Watching Websites

(I wrote up a detailed outline of this at Bitcurrent.)

I attended a panel on crowdsourcing support at the SIIA Software Summit. The panelists had some interesting statistics on what the crowds within an online support community are like, and what they look for.

Changing metrics for changing focus

On SAP Community Network, 90% of people consume information; 10% contribute it, and 1% are active. No news here — this is consistent with findings by Charlene Li, Jakob Nielsen, and others. But the data that mattered to the community changed as it matured:

• Early on, SAP attracted people because it had content you couldn’t get anywhere else, so the metrics that mattered were those of a content publishing system — who’s creating content, what are people reading most, and how good is the content you’re creating.
• After some time, the connections community established with other people started to matter more, and the focus shifted to tools for establishing connections — so analytics looked at who was befriending whom, regulating spam, and the like.
• Eventually, the site was popular enough that a community existed in its own right, and it became a point system for ranking and thanks. The focus was a reputation management system — and the analytics had to track leaders, scoring, and so on.

This happened over a period of 6 years, and the company invested heavily in things like member recognition. Ultimately, community members with high rankings were able to use this on their LinkedIn profiles, because it’s a sign to potential employers of that person’s expertise and ability to work with others.

The goal of your community changes your magic 1%

Over at Lithium, they also have a 100:1 ratio of consumers to active contributors. But they point out that the nature of that 1% varies depending on the goals of the community.

• If the goal of the community is to drive down costs, your ideal 1% is the folks who have the answers.
• If it’s new product ideas you’re after, then you care about the 1% of members who ask the best questions.
• If you’re trying to generate leads, your perfect 1% is the people who know others.

The payoff

The payoff for these communities is big. First of all, there’s the reduction in support costs. Each call that doesn’t happen saves the company $5-$10. But there’s also the fact that the community knows better than a single vendor. Every support problem has many moving parts — browser, router, carrier — and no one company knows all of the issues. But the community does. Vendors simply can’t afford to test with every possible combination. But communities, by definition, can.

Ultimately, support communities are one of the most popular, visible sources of community ROI. But expect to change the metrics you track as they mature and as the goals you’re after change.