Web Social Architecture

Your Online Community Beta Program: Metcalfe's Law and the Cold Start

Two Approaches to Solving the Cold Start Problem

Part of the challenge of launching online communities is offering something valuable for participants right from the start. Think of it this way: When the value of the system is provided by users, the system has nothing to offer its first user. So why would the first person to show up have any interest in sticking around?

This challenge is known as the "cold start problem," and there are two important ways to address it. One is to provide interesting content and present opportunities to engage with it. The other is to seed the community with participation from an invited group of people who have a reason to help you.

We've used these two methods in combination recently on Microsoft's The Art of Office and on Helio's The New Social Etiquette. In both cases, it's begun to work: we're seeing increases in both traffic and registrations. Content, and the value potential of the good content, is starting to accumulate.

But neither site is really there yet, in terms of offering instantly perceivable value. How do we get them there? And how long will it take?

Metcalfe's Law Provides Some Guidance

Metcalfe's Law says the value of a social system is proportionate to the square of the number of users:

I love that kind of stuff! Here it is expressed in a line graph:

 

Now, I do think you could overstate the relevance of this principle to online communities, because it assumes, for example, that all "value" is created equal. There's a lot more to value than volume of activity. I also think there's a ceiling on value, because at a certain point the value provided by the site starts to become redundant. However, Metcalfe's Law remains very useful as a way of thinking about managing an online community launch.

The line graph above has another assumption built into it: That the number of users increases in a linear manner. Not true out in the wild, but in a controlled beta situation, this is more or less "equivalent to" accurate (in the same way value is, according to Metcalfe, "proportionate to" the number of users squared).

Nonetheless, the important message of Metcalfe's Law holds true: Value can increase at an exponential rate relative to the number of users. (See my post on the early woes of Treemo for an example of the opposite being true.) This is especially important to consider in a beta environment.

Metcalfe's Law Applied to a Beta Program

A common misconception is that Web 2.0 beta programs are designed to provide a preview to the site before it's fully functional. Not so. That is, it shouldn't be so. Your beta program should be designed to accrue value for your site to the extent that casual users will consider it valuable enough to engage.

Your beta program should get you to the point where your site's value crosses the threshold of perceived quality--the point where new users instantly recognize the value of the site and want to participate.

 

But how do you know when you're there? How do you know when your users consistently think your site rocks? How can you be sure enough to bet the cost of your expensive launch PR campaign that people will like the site and want to use it?

There's one good way: Survey first-time users in your beta program, users who aren't biased by their familiarity with the site.

Yes, this means you need to keep adding users as your beta site matures. In, kind of, a linear way. Because only first-time users can tell you whether other first-time users will get the impression that your site rocks, that is, that enough value has accrued to make the value of the site obvious.

 

When the value of the site reaches a point where people begin to spread the word, then the increase in the number of users becomes subject to a network effect, increasing exponentially and sparking a proportional increase in the rate of value accrual. 

Like so:

 

You can imagine the potential: People go nuts, they throw you a parade, and so on.

Hey, it's worth a try! Why shouldn't you have a parade?

Talk Like a Pirate Day: Huh?

Yes, in case you didn't know (hardly seems possible) September 19th is International Talk Like a Pirate Day. You can get all the details on the official site.

Here's the thing: The world needs good, regular doses of fun and silliness. I'm totally into it. Tomorrow I'll wake up in LA and head to the airport for Seattle. Anyone who talks with me and seems even remotely to have a sense of humor will be dosed with some fun. That's right. Stand back. I'm going to talk like a pirate.

And, I have to ask myself. How did I get interested in this, exactly? This whole thing is a really strange idea.

It's the power of the Internet, and more particularly of the social Web. Things like International Talk Like a Pirate Day are out there, things that are fun, and silly, and are not exactly dumb. How do we hear about these things? How, for example, did you hear about International Talk Like a Pirate Day (I just love saying those words)?

The ways we hear about these things, these random online things of which there are untold thousands, but only a few we notice, are through the channels of the social Web. So do tell, how did you hear about the most wonderful thing of September 19th, International Talk Like a Pirate Day? I'd love to know.

Fame! Fun! Free Snacks! Call for Speakers for Seattle UX Event

Last month, ZAAZ hosted an event at our Seattle office for the local user experience community with a group of four speakers. Lee Lefever of CommonCraft spoke on online community, Paul Ingram of Ingramlabs spoke on "inventrepreneurship," Jason Carmel of ZAAZ and Wishradar spoke on optimization, and Emma Rose of the University of Washington spoke on design ethnography. You can read my wrap-up here.

The event was a ton of fun. Each speaker gave a 15-minute talk, and we ended with a panel discussion and audience Q & A. The level of participation was really exciting. We had great audience questions, and the feeling was low-key and casual. The format is meant to encourage socializing (as is the beer).

We'd like make the event a regular series, and to do that we need you. If you've got a topic you'd like to talk on, drop me a note (ryant at zaaz dot com) and let's discuss it. This is a great chance to polish up the public speaking skills in a low-pressure, fun environment.

I'll figure out the schedule once we have speakers lined up and can juggle everyone's calendar. We're thinking sometime between now and the holidays.

Interested? Get in touch!

A Sure Sign Facebook Is, In Fact, Taking Over the Whole World, and How You Can Too

I have been putting off joining Facebook. What a Luddite I am! Like everyone else, I've been, over the past several months, getting slowly buried in Facebook invitations from friends and acquaintances who've "gone over." But I already have a dozen Internet homes. I already have a multitude of connections with my people, in a range of communication channels. I haven't felt the need to add another. Today, finally, the last straw arrived.

I received a Facebook invitation from my father.

OK, OK. I'm signing up. Resistance--clearly!--is futile. Not that my dad is technophobic. It's just that, I'll say, he's not an early adopter.

As I went through the signup process, I have to say I was pretty impressed with the add email contacts feature, and I think it probably explains how it came to pass that I got an invite from my father. When you join Facebook, it asks you for your email addresses and passwords, and it logs in to your accounts, scanning your contacts to see who's already using Facebook. It offers you options to "friend" (the best new verb, by the way, since "google") existing Facebook users and invite the rest to join.

The beauty of this approach is that it taps into the existing networks of relationships we formalize in our online address books. By adding a person to our contacts, we effectively say "I know this person." And what a great service, for Facebook to "hear" that declaration and make it importable. We're saved having to manually add contacts, and Facebook takes over the world, spreading itself like a salty joke across the Web's relational architecture.

I can imagine a couple other opportunities to use the same kind of mechanism. What about, for example, our instant messaging contacts? What about the blogs we subscribe to? What about the lists of permitted visitors to our YouTube channels and our Flickr photostreams? Our contacts on social media sites? What about other members of our online communities?

Facebook's contact finder thingie is a great example of a mutually-beneficial service built on an architecture that transcends domain. The network of human relations crosses boundaries of web site ownership, Web service, and platform, but it is, nonetheless, machine-readable.

The Internet is rich with relational architectures. Too rich, in a way--if you were to envision relationship "services" that tapped into all our established contacts and relationships, across all our online identities, the need for some kind of automated classification of contacts becomes pretty clear pretty fast. There's a reason we group contacts, a reason we have different contacts in different places.

Some day I'll blog about a typology of online relationships. Maybe there's a good one out there already--please let me know!

It all makes me wonder: Do you think the old man has my RSS feed? Scary.