Perhaps you have found yourself in the following situation.
You were invited to a meeting about a new project initiative. There was no meeting agenda. Just several people talking about some loose ideas concerning a project they should undertake for their organization. When the meeting was over, no real decisions were actually made. Oh, except one decision…I almost forgot. They scheduled another meeting.
Project brainstorming and ideation is often painful to watch and even more painful to participate in. KPIs handed down from upper management are seemingly unaligned with profitable activities. Goals of the project are unclear. In short, it is a mess.
This phase of project development can be particularly frustrating when projects involve data. Project participants fall victim to the illusion that they are actually doing something worthwhile and are logically driven when data is involved. Unfortunately, this is not always the case.
Data availability compounds this problem even further. It often shapes the entire course of the project. The desired outcomes of the project become shaped by the datasets already available to the organization. In turn, executives are presented with limited options, creating an opportunity for poor business decisions to be made.
This is one of the major problems with data science today. Analysts simply want to take the available data and manipulate it this way and that. However, it is an exercise in futility if the data manipulation cannot provide valuable insights to business executives.
Every organization, which utilizes more than two employees, suffers from these same issues. This is one of the reasons why I have developed the Strategic Cartesian Framework.
The Strategic Cartesian Framework is based upon four main principles. Each of these principles (which I often refer to as axes) is named after some part of the Cartesian coordinate system. This ties these principles back to data science and therefore makes them easier to remember.
The Cartesian coordinate system works by utilizing a origin of zero while presenting us with three axes. It allows us to think in 3D.
Many times in problem formation and project ideation, we allow our existing data sources to influence the way we define the problem. This is wrong. By working “Zero Based,” we free ourselves from the constraints of our known data sources when defining the problem and project. Only then do we source the data.
We live in a world that is filled with so much data; it’s easy to believe all of it is significant. This is a critical error! The Determinative Axis states that we should only focus on that data which is truly significant and really influences the variables in our problem.
The Declarative Axis proclaims that our reports and our outcomes should be clear and concise. The entire purpose of data science is to acquire actionable knowledge about the state of things as they are and as they will be. If one’s findings cannot meet this goal, then they are in vain.
The Decisive Axis asserts that our findings should allow business leaders to select from multiple well-defined courses of action. This can only be achieved by working with a data scientist who understands business opportunities, economics, and risk.
I never approach a new project without first thinking in terms of this framework. It has allowed me to seek out more meaningful outcomes and thus produce better projects.
Due to the framework’s simplicity, I can reference it at any phase during a project.
During ideation, I might wish to know “What sort of actions is management prepared to take based on our findings?” You would be shocked how quickly that one question can scale down a project and cut away all of the unnecessary fluff.
Here is an example of some Zero-Based thinking. “What do we really want to know?” Not what can we discover with the data we already have.
Utilizing the Determinative Axis, we may need to question what data is truly significant. Utilizing mathematical equations to decide upon significant variables can be misleading at times. It may be prudent to utilize critical thinking, interviews, and other methods depending upon the situation. Wouldn’t it be a tragedy to have found that one variable in your data, which proved to be mathematically significant, only later to find out it was a coincidence or an anomaly? That occurs more often than you might think. Sometimes research needs to progress beyond just CSV files.
Yes, we have machine learning. Yes, we have neural networks. But we cannot afford to neglect common sense.
If the Strategic Cartesian Framework is something you would like to learn more about,then feel free to contact me. I would gladly present this framework to your organization.
Jonathan Adams
August 1st, 2025