Hi Everyone:
Can you think of any other points other than what I have listed below when measuring Community Growth?
Layla,
That looks like a well thought out list...
Have you enabled Accepted Solutions in your community? If so, I might throw that in the mix, and perhaps some derivatives of it...
How about some qualitative measures, like the number of, and level of participation of your star members, your super users? Is your community attracting and / or developing more of these types? Of the ones that you track, how does their participation trend over time?
How about some external analytics that help to understand traffic sources feeding your community, and perhaps the number of links in and out of your community to help gauge how influential it is becoming. Being linked to and referenced from other communities, perhaps?
Mark
This is a wonderful topic Layla!
There a huge number of metrics to choose from, and part of the danger for me sometimes is information overload. So I like to pick a couple of measurements that tie directly to what I am tring to achieve with my communities and focus on those.
'Growth' can be a broad goal, so sometimes it helps to focus in on exactly how you want to grow:
The key for me is to focus on what I want to improve and find metrics for that. Then I can measure the relative success of different improvements as I make them.
Good question and good responses. Layla, in terms of community growth I usually recommend that communities start by measuring seven things:
Posts
Page views
Registrations
Searches
Time to Response
Posts per Board
Top Posters
The goal is to cover the full spectrum of community usage. Posts are obviously key -- no content, no community -- but since most community users never register and never post, you need passive metrics like page views and searches too. (Per your list, sessions and unique visitors are also helpful here.) Registrations and posts focus on that smaller percentage that participate (logins and time on site also address this segment), and top posters focuses on the tiny percentage that contribute a large percentage of the content. All of those segments need to be healthy for a community to grow.
At the next level down, there are metrics that are helpful in analyzing the trends that underlie changes. You want to break down posts into threads (or OPs) and replies, because you want to know when post count rises or falls whether the change reflects questions or more answers or both. (Answers come disproportionately from a small number of people, and questions come from everyone, so how you'd address a drop, for example, would greatly depend on which metric has dropped.) You want to measure time-to-response because it will tell you whether people on average are getting a response in a day or less, which is generally what people expect. Anything less will inhibit growth. Same with posts per board -- there's a ratio at which communities continue to grow and thrive, and below which communities stagnate and fail. You have to keep on the right side of that equation even if it means reducing boards or putting off internal demands for new forums.
Beyond that, there are other metrics you might view only when you want to dig further down into a particular trend or question. We have more than 100 metrics in the system now, and that's why.
Hope this helps!
Google Analytics is great for any Community Manager to have in their arsenal; since it allows you to look at the bona-fide Web analytics while considering the dynamics of online community.
Personally, I focus on these metrics, while acknowledging the others that were previously mentioned above:
Hi Layla,
I want to point out an amazing thing that seems to be happening. We're in week 32 of our forums and at around 7000 registered users. By estimates using the 90:9:1 rule we can roughly guess we've had about 70,000 unique visitors and expect about 50-100 fairly active users. Pretty much spot on.
In the last 4 weeks, we've seen something cool. While our new registrations have remained steady at about 300/wk, our posts/wk have gone up 40%. Even more incredibly, our time to response has dropped 15%. We have not stepped up moderator/employee contribution. The users are really started to help each other.
This opens up another question to Lithium and their partnership with Omniture, we can get time to response, but can we see that number filtering out employees and moderators?
Thanks,
Dave
Dave wrote:This opens up another question to Lithium and their partnership with Omniture, we can get time to response, but can we see that number filtering out employees and moderators?
At present there is no way to get metrics for Time to Response by role today; though I too think that would be a useful way to slice and dice employee participation.
Hey Dave, you feel like adding an idea to the suggestion box or the new ideas area? ![]()