Today’s topic is all about using data analytics to make good decisions.  As leaders we all ask ourselves, how can I make a good decision on what the data is telling me?

What Is The Challenge

It’s 2011 and we are all awash in a sea of data.  We all know that making good decisions means sorting through this data, finding the real important or “key” data, and then combining that with other “key” data to piece together the “big picture” and make decisions.  So how can we do that as marketing leaders (and in other disciplines as well)?

The Solution

I don’t know how you do it, but here’s how I do it.  I have used this model as both the leader of Microsoft’s Worldwide Partner Group and the head of Marketing & Operations for Microsoft's US Subsidiary.  I think it works well for large, macro situations like what I have found in these two roles, but I also think it works well for smaller, more tactical projects and organizations.  The key is that it helps you quickly identify what data you need to drive sound decisions, and then supports a very quick build of an underlying process (or system) to support that decision cycle.

• First, I identify three archetypal users in my organization, in this case they are marketers and business decision makers and then I work with them to determine what decisions they need to be making and the frequency with which they need to make them.  This is not an all-inclusive list of possible decisions but rather a very precise list comprised of the critical decisions they will have to make.      
• Second, I take that list of decisions and model the process that I need to execute to get data for them and the analytics they can use to support their decisions. 
• What data will my users need to support their decisions?  Be precise.  Ask yourself if you have the data in your own systems, or does the data exists somewhere else?  How do I map out getting the data?  This is the process by which you get data. 
  In my organization today, there are 36 different data sources that have data that shows me what is happening in our sales and marketing efforts.  Sales scorecard data, licensing data, marketing spend data, sales pipeline data, the list goes on.  . 
• Once the data acquisition process is written down, then I can build an analytical mashup.  This is a model, a prototype, that is literally put together quickly with a whiteboard and a piece of paper.  This is not a big investment; it is a quick build that allows you to see if the process is repeatable and if the process can get you what you need data and analytics wise.  This is my user experience design prototype and you need to ask your archetypal users if this will work for them.
• Third, I now go back to my archetypal users and make sure they can make decisions using the data and analytics I am proposing.  The hardest part of all of this is being laser focused on the archetypal user and limiting the top points of data that they need to make decisions.  The users are always going to want to throw in the kitchen sink and will always have 10 more questions.  It’s all about what the user needs in terms of precise data and analytics needed in order to make good decisions.   
• Fourth, then it’s time to build the system.  In my case, The system can be complicated or simple.  It could be a BI system that pulls from 36 different data sources using SQL BI.  It could be simple, an Excel spreadsheet with automated data pulls. 
  Here are some examples from my current organization:
• Our Scorecarding Dashboard:  This system allows me to see how our business (sales, marketing, operations) is running in near real-time.  My first level of pivot within the analytics lets me see the details of how the business is running.  My second level pivot lets me drill into the complexity of our sales segments (Small & Medium Business, Large Corporate Accounts and Enterprise Customers, Consumers, Public Sector, etc...). 
• My Marketing Campaign Tracker:  From a master screen that tracks all of my marketing campaigns across the US Subsidiary I can drill down to see how many people I am touching with each campaign, key data such as what the cost per click is on each campaign, and pivot a wide variety of factors to determine which campaigns are most effective (and vice versa).  This lets me make adjustments to a campaign WHILE THEY ARE RUNNING. 
• The last step is to never stop repeating this process. Once I have completed step four above and I have the system built and deployed, I revisit how to make it better.  At least quarterly, but monthly is even better. 
  Why?  Oftentimes as leaders we get locked into thinking that projects are a linear thing and I believe they are not.  In reality, we are always circling the optimal solution and if we don’t get it right on the first iteration, we need to do it again and each time we should get closer to the target.  Think in circles instead of lines when you are running your projects. 

Bringing data/analytics together w/your particular discipline (such as marketing) in a closed loop system can help speed your company’s internal decision cycle.  Are you in a progressive, insight driven organization?

What do you think?

 Allison