In my post discussing Top-Down BI Strategy, I introduced a chart which I call the “Data Mass Value Curve”.

Value of Data vs. Amount of Data

Data Mass Value Curve

This graph identifies an exponential increase in the value of the data as a function of the amount of data available. That is to say that the more data you have available to you, the value of that data increases exponentially. But why is this true? Why does the value of the data available exponentially increase?

Why is the Increase of Data Value Exponential?

Take the example of a credit union who only has access to their core data. This credit union will only be able to provide a one-dimensional analysis of their membership. They can return simple requests like, “how many of our members have share drafts” or “how many of our members have vehicle loans”. Those questions are simple and do not provide particular insightful analysis.

Now look at the credit union that has access to both their core data and their CRM system. Questions that were once simple like “how many of our members are over the age of 65?” become “how many of our members over the age of 65 call into our call center more than once a week?”. The questions are much more insightful and the integration of data systems allows for more complex questions to be asked of the data. Instead of asking questions only about the core or only about the CRM system, you can ask questions that draws data from both systems at once.

As the number of data sources available to us increases, the number and complexity of questions we can ask of our data continues to increase. With three data sources available, we can draw analysis from one, two, or all three sources at the same time. This is the crux of the exponentially increasing value of data as the number of data sources (or amount of data) increases. This level of multi-dimensional analysis is critical to a business intelligence program that delivers significant tangible benefits to the organization.

Collect as much data as possible, right?

Your first thought might be, “let’s collect as much data as we possible can”. But there is a slight catch – the “Data Mass Value Curve” assumes a proportional level of data expertise that can properly and efficiently utilize the data available. Often, the limiting factor of an organization is the expertise with which they can utilize the data available. When the data expertise barrier is reached, the increasing amount of available data provides no additional value.

Data Mass Value Curve_Barrier_1

In this curve, the expertise barrier is reached quite early and the value of the data is stifled. The actual value of the data is a fraction of its potential given the limited expertise to utilize the data properly.

Now let’s look at an organization with a bit more data expertise available:

Data Mass Value Curve_Barrier_2

Notice that the expertise barrier is reached farther out on the curve. This organization is not able to completely leverage the amount of data available to them, but they are able to make more of an impact with their greater expertise.

The organization below has expertise that can provide nearly all the value possible:

Data Mass Value Curve_Barrier_3

This organization has the necessary resources, talent, and efforts to be able to properly utilize the data sources available to them.

How does this help me today?

This “Data Mass Value Curve” naturally is a bit abstract. However, the concepts can be applied to your BI environments. Ask yourself some simple questions:

  1. How many data sources can I access in a SQL database format?
  2. Of these data sources, how much of the data is integrated? Do you have data marts or a data warehouse in production?
  3. Do I have the necessary staff, resources, or talent to be able to effectively utilize and analyze this data?

After answering those basic questions, you can start to gauge where your organization sits on the “Data Mass Value Curve”. Once you have established your location on the curve, you can begin to roadmap a strategy to achieve proper data utilization and returned value. Depending on the answers to the questions above, you may need to hire or train SQL and BI talent. Other organizations may have a necessary level of expertise but not enough integrated or accessible data sources available to provide value. Abstractly understanding your organization’s location on the curve can help you drive your business intelligence investment in the right direction.