comScore: Facebook Passed MySpace in the US for the First Time in May.

It’s been a long time coming, but Facebook has finally passed MySpace in terms of total US uniques, according to comScore. In May, comScore reported 70.28 million US uniques for Facebook up 97% year over year, compared to 70.26 million for MySpace down 5% year over year.

Blogging this for posterity’s sake.


19
May 09

The State of Things

On Wednesday,  May 20th I’ll be appearing on WUNC’s excellent radio show “The State of Things.”  As I listen to TSOT almost every day, it is pretty exciting to get a chance to do the show.  We’ll be talking about social networking and its recent growth in popularity.  If you’re local, tune in at noon tomorrow – or stream the show online at WUNC’s website.