Growing Successful Communities

[#1]

This week Lithium began sending out a weekly Community Health Report to our customers, and as the Lithosphere is one of our communities I also received a report with our community's CHI score. So I thought it might be interesting to share CHI scores, and discuss what they mean to you and how they help you put your community's health in perspective.

For instance, the Lithosphere's score stands at 417 as of last week. As a small B2B community, its not too surprising - larger, consumer communities always feel more active than their B2B cousins. But there's definitely a good deal of room for improvement for me to get that number up. An while I can still show that we are getting tremendous business value from the community, I now have an objective measure telling me that there is even more value out there for me to achieve!

For those who received the report, I'd be interested in hearing what your CHI score is, your first impressions and what ways you might think of using this new tool in the future. For those who haven't yet seen their score or are not Lithium customers, how do you think you could use CHI for your community?

[#2]

A quick note on a couple of related postings:

• Michael Wu begins a series on the different health factors of CHI - and check out his other posts on CHI as well!
• Paul Gilliham (aka bladefrog) posted a blog article about CHI and his community's score that you should definitely check out!

If you see (or start) any conversations about CHI in communities, share them here!

[#3]

I've been reading all this and thinking a little about the different measures.  Today, as I was doing a little housekeeping deleting placeholders, I realized that they are messages with no reply.  Do they affect the responsiveness metrics, or are they filtered out?

We hadn't cleaned them up in a while, so even with advanced search and batch processing, it was kind of tedious getting rid of them.  Potential's idea for Auto Expiry of Placeholders looked good the first time I read it and seems positively brilliant now!  ;-)

Thanks - Jane

[#4]

Hi, Jane,

Thanks for the question. Currently, the message placeholder does not affect the responsiveness health factor, because we excluded all messages that do not have replies, when we compute the average response time between successive messages. I know you are probably thinking "WHY???" And the reason is that it is already accounted for in the interaction health factor. If a message has no replies, there is no interaction, so Interaction = 0.

However, we did not exclude read-only messages, such as placeholders, announcement, or other read-only messages. If you have a lot of placeholders, it will lower your CHI. However, the effect comes in through Interaction, not responsiveness. So, it is a good idea to periodically do some housekeeping and delete the placeholders.

There are already plans to exclude read-only boards and messages in the next reformulation of CHI. So please bare with us. I've started a miniseries of blog articles on the current and future of CHI and their health factors, so you can read about all the changes we plan to make.

[#5]

Mike,

So I was kind of on the right trail thinking that the placeholders would negatively affect the CHI - I was just looking at the wrong health factor.  I'm happy to hear about the plans to filter out placeholders in the future, and I hadn't even considered that the read-only threads and boards would affect the CHI as well!

I wonder how many times I can bump Potential's idea about automatic placeholder cleanup before Scott smacks my hands?  ;-)

Thanks for explaining how it works!  This is pretty interesting stuff and gives me a lot to think about!

Jane

[#6]

---

jloyless wrote:

I wonder how many times I can bump Potential's idea about automatic placeholder cleanup before Scott smacks my hands?  ;-)

---

It's all queued up just as soon as we complete this new bump-prevention feature, Jane. That one consists primarily of a electric jolt through your mouse or other pointing device that feels just as if you've been smacked sharply on the back of your hand.

[#7]

Eek!  Eek!  Eek!  ;-)

[#8]

Hello Jane,

Thank you for your interest. You're definitely on the right track.

These kinds of conversation is exactly what we like to hear. It will really help us make CHI better and more suitable for you needs. So keep the discussion going. If I forget to reply because I got so busy reformulating CHI, Scott will come smack me too... 

[#9]

I just got my community's first CHI.  I have hereby decided to pronounce it "Chai" like the tea, not b/c I love Chai tea but b/c I like the smell and I like hanging out in coffeehouses, reading, and otherwise looking smart.  Boy, can looks be deceiving!

The score was 549, and it was dragged down b/c of the low score of liveliness.  We had a major spike of activity about 8-5.5 months ago.  I mean, it was crazy.  Insane.  We were wondering whether we should shut down the forum for a while (we didn't).  Anyway, the CHI is still judging the current, much calmer, state of the forum by the tail end of that spike, and so that's funny.

On the good side, one of the big bosses is interested in using the CHI as a metric for judging the forum's performance and value, so that's a cool thing.  And given that the CHI will soon stop looking at the big spike, the score will probably be on its way up! 

(Now I'm just wondering whether I should explain that in alot of detail to said big boss, or let it ride so the forum will look better when the CHI goes up!  Haha!)

[#10]

Hey Guys,

Great stuff on the CHI. Bladefrog and I have been looking at CHI for Juniper's J-Net community and his blog is indeed an excellent summary.

I was wondering if Lithium plans to make the CHI report available on a per category/board basis in the future? I realize part of the compass like liveliness may depend on variables such as number of boards, but we've been getting requests form business units on how their particular areas in our community are doing, and something that's a bit more sophisticated than post count and time to response would be great.

Not sure if it even make sense to do this considering how CHI is calculated at this point, just wanted to toss the idea out there and see what you guys think.

Thanks

Anton