I am pleased to be back for my sixth Enterprise 2.0 conference in Boston. Here is a link to a summary of last year’s notes. This is another of my notes for this year. There will be more to follow. I attended session - Got Numbers? Community Metrics and Analysis led by Jillian Bejtlich, Community Manager, Autodesk. Here is the session description for the workshop:
“For anyone working in community management, you’re well aware it’s not all about sitting on social networks all day chatting it up. Users and sometimes even our own co-workers are shocked to find out we are the ones pulling massive reports and trying to make sense of millions of data points. Who knew community management was so focused on mathematics?
As a member of marketing, communications, support, or whoever the community is run by, you’re probably accustom to being asked for a variety of metrics and ways to prove success, profitability, and efficiency. But where in the world do you start? This session will help guide you through some of the actual practices of putting numbers to work. We’ll go through some simple ways of creating valuable and easy to understand analyses, how to find something worthwhile in the massive data files, and ways of sharing your mathematical discoveries with others in easy and comprehensive manners.”
Jillian said numbers are your friends, but data analysts are needed. There will be 190,000 data analysts needed by 2018. Community managers need to have analytical skills. Jillian realized she needed to develop her analytical skills as a community manager. She said that many questions that managers are asked are actually numbers questions. You need to know where to start.
Jillian said she came from an engineering background where you are taught to think. She asked how you solve a community problem using techniques from a physics problem. There is known and unknown information and formulas to help us get from start to finish. For example, what if you are asked is the community is fast enough in its responsiveness. You need to ask a few more questions to clarify the question. Who is the result for? Who is the question about? Who are the players in the formula? Who can help me find the answers I need? Answer these questions first before you start to create the required report.
Then ask when is the answer due? Is there a relevant date range? Will this report be recurring? Is this a one off the situation is quite different. If it is recurring be sure to notice lessons learned for the next iteration.
Then ask “how” questions. How will I get my data? How will I share my data? How will I verify my results? Look to see how accessible the data is and will it be available on an ongoing basis. Be sure to be able to share the raw data.
Then ask the “what” questions. What are my constants? What don’t I know? What is the end goal? What could go wrong and slow me down? What is the best possible outcome?
After asking the questions fill in the knowns, determine the assumptions, and determine the unknowns. So after the data collection what do you do next? You want to consolidate it. Narrow down the essential data and ignore the rest for the moment if related questions show up. Then solve the assigned question. Be sure to double check for mistakes and consistency. They do happen.
Now create the report. Use visualizations and graphics. Then think before you share. Keep in mind who you are sharing the data with. Is there sensitive data? Will someone be doing further analysis? Is it a written report or a presentation? Will you get follow up calls on additional questions? Anticipate possible questions.
Four major take ways: First, before you start, make sure you know what your goals are. This will save a lot of time, effort, and embarrassment. Second, listen to your brain. If something catches your eye note it. It may be important. It may be useful later. Remember nine out of ten times your first guess is your right answer. Third, chill out when chaos breaks out. Be focused on the big picture and do not freak out over little stuff. There will be anticipated things. Fourth, focus on one metric at a time.
This was a really clear and useful presentation.
Comments