Just a short post on this but thought it’s worth sharing some thoughts around terminology. What’s in a name indeed…  It’s amazing how everything these days is referred to as business intelligence. I love that people are visualizing their data, drawing insights and creating dashboards. But whenever I present to a group of business folks on topics ranging from deriving insights from consumer networks to how to measure success, I’m sometimes surprised that their questions involve the wrong term. People might ask about analytics for their enterprise collaboration network, when actually when you are just starting out, plain old statistics might do (how many people joined, how many posts and so on). What metrics are important to your organization though is a different question.

As an example:

  • Statistics: How many posts were tagged with #safetyday
  • Analytics: What proportion of posts were by people who had had prior safety training (e.g. members of safety training group). Which posts were shared widely or had most discussion?
  • Metric of your network: Number of people participating in an organizational campaign (goal is to have over 1000 participants for example)
  • Metric of your organization (that you derive your network efficacy and measures from): Safety awareness in 100% of the employee population

From there we can go a lot further. Imagine that your organization sells safety equipment. Perhaps your organization has a Twitter account, a Facebook page, an external facing community and data from your CRM system. You run a similar #safetyday training campaign for your customers. We can use similar information as above and then:

  • Analytics: Cluster data from CRM system and posts in community site for specific accounts to judge impact of campaign
  • Predictive analytics: correlate key account data from CRM system, sentiment analysis on Twitter and time to sale to create forecasting model for when you should run a campaign to support product sales cycle
  • Discovery: Use machine learning (machine learning in its simplest form is a way of creating and developing a model to crunch data) to predict sales person effectiveness based on campaign engagement, correlate over multiple campaign data (if customer bought this one year ago, they buy this unrelated item later). This produces insights to help you fine tune campaigns but also much bigger aspects like building products and solutions people have unfulfilled needs for.

You might think that I am far out on this, when in reality I am not even scratching the surface. For years, companies ranging from Zara to Walmart have mined data for insights to create products and choice that people want to buy.

Here’s a handy dandy BI terminology reference chart that you can download and feel free to adapt for your own needs on what the terms mean (in total layman language) – I realize so much of the detail is lost in the brush strokes on this one.

I feel it’s important to understand the terms because it’s hard to understand what you are trying to measure and why you are trying to measure it and how you will measure it, without being accurate in your terminology.