All posts by Mukund Mohan

My discipline will beat your intellect

Measuring to get a desired outcome from communities: Best practices from Marc Siegel, IBM


Marc Siegel runs IBM Rational’s online developer community. IBM developer community has over 650+ forums (one or multiple forums exists for almost all developer products of IBM. This represents over 1.5 Million users. Marc purveys thousands of users, and facilitates discussions and keeps “the fun going”. You can reach him via email with some questions.


I met Marc at Le Boulanger in Willow Glenn, yesterday over lunch. He has a very interesting background. After rising up the ranks at several large organizations (including NASA – so he’s as close to a rocket scientist as you can imagine) to manage a team of people, he wanted to get his hands dirty again. He joined a really fun startup – Catapulse, ( know it was fun since a couple of other folks from there remember it fondly) which was acquired by Rational soon after he joined. He joined Catapluse with the express intent to build and grow their online presence.

Here are his top 3 best practices based on years of working these communities:

1. Understand your user behavior, usage patterns and help fit the community into their work (instead of the other way around). Unix adminstrators and programmers were the first “developers” of the products that he was responsible for to build a community. Now, if you know these folks at all they are not exactly patient with the “Windows crowd”. The online medium of community though gave these developers a great avenue to share scripts, information and tips and tricks on getting things done quicker. With minimal frills and a simple design this community has thrived and grown in its significance.

The other interesting point was that not most of the developers were at their desk waiting to help others. Some of them had a 45 min train commute and would love to get the questions and other information to their laptop and work with it offline and respond back the next time they were connected. This is a great example of “making it easy” for them to be a part of the community and “fit into their pattern of work”.

2. Graduate from measuring simple metrics to the ones that affect change. Most metrics collected by Marc today are around page views and user signups. The key project for 2007 is to get the next level of support metrics – how quickly were questions answered, how many questions were answered, etc. These are a better reflection of the objective measurement of the community.

He mentioned the folks from SAP said their developer community was so strong most questions to the community forums (88% of questions) with a mean time to solution of <1 hour. That is a great metric. I would love to learn more from SAP and share it with our readers.

3. Understand the usage of technology for the benefit of your community more than its capability. Most communities are okay with just a message board and simple chat. Putting together a blog and wiki (just because they exist) does not really help UNLESS you integrate it with a process that resolves a problem or supports a use case for the community members.

From Future of Communities: Communities, Social Networks, Collaboration Networks – what’s the difference

I posted this to Future of Communities a few days ago.

Hint: I dont know for sure, since I get easily convinced by different people they are the same and they are all different.

I must admit I was not surprised when I was asked by a prospect
yesterday when I was using the terms “community, social network,
collaboration ” all in the same sentence and trying to give him
examples of each.

He said “I think your strategy is to either convince me you know more or confuse me – please tell me which?”

When I think of the word “community” to me it
usually refers to a group of people that are linked together by
something(s) common. These group of people are joined together for a “common goal”. But there is something they are trying to achieve together that they believe they cannot achieve alone.
When I am expressing “social network”
the perception in my mind (right, wrong or otherwise) is Myspace,
Facebook, etc. This to me is a set of people that are trying to further
their own “self goal” within a group of like minded people.

So a social network has individual first mentality, as opposed to a community – which has a group first mentality.
Collaboration Networks are a nuance. I tend to think of them as temporary groups of people “aligned to a same objective”.
Note I said objective NOT goal. These are more time-bound, with
specific criteria that they must meet for the team and inviduals.

Here is a chance for 3 people to win a Starbucks gift card each. See below.

Lets take some examples. You can put in the comments section what
you think these fall under. All you have to do is put the number and a
letter A, B or C next to it. I will put the first comment as an example.

Bonus points for adding to the definition of these terms.

A= Community, B= Social Network, C=Collaboration Network

1. Youtube
2. MySpace
3. Threadless.com
4. We are smarter than me
5. Intranet for HR folks to share information with the rest of the company on policies
6. Message boards for customer support online
7. Discussion and sharing forums for customers and prospects to understand issues about a particular industry
8. InfoWorld IT EXEC connect
9. Quest Software online groups
10. BMC Software online groups

Understanding the economics of support communities: Best practices from Joe Cothrel of Lithium


Joe Cothrel
has a very deep and rich background with Communities. Having been at
Participate.COM which was one of the first moderation service
providers, he has been researching, managing, running and developing
communities for at least the last 10 years. He is currently the VP of Community Management for Lithium and also blogs there.


Joe works out of Michigan and his team
provides
community management and moderation services to customers using Lithiums on-demand
software
.  Based in Emeryville, California, Lithium was founded in 2001, has about 50 employees and supports more
than
50 customers including Dell, Cingular, Comcast, Linksys, Sony,
Salesforce.com
. Lithium
provide
s a hosted solution for community and
price
s its product on a monthly
basis based on page views. They also charge a one
time setup
fee.
Software fees start at $1300/month. Joes team provides ongoing services in moderation and
management
as well as one-time assessment and
strategy projects.

Since Joe has a good background on support communities, we discussed a
particular issue that consistently comes up with Support
executives:

“I am concerned that if I start a support community,
I
’ll need a very large
staff
to manage and
support
it

This is an understandable concern: the goal of support is
to
keep customer satisfaction levels high and do it at the lowest possible
cost
, and the number one cost of support organizations is
headcount.
  However, the truth is that support
communities
typically dont require a large staff.  

Here
are
three best practices that I learned from Joe to leverage the power
of the community and keep your support costs low:

1. Define your
interfaces
.  
Effective
and efficient support communities are well integrated with existing processes in customer care, technical
support,
marketing, product development, corporate
communications,
and
legal
One of the first questions Joe asks in his
community management working sessions is
Who in your
organization will touch this community in any way?
  You can then begin
to think about whether

the interfaces are
routine (i.e.,
escalati
ng unanswered questions to support
specialists),
periodic (reports or notifications on topics being discussed
in
the community), or ad hoc (i.e., product managers participating in
discussions as their interest and time allows). 
Needless to say,
when communities are isolated from other processes they take more effort to
manage, because everything falls to the community manager.

2.
Develop

your users.
   Joe made the
interesting observation that while
we all talk
about
companies needing to learn and change (Cluetrain
Manifesto
, etc.), we forget that customers are learning and
changing too. 
Community management
is, in part, a process by which we help
customers develop their competencies in collaborating
with
us and with their peers If we dont help them learn and grow,
they
ll simply create the
same problems for us to contend with over and over again.

3.
Develop your leaders.
  According to Joe,
every
community is, at a minimum, two communities:  the community of the
average user, who may come once a week, once a month, or even
just
plain once, and the much
smaller
community of superusers or hyperaffiliates, who are 10 times or
100 times more active than the average user. 
Its not uncommon to see communities
where
0.5% of users provide 60% or 70% of the
answers. 
Are these people important in making
customers happ
y and reducing your costs?  You bet they
are. 
The economics of communities ONLY work if you let the community help
itself.
  The temptation to have support specialists answering questions
in
the community is strong. But you have to be
careful not to set a pattern in
the community that this is the place to get answers directly from the
company.
  You may never fully benefit from the knowledge and experience of your
customers
. 

The Myspace Effect: Why are businesses building social communities? – Interview with Michael Wilson

<img src="/images/64360-56413/Michael_bw.jpg”> I had a
a insightful discussion with Michael Wilson, co founder and CEO of Small World Labs.
Founded in 2005, they are a 15+ people company based in Austin, Texas.
Small World Labs provides a hosted solution for online social networks
and communities. They currently have about 45+ customers including KVIE (PBS radio station), Save the Children, Bootstrap Network , Dallas Morning News 
SportsGist.
They price their offering based on the number of page views the community
generates monthly, plus an initial setup fee.

Michael’s parents are both deaf, but their lives changed with the advent of
Internet communities by allowing them to communicate and interact in a way
unimaginable before, which was the seed of inspiration for Michael. They have
the vision of allowing companies, customers break down the barriers to
communicate with each other and enrich lives.

He clearly sees the market for online communities moving away from the early
adopters to the early majority
, driven by “The Myspace
effect”
, which we have written
about before
. Companies large and small having seen the sheer numbers and
size of Myspace and the cult-like following are actively looking at their own
constituents to see how they can foster discussions and conversations to their
benefit.

Here are the top 3 best practices he is recommending on the path to success
with social communities:

1. Align all user incentives (loyalty plan) towards goals for the community.
He advocates writing down clear objectives for the community e.g.:

a) # of
page views you wish to see annually / monthly / weekly – this assumes you care
about page views – which in a lot of cases translates to CPM (based on an ad
driven model)

 
b) # of
users you want to join as active participants in the community.

c) User facilitated interactions. There was a good piece on “Conversation
Index”
by Don Dodge, which discusses this in more detail. How to track
that you are having a conversation with customers instead of talking at them?

 

2. Dedicate
resources
and align their incentives towards the community goals. E.g. If
my bread is buttered (as a Marketing person) on number of leads generated, and
the community “project” is a night job, the misalignment is obvious.
Instead, ensure that the marketing person gets bonus paid on achieving the
community goals.

3. Measure the metrics that matter. Most social networks measure
tangible metrics, but not those that actually contribute to the growth of the
community or its vibrancy. If you have support community, maybe the metric that matters is more how quickly can users get the
information you need, instead of page views, which might tell you something,
but nothing useful or actionable.

There was an extremely compelling discussion I had with him on the Communities
being “Community Centric” vs. “User centric” – which is
about building a community for the sake of building it, versus driving the
community to build it. More on that later.

From Future of Communities: How to measure effect of communities at the macro level?

I published this in the Future of Communities a week ago.

It is absolutely facinating that some metrics can capture the
imagination of the masses very quickly. Some examples include the
airline industry (which I am told before the advent of SouthWest) did not care about cost per passenger
mile until this new metric has become a key determinant of
profitability. Similarly the Hotel industy used to care more about
Occupancy Rate until a few years ago, when RevPar became a more important number.

The question that I have been asking several experts in the community area is:
At the macro level what impact will communities and social networks have in economic terms?

I have heard several responses which I am trying to summarize below,
but to be honest I have not done a detailed assessment enough to claim
what I am sharing below is comprehensively thought out.

1.  Employee Productivity:
This is a number that a few Wall Street analysts have been tracking for
a few years. In simple terms take the total revenue, divide it by
number of employees. Since internal communities (collaboration
networks) will allow for employees to tap into the large knowledge,
contacts and networks of all the employees of the organization, we
should see a vastly improved employee productivity rate for companies
that adopt communities faster than those that dont.

2. Transaction Friction Loss:
Several prominent thinkers and experts have talked about reducing the
friction “tax” that is paid due to inefficiency of information
availability. In simple terms, I know something about X, but its
useless to me for most parts. Someone else is looking for information
about X but she knows how to monetize it. How do we connect her with me
at the quickest time to monetization? Google, as an example has reduced
the friction from lack of information (to a certain extent). Its not an
easy metric to track or come up with, but this metric should reduce
since networks will reduce the time, effort and money required to
connect each other.

3. Degree of Seperation. Degree, Network and Betweeness Centrality.“Social
network analysis [SNA] is the mapping and measuring of relationships
and flows between people, groups, organizations, animals, computers or
other information/knowledge processing entities. The nodes in the
network are the people and groups while the links show relationships or
flows between the nodes. SNA provides both a visual and a mathematical
analysis of human relationships. Management consultants use this
methodology with their business clients and call it Organizational
Network Analysis [ONA].”

Measuring how connected people are should reduce the degree of
seperation with communities from 6 to less than 4 is what many people
are expecting. The next question is:

“What is the value of reducing our degree of seperation?”

What do you think?

The trouble with best practices

I have to admit I am not the expert at a lot of things related to communities. Being someone that ran 2 before makes me a practitioner who has made many mistakes, but not all the mistakes one could make. Talking to a lot of other folks running much larger communities than the ones I did gives a very different perspective and set of new learnings.

My personal opinion is that most people make the “same variation” of mistakes multiple times, derive a different outcome each time and call it “learning from a new mistake”. Afterall we all want to think we are smart and learn from our mistakes so we dont make the “same mistake twice”.

This brings me to the point of best practices. If you dont try 3-5 new “ideas” or “concepts” with your community each week/month/year then you end up with 2 problems:

1. Waiting for someone else to try something that is a great way to engage community members or facilitate a discussion forum or foster a thriving atmosphere in a community. This gives them a heads up first mover advantage and gets their community more adoption. Before you think this does not hurt your community, remember that most individuals cannot possibly be a part of more than 3-5 communities and feel engaged in all of them. So you are competing with other communities not in your industry or area of focus.

2. Best practice implementations have nuances, which have to be tweaked over time before they become “known best practices”. That time if not on your side, will slow your community’s chances of growth.

The trouble with best practices is that a proven set of people (early adopters) have tried something and after a resonable number (lets say 30-50%) of the people have tried the same and found it works, it goes mainstream. Like Neil Patel points out in his 5 Sure fire Social Media Headline Formulas, most become “formulaic” and the new best practice becomes doing something different.

Online Communities overall are relatively new (some would venture even only 3-5 years old) so there are a few well known best things to do, but by and large, there are more unknowns and most are very confusing.

The one takeaway:Its good to follow some best practices, but dont hesitate to try that important 3-5 new things each week / month / year so you can write your own best practices.

Best practice: Internal Communities: Managing the GROWTH of DATA

There have been several discussion I have had in the last few days regarding Internal communities. Most recently I was involved in a discussion with 3 folks: One person works for a very large software provider and is in the process of trying to build an internal sales community of over 1000+ people. Another person was responsible for a marketing community (collaboration) for over 50 support personnel. The third was trying to put together a collaboration intranet (not a regular intranet) of all their employees for technical document sharing and discussion.

Regardless of size they all had two questions that were consistent across the board:
1. How do we manage data growth (document, messages, asset) to ensure that we dont have to filter through a lot of useless, outdated information before we get to the right content?
2. How do we ensure automated excerption of comments and messages so we dont have to read a lot of useless information before we get to the 3 sentences that matter?

Most people cannot read a post like this with so many comments, that its hard to understand the top 3 points for and against.

I am going to tackle the first question now and cover the other question later.

At my previous company our internal collaboration content management system was a black hole. Everyone would say “you will find it on the discussion board” as the answer to every question, but the person searching would end up:

1. Getting 3 versions of the same post each dated differently but with slight variations. E.g: A thread which tells the release date changed 3 times on 3 dates, but the last updated date would be wrong.

2. Get too many posts but still not the right information. There would be 10 replies to a question but 90% or worse all would not still answer the question at all.

3. Multiple different discussions from different authors with conflicting data – the engineering VP would put the release date as 3/15, the release manager would have another document with the date as 4/15 and the product manager would have a third date, 3/30.

1. Encourage user tagging: Without making it extremely onerous, the best way to stop data growth and conflict is at the source. Tagging is a great means to ensure that similar documents are clustered appropriately. Ensure that content publishers can put the right content upfront and tag it appropriately.

2. Setup an active purging and archiving policy. There will be multiple discussion threads and comments, and multiple statistics have proven that there is a 80% to 20% noise to signal ratio on all message boards. An active policy ensures that it is granular by type of board, time elapsed and audience type of the participants.

3. Engage first then facilitate, finally moderate: Get internal users to participate. The best killer for any community is lack of participation. Then get the party going by encouraging (positive reinforcement of good valuable content, finally moderate unnecessary (or repeated) comments. No point in having 100 people having the same answer to the question or replying “RTFM”.

What are best practices that work for you? Let us know.

The personal blog of Mukund Mohan