Note: Part one of this series can be found here.

On blogs I visit discussing social media, one ongoing debate concerns metrics. Some claim that metrics for social media projects are not meaningful; some claim that new metrics must be developed to gauge social behaviour. Some even claim that metrics aren't needed.

I believe that it is essential to have meaningful key-performance indicators in place for a social media project, as you would for any other project. You must know what success will look like to know if the project is worth repeating or not. Some of these metrics are familiar Web metrics, some are more similar to offline advertising, some are similar to PR metrics, and some are indeed brand new and hard to measure.

For ChangeEverything.ca we measure a few things that are familiar to anyone who runs a website:

• Unique visitors
• Time on site
• Referring URLs
• Natural search results
• Number of registered users
• Number of active users (need to define for yourself)

We also measure "Web 2.0" stuff:

• Our Technorati authority ranking
• The number of RSS subscribers
• Conversions of visitors to contributors (people who make it through the registration or content-creation funnels)

Like offline media, we measure how many people in our geographic region are aware of the site, just like we track awareness of our television campaigns. We also measure how many can link it back to our brand.

Like our PR, we try to measure the earned media the site has garnered for us, which has been significant, with lots of positive coverage on TV, radio and newspapers. We also keep track of what bloggers say about us. We consider a blogger writing positively about ChangeEverything.ca to be an unsolicited third-party endorsement. Happily, almost all blog posts so far have been positive, as have our earned media. It's difficult to criticize this project since most press coverage is about how the site has helped the community in some way. And that leads us to the most interesting metric by far.

Where we need new metrics is on the issue of real-world impact. This was not a metric we had in place prior to launch—honestly, it never occurred to us. But it became necessary because of activities happening in the real world (remember that?) due to the influence of the site. The first time this became obvious was after a bad snowstorm Vancouver in November 2006. The site's amazing community moderator Kate created a post called Got Hats? and asked for people to donate warm clothing and blankets to the homeless. This initiative took off and over the next few days, we estimate that more than 4,000 pieces of clothing, blankets, pillows, and, yes, hats were donated to local shelters, all via communication and organization on the site. The change occurred while snow was still on the ground, while the need was still very real, and even a matter of life and death. It was the first clue that we were onto something truly important.

There are many positive traits the site lends to the Vancity brand. ChangeEverything.ca is more than a bunch of people discussing local issues they want to change. The site has created real impact in the communities Vancity serves. Since the Got Hats? episode, we have seen the impact of the tremendous exposure a woman on the site has received for blogging in depth about her valiant attempt to give up plastics in 2007, the successful implementation of a bike share experiment in Vancouver and now a contest where people can win $1,000 to give to an organization making change for the good (appropriately called ChangeSomething).

Traditional Web metrics can't measure the human terms of this impact, and that's the beauty of social media. It spills over into people's lives, because people are in the driver's seat. We need to expand our view of key performance indicators for social media so they reflect the project's success, which now includes the true impact of these projects on our communities.

I think that explains why those of us who advocate for the appropriate use of social media are so passionate about our work.