My name is danah boyd and I'm a Principal Researcher at Microsoft Research and the founder/president of Data & Society. Buzzwords in my world include: privacy, context, youth culture, social media, big data. I use this blog to express random thoughts about whatever I'm thinking.

Relevant links:

Archive

Can the iPhone hit crucial network density for noticable cluster effects?

On Friday morning, I was shocked to find my always-empty neighborhood AT&T store host to a long line of iPhone cravers. What shocked me even more was that the diverse group didn’t look like typical Apple consumers. They sold out quickly and are still sold out. I remarked on this to the cab driver and he smiled and raised his Gen 1 iPhone, telling me that his cousin wanted him to borrow it for a few days to convince him to get one. His cousin thought it would completely change what it meant to be a cab driver in LA. Not only would it give real-time traffic info but it would let him know where his fellow cab friends were with ease. My driver was starting to agree with his cousin (who should definitely be earning commission for his iPhone sale).

I had never thought about the cab driver case. Cab drivers in my city are always so excited to see a familiar face on the road and they wave enthusiastically. Those who hang out at the airport have strong networks of fellow cab drivers who wait with them. While they’re always tethered to their company, the iPhone would let them connect to one another all day long. I could just see the joy in this driver’s face as he imagined when he’d be able to look at the screen and see all of his friends on the map buzzing around the city alongside dots telling him which surface streets to avoid.

I’ve been anxiously awaiting this launch in the hopes that it might show the power of cluster effects wrt mobile phones. Cluster effects describe the emergent practices that occur when the density of infrastructure adoption in a social network reaches a critical tipping point. In other words, cluster effects are the cool things that people do when all of their friends can do the same things. We take cluster effects for granted in the Internet space because, by and large, entire friend groups can jump onto a computer, grab a browser, and login to a website. In terms of clusters, the barriers to Facebook or MySpace are more personal than infrastructural. (Those who lack general access tend to have friends who lack access.) Mobile phones are different. Even if all of my friends have a Nokia N95, the likelihood that we’re all on the same carrier with the same plan is next to null. The result is that I can’t install an app onto my phone and expect all of my friends to be able to play along. This kills mobile social software from the getgo.

So far, there have been few examples of dense mobile adoption platforms. There’s the Crackberry, but that audience isn’t exactly the most innovatively social. The Sidekick was impressive amongst deaf communities and urban youth, but T-Mobile managed to lock that puppy down so heavily that no innovative practices really emerged. Still, if you look at the AIM usage in those clusters, you get a good indicator of the potential. And that’s all folks.

The iPhone has the best chance of hitting that tipping point of anything out there. For the most part, everyone is stuck on AT&T. And everyone gets a data plan. And the phone is semi-open. The price is still out of reach for most high schoolers who rely on parental pass-me-downs, but it has a decent chance of hitting other clusters. I was banking on urban 20-somethings, but I love the idea of it hitting cab driver clusters.

Right now, a phone is primarily a 1-1 communication device and, if you’re lucky, an information access device and a portal to the web. Interesting things can happen when the mobile is a platform itself. In other words, when you can assume that everyone around you has the same tool, you can start doing networked activities that don’t rely on a website. Cluster effects in mobile will be what happens when the LCD is not texting. From there, you can innovate. Sure, we’re going to see a plethora of mobile social network sites and mobile location friend services and mobile dating and mobile media sharing communities. The first wave will always be a translation of the web. But once you have cluster effects, you can also start innovating and finding new services and tools that allow people to connect in meaningful way. New games can emerge. New social services. Innovation in this space will be iterative – it will involve throwing things out to the market and seeing what consumers do and do not do. It will require iterating based on their practices and not trying to shove those curvy creatures into square holes. But there’s no point in leaving the starting block until cluster effects are underway because, sadly, iterating in imagination land inevitably leads to techno-utopian fantasies instead of meaningful applications.

Gosh do I want to see cluster effects triggered in mobile space. There’s such great potential for interesting things to take place. Sure, I’d rather see it take place on open platforms and open networks. And I am a bit worried that, without openness, we’re going to see some not-so-good side effects. I definitely share Zittrain’s fear of non-generative technologies. But part of me would rather fucked up market effects trigger cluster effects instead of governmental decrees. We all know that something has to break in mobile somewhere sometime soon. Our options are limited. Option 1: all carriers and handset makers need to start playing along. Option 2: some combination of handset/carrier triggers massive adoption. Option 3: municipal wifi emerges, allowing the web to serve as a temporary bridge. Option 4: governmental intervention demands platform infrastructure. These options all have downsides… Option 1 is a pipedream. Option 2 creates a monopoly risk. Option 3 will take a long time to unfold and still requires handset compatibility. Option 4 is more realistic in some countries than others.

Anyhow, there’s a decent chance that Apple & AT&T will screw this one up, but they have the best chance to hit Option 2 right now. And really, I’m bored. And I want a new phenomenon to study. And I want to see what happens when people can do weird and interesting mobile-based social stuff. I’m especially curious how this might affect mobile-centric populations, although that’s still a ways off. But yeah, possibility is in the air.

So…. AT&T, Apple, and Market Research Firms: I strongly encourage that you watch the network density of iPhone adoption. (Note: raw numbers don’t matter… you want density of adoption amongst pre-existing friend groups.) If there’s anything you can do to encourage network density, you won’t regret it. If you can tip full clusters to the same platform with all-you-can-eat plans, you can launch all sorts of interesting things that will fundamentally alter practice and change the mobile landscape. Please don’t screw it up.

Print Friendly

8 comments to Can the iPhone hit crucial network density for noticable cluster effects?

  • How are cluster effects different from “network effects”, or in a more economic sense, the “Mathew effect”? As I see it, clusters tend to get analysed as conceptual categories (as in the data mining notion of “cluster analysis”) or sometimes as physical effects. The network effect is more aggregative and can swing wildly out of control (which is why they are often studied analytically as log normal). I think the political and social effects are likely distinct too, although I am unsure what exactly they would be.

  • Could ‘a good browser on a phone’ be enough to drive this? Does it have to be a monolithic uniform platform to cause cluster effects?
    I’ve definitely seen cluster effects around Twitter…

  • Christina Pikas

    What about Android? Maybe that will have some of these effects?

  • Quinn: I see cluster effects and network effects as slightly different, albeit complimentary. Network effects describe how an individual’s adoption is influenced by and influences those around that person such that an individual benefits from all of their friends adopting the same platform. Network effects are core to reaching critical mass and tipping points. In my head, cluster effects are what happens once critical density is achieved. They are the practices and dynamics that take place once a cluster is entirely engaged. Rather than being the effects that solidify the network, they are the effects that are caused by the network having formed. I don’t think that I’m being entirely clear so I’ll have to think of a better way to describe this.

    Kevin: a good browser + universal data plans would equal a mobile web. Yes, that could work, but I don’t even see basic browser compatibility across phones (let alone data plans).

    Christina: When Android is adopted en masse, we can talk. Best it looks now, it’s just another platform competitor in an already diverse market that is going to be adopted by some phones and some carriers. What I’ve seen so far suggests that it will be adopted by quite a few, but not in meaningful clusters.

  • Steve

    I wonder to what extent cluster effects are permanent versus transitory. In the seventies CB radios were a big deal, and a “cluster efect” was created amoung professional truckers which was extensively celebrated in song and cinema. As far as I know, CB technology still exists. I suspect that most drivers even still use it. But nobody seems to make a big deal about it. It’s kind of off the radar. (“Tired” not “Wired”.)

    Will all the hype around mobile fade when a new generation of adults, now with emerging family responsibilities, realizes that there are only a few basic things that are actually worth doing with a mobile comm/data device?

    (I emphasize the transition to parenthood because that seems to be a point where one’s relationship with one’s peers becomes less of a motivation factor for life activities compared to the parenting/family dynamic).

    Just a thought,
    -Steve

  • ray

    very interesting – thanks i enjoyed reading your piece. a few more things to consider re. strategy for the iphone. in the new economy business models have clustering on two sides. i think u sum up the dynamics from the consumer side well. other points to perhaps mention re. clustering:

    industry clustering and strategic interdependence wrt control points (i.e barriers around your bizmodel):
    what’s happening to the value network? how will apple shape the industry value network and re-shape the industry cluster? is it who takes what – ultimately means who gets what.

    the important point here is that apple is creating a industry cluster and strategic control point around its app store. such cluster effects can be a great barrier of entry – esp for a business model. this particular control point is what will ultimately determine the success of the iphone. on the other side of the business model – an ecosystem of value add apps fueled by the cluster effects of an online market (which is what you described very eloquently.

    oh. but wait. Google…hot on apple’s heels – they’re launching their own app store – probably for android (when its launched 😉 ) **apple shakes fist in air…why i outta…”

    my $0.02

    ray33

  • Yep. I’ve thought for along while now that ClusterFx will be the joy experience that propels mobile webbing/creative roaming from niche to mainstream.

    Like-minded nomads roaming and clustering easily, on the fly, around shared interests, with the opportunity made evident and real by location-based social networking, etc.

    Say, for instance, you’ve got a serious physical malady, and you’d like to get some therapeutic discussion (and maybe even a little massage) going with fellow sufferers in your area, especially those who enjoy a little Miles Cowperthwaite with their Miles Davis, and you happen to be cruising past a boarded up Starbux, and your iPhone/Touch suddenly buzzes in your pocket…

  • It’s interesting that you write of the taxi driver seeing where all his taxi driver friends are on his iPhone.

    The iPhone is the one platform that can’t deliver on this properly, because it does not support background apps (minimized apps). So for your taxi driver friend to see where all his taxi driver friends are *right now* they would have to be actively running the app on the iPhone.

    iPhone users don’t tend to leave apps open to enable that real time context reporting.

    All the other smartphones do allow background apps (Blackberry, Android, Windows Mobile, Symbian, UIQ, Palm Pre). It’s just a pity their user experiences pale in comparison to the iPhone.