social networks pick better roommates. RAs rejoice

So the Chronicle of Higher Education is reporting that more and more colleges are leveraging social networks to better match incoming students as roommates. (Cue the grampa voice: "In my day, we just threw them together in a room and let the RAs sort it out.")

I was one of those lucky Resident Advisors back in college forced to fix these personality clashes. And maybe that's where I cut my patience and diplomacy teeth mediating disputes between mis-matched roomies.

I really had to wonder how often the housing folks simply gave up trying to match students as roommates based on the information supplied by the kids who knew damn well mom & dad would be glancing at their proffered details before they went in the mail to school. How many non-smokers were curiously paired with chimneys? How many bookworms found themselves rooming with the life of the party? It was all lies on the form, and the staff could only work with the cards they were dealt. Besides, the info that seems to most indicate how well we'll click (religion, attitides toward sex, political beliefs, party habits) isn't the kind of stuff that colleges care to, or are allowed to, keep in their files.

I don't even remember what I put down as my habits/likes/dislikes on the form back in 1987, but I do remember the guy I got matched with as an incoming freshman was, um, curiously odd (no offense, Roberto, if you're still stalking me all these years later). I of course blamed the system. There's no way I was that eccentric. Nuh-uh, not me.

So now, knowing that kids are already scouring MySpace and Facebook for the sordid details that'll repel or attract them to their next roommate, the colleges are getting wise.

Tulane University recently announced a partnership with RoommateClick, a service that allows incoming students to select roommates through a closed social network.

RoommateClick provides about 10 colleges with customized roommate-choosing networks, where students fill out a questionnaire regarding their living habits (cleanliness, smoking, etc.) and provide profile photos and other open-ended profile information about themselves. Students can then search the closed network to choose roommates if they don’t wish to be assigned matches by the housing department.

At $5 to $10 a student, this could be quite the lucrative market if it's executed properly.

Heck, I'm sure the RAs would even chip in to prevent the headaches on the magnitude I experienced long ago.

(A Card Sort in Progress photo by Yandle)

wading into the world of realtionship management via CRM

Part of my function at work is to help our philanthropic investment firm understand its own aggregate social network so we can begin to leverage these connections to more effectively achieve our mission of creating opportunity for people to make their own lives better.

Put another way: we need to understand who we know and what they know and how to get to them.

Right now, each of us tracks our own individual networks using whatever works best for us as individuals (LinkedIn, Facebook, Address book, rolodex, those squishy bits between our ears), and most of us are really good at managing our own networks. Simply put: we wouldn't even be working here if we weren't good at networking.

But we're going to start growing gangbusters here, and with each new hire, the likelihood of overlapping networks grows and the likelihood of well-intentioned retracing of steps increases.

So how do we roll up each of our networks into a shared network that each of us both contributes to and leverages in a way that doesn't suck the social capital out of these connections?

• Is it a hoarding bias we have to overcome? No. I haven't encountered a single person who doesn't want to offer up all their contacts for the benefit of the group.
• Is it a technical problem we have to solve? Yes, partly. It's not a matter of simply dumping everyone's address book into a big shared file on the server. That's a band-aid solution. We need something that meets the who/what/how criteria on-demand and accurately.
• Is it a human process problem? Yes, partly. This touches on the squishy bits between our ears method that all of us use to track our social networks. We're going to have to find ways that are in-the-flow of our work to capture and record the states of our networks. No matter the technical solution, there'll have to be some process changes, and if I had to guess, this'll be the most difficult part of the relationship management problem

Understanding this network is key for us to best manage the whole investing process from deal flow pipeline to deal-making to post-deal portfolio management. I'm not looking to implement the Pipeline-to-Post-Deal management solution in one fell swoop, mind you. I think a shared contact database is the best place to start, and we can build out from there.

As a first step, I'm looking to hire someone to do a survey of similarly structured organizations (Venture Capital or Private Equity or Financial Services) to see how they handle CRM, especially how it integrates into the deal management process from pipeline through deal-making and out to post-deal management systems. All my informal surveys so far have yielded a hodgepodge of home brewed solutions like "we track in a big Excel spreadsheet" or "a custom Access database" or "we leave it up to the individuals to track that stuff."

I've got feelers out to a couple folks to see if they'd be interested in doing this survey on our behalf. Do you know of anyone you'd recommend for the project?