One of the big challenges of AI teams is ensuring they deliver business values. And not only any business values but the ones that can improve business objectives. Hence investment from businesses in AI is significant (building teams, training, R&D, hiring specialists, involving domain experts, purchasing subscriptions, hardware and tools, and so on…); if AI efforts do not improve the business objectives in organizations, it will face an existential threat. So the higher the investment, the higher the expectations are.

The root cause of the problem is a gap:

“Data and analytics leaders are investing heavily in analytics, machine learning, data science, AI, and other related technologies. The belief is that such technologies can drive better decision making and thus business outcomes. However, there is a real gap between this belief and the idea about what a decision really is.

Analysis by Andrew White, Gartner

There are few approaches to cover this gap; there is the decision intelligence framework from Gartner and AI Canvas framework. They both have pros and cons and will not address the gap.

The good thing with AI Canvas is that it incorporates the business outcomes and data requirements at the project's definition level. If we look at the individual project, it creates a very good explanation of how this particular project can be implemented and what are its requirements. It is a single-page project scope to communicate and define a shared understanding of the project opportunities and requirements. However, it does not address the above gap and certainly lacks how these projects connect to organizations’ business decisions. In addition, it looks at AI field from a very narrow and data-driven lens.

On the other hand, the Decision Intelligence framework from Gartner, looks at AI from a wider lens, and it is definitely inclusive of many different techniques and technologies in the AI field. However, it lacks a practical step-by-step guide to build the successful projects and how they are related to organizational values that derive most important objectives.

What is the alternative to address the gap?

Decision-Centric Approach® is a methodical approach that combines people, processes, data, and rules to ensure organizations can make optimized, situation-aware, and customer-centric business decisions.

Decision-Centric Approach® fills the gap by being inclusive of all technologies and techniques in the AI field. It creates a map for key business decisions and their correlated business objectives holistically in the organization. Also it depicts the sub-decisions and other decision requirements that influence the key business decisions.

How does it work?

As we discussed already, business decisions are the fundamentals of organizations. Any effort in organizations is to make sure business decisions are made and executed more effectively and efficiently. Therefore, that’s why our approach is called Decision-Centric Approach® to make sure business decisions become first-class citizens or organizations.

Let’s have a look at Decision-Centric Approach® at a 10000-foot view:

DCA-with 3 Titles for AI Strategy

As illustrated above, it has 3 stages of (a) discover (b) automate (c) scale.


At the Discover stage, we try to understand organizations from business decisions' viewpoints. We should look at data and the decisions' log to build a decision audit of the organization. The goal is to:

  • To identify what business decisions are made in organizations
  • To understand how those business decisions are made and executed
  • To determine how business decisions influence each other
  • To understand the far-reach impacts of those decisions

The outcomes of this stage are blueprints of business decisions in the organization. These blueprints are graphical visual models of how business decisions influence an outcome which they are strongly correlated with a business objective.

These blueprints provide significant value to organizations. They are the source of every single improvement you may or may not need to make. They give you transparency on how decisions are made. They establish trust between all stakeholders (internal and external).

The outcomes of this stage are a set of artifacts about how decisions are made and executed in an organization. They are the blueprint of decision-making in organizations. You can stop here and at this stage. You can manage these blueprint artifacts, use them for clear and transparent communication, update them, and keep them current.

Without these blueprints, you have no way to improve any part of the decision-making of the organization and certainly no hope of understanding how decisions are executed. This is the decision management part of the Decision-Centric Approach®.


Based on the established blueprints of the previous stage, it becomes very clear on what decisions are the most valuable business decisions (i.e. key business decisions) for a specific business objective. Every key business decision are correlated with one or more business objective. In addition, these key business decisions are way too complex so you can drill down into them and understand their hierarchy and sub-decisions.

Based on your organization's technology and team competency, you can choose to automate any of those sub-decisions that influence the key business decisions. No more type 3 errors, no more using all the right data and math to solve a wrong problem.

For those sub-decisions, you can use a framework such as AI Canvas framework or the decision intelligence framework at the implementation and automation stage. You can choose the technology of your choice depending on the type of decision and its requirements. The key factor in automating is whether these decisions are repeatable and significant enough so that automation can help:

  • Increasing the organization's operational capacity
  • Reducing error rates, bias, and personal interpretation

But please note that automation will not necessarily improve the outcome of “key business decisions” and requires experimentation.


Business decisions are complex in nature, and they are multi-dimensional with far-reaching impacts. Therefore, hoping to just improve business objectives with automated decisions alone is a false hope. To improve key business decisions associated with business objectives, the monitoring, measurement, and experimentation with multiple variations of sub-decisions are the critical parts.

At the Scale stage, you will be able to

  • Integrate the outcome of automated decisions into systems and processes
  • Define, measure, and collect the KPI of the automated decisions
  • Measure cumulative impacts of the automated decisions to their key business decisions over time

Now you are ready to create experiments for automated decisions. Start doing some A/B testing and introduce the alternative to an automated decision, and check which one is actually improving the business objectives better.

All of this is possible because you have a clear picture with the business decisions' blueprints to understand and see the far-reaching impacts of business decisions.

Book a Custom Demo

First or last name is too short


Decision-Centric Approach® is a Decision Intelligence framework that provides a practical guide and steps to identify, design, implement and scale AI projects. It fills the gap of defining explicitly what business decisions are and how they influence each other. It clearly illustrates the far-reach impacts of business decisions to the business objectives and the ecosystems of partners, projects, and processes.

Learn more on Decision-Centric Approach® and how it can help your AI team and your organization.

Last updated March 1st, 2023 at 10:16 am, Published February 24th, 2023 at 10:16 am