Berkman@10: "The Future of the Internet" Conference

I'm excited to be attending the 10th anniversary conference - "The Future of the Internet"- for the Berkman Center for Internet and Society at Harvard Law School from May 15-16. I'll be soaking up insights from the likes of Jimmy Wales (Wikipedia) and scholars Yochai Benkler and Jonathan Zittrain.

- Stephane

Puzzler #1: How to Add Companies

Our team often bumps into interesting design dilemmas. We'd love to hear some ideas about them. Here's the first that I'll share on this blog: how to add companies to our database.

Our users have told us that they'd like to be able to add companies to the database. And we think they ought to have that power. This raises two sticky questions, however.

1) How do we maintain a clean user-generated database of companies?

Companies are always merging, changing their name, going out of business - it's a dynamic market. We'd want to avoid seeing "Exxon," "Mobil" and "ExxonMobil" as three separate companies in our database. (Exxon and Mobil merged a few years ago to become ExxonMobil.) We'd also want to avoid a plethora of duplicate entries from typos ("Exon"). These challenges could be managed by setting up a process for our community or our staff approve a newly added company before it is formally incorporated in the database. We could also try some sort of auto-fill feature when a user is entering a new company to reduce the frequency of typos.

2) How do we encourage our budding community to reach "quorum" for a given company?

We aggregate the information provided by our users into scores for each company. But we can't in good conscience post a company score that is based on only a handful of user reviews. We need a minimum number of people - a quorum - to review and rate a company's performance before we can be confident that the aggregated score is reliable. We don't know what that minimum number is yet; we'll test our model to find out.

The problem is that increasing the number of companies in the database will probably reduce the likelihood that any given company receives quorum. This is especially the case in our early years when the size and capacity of our community are limited. If five hundred users each add and review a different company, we'll end up with five hundred reviews but not a single company score that our community can use.

One way to solve this is to create incentives for community members to focus on certain companies, one industry at a time. Our users earn status through the quality of the reviews they write, culminating in a Contributor Score. We might simply offer bonus status points to users who write about a company that is part of our current "focus industry." Users would be free to add companies but encouraged to pool their efforts efficiently.

Have any brilliant solutions for us? Please share.

- Stephane

What's Cool

There’s a David Brooks op-ed in the New York Times today about social entrepreneurs. It’s a good overview of social entrepreneurship, and Brooks brings up interesting ideas like America Forward’s proposal to create semipublic funds that invest in local organizations.

It does, though, underline the fact that social entrepreneurship is appealing to conservatives because it involves a smaller role for government in meeting social needs. This is not necessarily a bad thing, but it makes me think that ‘social entrepreneurship is displacing the public sector because it’s more efficient’ and ‘public sector failures create vacuums that are being filled by social entrepreneurs’ are equally plausible stories.

It’s also interesting the way social entrepreneurship is ‘cool’. Brooks writes about “some of the smartest, most creative people in the country” who are building new social service models based on a “decentralized worldview.” (This article makes me think our team ought to be dressing snazzier.) I think he’s right to be excited, though some of the more hopeful next-big-thing rhetoric reminds me of what people say about any trend that arises quickly and seems like it’s going to revolutionize everything.

But what’s striking is reading this op-ed next to articles on the current financial crisis. (David Leonhardt’s is particularly good.) Over the past ten years I’ve seen investment banks recruit lots of bright and confident young people from college campuses. The quantities of money involved made i-banking cool in a ‘masters of the universe’ kind of way: you could step up to the plate and conquer financial risk and see how much wealth you could create before turning thirty.

But now it looks like nobody knew anything. (If you’ll pardon the snark, some of these guys made millions for not understanding complicated investments. I, on the other hand, would have been willing to not understand complicated investments for free.) In a great profile of the investor Blaine Lourd, Michael Lewis says: “One day, someone may look back and ask: At the end of the 20th century and the beginning of the 21st, how did so many take up financial careers on Wall Street that were of such little social value?” That might be overly harsh. But I wonder if this crisis might reduce the allure of finance for an entire generation of smart, creative, impatient young people- and if social entrepreneurship will be the new cool thing for them.

-Isaac

Cumulative advantage

It’s an exciting thing to write an inaugural blog posting! This is the official blog of Citizens Market, an online database and community dedicated to promoting socially responsible choices in the marketplace. Soon we’ll have an official (and properly epic) “About Us” section on the homepage. The short version is: we’re a six-person team based in Cambridge, MA who wants to make it easier for consumers to inform themselves about the social and environmental behaviors of corporations.

We’re not sure what this blog will be, exactly; partly news and analysis of developments in responsible consumption, partly updates on our website’s progress, the occasional “what-does-it-all-mean” musing. This, like our platform, will evolve.

Today I’m wondering about the factors that ultimately will determine the success or failure of our project. I recently re-read this article by Duncan Watts about cumulative advantage. Watts and collaborators created online worlds where users could listen to, rate, and download songs by unknown artists. In one of the worlds, users only saw the song names (i.e., they were making independent judgments about the music), but in others users saw how many times the songs had been downloaded by other participants (i.e., they were subject to social influence). Success in the social influence worlds was unpredictable: different songs became big hits in different worlds. There was little correlation between what was popular in the social influence worlds and the ‘objective’ quality of the song, measured by popularity in the independent judgment world.

It’s a great article, and/but the lesson is many things become popular not because they have inherent ‘likeability’, but through a series of arbitrary and chaotic interactions. In these experiments, the hits didn’t even result from mobilizing Malcolm Gladwell’s connectors, mavens, and salesmen; they just emerged from random forces bouncing around in a ‘rich-get-richer’ environment.

Our project depends on getting a critical mass of information and users. But it’s possible that we could do everything right and still not take off.

On the other hand, these cumulative advantage effects seem to emerge in ‘rounds’ of popularity games. But in the real world, what’s a discrete ‘round’? If you’re determined and you keep incorporating feedback and re-launching your project, you increase the odds that people will recognize your awesomeness.

-Isaac