Big Data

Artificial Intelligence: 6 Step Solution Decomposition Process

Bill Schmarzo By Bill Schmarzo CTO, Dell EMC Services (aka “Dean of Big Data”) March 5, 2018

It’s simple. The conversation is simple because the objective is simple:

How do I become more effective at leveraging (big) data and analytics (artificial intelligence) to power my business?

Success with artificial intelligence doesn’t begin with technology, but rather the business, and more specifically the people and processes running the business. Before deploying technology, leaders should seek to understand (envision) how artificial intelligence could power a profitable business, and drive compelling customer and operational outcomes.

Collaboration with stakeholders and key constituents is critical to understanding the decisions and needs of the business. While every organization’s needs vary, there exists a consistent, transparent process that can drive a more stable and widespread adoption of artificial intelligence.

Note: throughout this blog, when I use the term “artificial intelligence,” I mean that to include other advanced analytics such as deep learning, machine learning (supervised, unsupervised, reinforcement), data mining, predictive analytics, and statistics (see Figure 1).

Figure 1: The Evolution of AI, ML and DL (Source: Nvidia)

Artificial Intelligence Solution Decomposition Process

I teach a “Solution Decomposition Process” course at the University of San Francisco as well as at other universities whenever I guest lecture (like I will be doing at NUI Galway, Ireland on March 16 from 6:00pm to 8:00pm). I also spend considerable time teaching the “Solution Decomposition Process” to executive teams to help them successfully adopt artificial intelligence into their business. Whether lecturing or meeting with a small group of executives, I always begin by first addressing a basic question:

What do I mean by success?

If success is simply adopting and deploying advanced analytics technologies, then you don’t need strategic guidance to chart that journey.

However if success means deriving and delivering business value, and becoming more effective at leveraging big data and advanced analytics to power your business – then I have the process for you (see Figure 2)!

Figure 2: Solution Decomposition Process


Figure 2 outlines the “Solution Decomposition Process” that is designed to ensure that artificial intelligence is deriving and driving new sources of business value. The power of this process is its simplicity. By staying focused on the business or operational objectives and tasks, businesses can successfully transform how they use data and analytics to produce optimal outcomes.

There are six key steps to the Solution Decomposition Process to undertake before deploying AI solutions to derive and drive business value. Let’s take a quick look at each.

Step 1: Identify and Understand Your Targeted Business Initiative

Modernizing your data center is not a business initiative. Moving to the cloud is .not a business initiative. Installing an “Analytics as a Service” platform is not a business initiative.

So what is a business initiative?

A business initiative is a senior executive mandate that seeks measurable and material financial impact on the value of the business.

Here is a checklist of the key characteristics of a business initiative:

  • Sense of urgency mandating results be delivered in 12-18 months
  • Important to success and survival of the business
  • Compelling and material financial impact (ROI)
  • Clear business executive ownership – someone on the executive team is not sleeping at night due to their concerns on this initiative
  • Analytically friendly in that customer, product, and operational insights have material impact on initiative success
  • Bounty of potential data sources to be mined for actionable insights in support of the business initiative
  • Strong CIO leadership and IT business collaboration

A thorough understanding of the business initiative is key before starting the journey, including:

  • What are the targeted financial outcomes or returns from the business initiative?
  • What are the metrics that will measure success?
  • How will this initiative impact the customer for better or worse?
  • What are the potential (and likely) impediments to success?

For example, PNC Financial Services Group’s annual report mentions the business initiative to “grow profitability through the acquisition and retention of customers and deepening relationships.” We will use this “increase customer retention/reduce customer attrition” business initiative for the rest of this exercise.

Figure 3: PNC Financial Services Group 2015 Annual Report

Step 2: Identify Your Stakeholders and Constituents

Next, identify the business stakeholders and constituents who either impact or are impacted by the targeted business initiative. This includes internal stakeholders (e.g., sales, finance, marketing, logistics, manufacturing) as well as external constituents like partners, suppliers, and don’t forget, the customers!

Start embracing some simple Design Thinking techniques. Create a single-slide persona for each stakeholder and key constituent as a way to make the key stakeholders “come to life” (see Figure 4).

Figure 4: Create Persona for Key Stakeholders


See the blog, “Design Thinking: Future-proof Yourself from AI,” for more insights about the role of design thinking, artificial intelligence and machine learning.

Step 3: Identify Key Decisions

Next, identify the decisions that the stakeholders and constituents need to make to support the targeted business initiative. Be sure to invest the time upfront to identify, validate, vet, and prioritize the decisions because: 1) not all decisions are of equal value and 2) there may be some decisions that need to be made prior to other decisions (see Figure 5).

Figure 5: Identify, Prioritize and Create Decisions Roadmap


See the blog “The #1 IOT Challenge: Use Case Identification, Validation and Prioritization” for more details on how to identify, validate, and prioritize the organization’s key decisions.

Step 4:  Identify Predictive Analytics

The next step is challenging because it requires organizations to change their mindsets with respect to how they currently leverage data and analytics. Organizations need to guide their stakeholders through a process of identifying the most important predictions that will support the targeted initiatives. This process starts by identifying the most important questions that the stakeholders are asking today in support of their key decisions.

Questions can then be converted into predictive analytics. For example, instead of asking: “What was customer attrition last month?” we want to predict: “What will customer attrition likely be next month?” See Figure 6.

Figure 6: Creating Predictive Analytics


See the blog “Business Analytics: Moving From Descriptive To Predictive Analytics” for more details on the differences between descriptive, predictive, and prescriptive analytics.

Step 5: Brainstorm Data That Might Be Better Predictors of Performance

Next, we want to collaborate with the business stakeholders and constituents to brainstorm what data they might need to make those predictions. Continuing with our “increase customer retention / reduce customer attrition” business initiative, we highlight one of top predictions:

“What will customer attrition likely be next month?”

To support the data brainstorming exercise, we would simply add the phrase “and what data might I need to make that prediction?” to the desired prediction. The results of this exercise might look like Figure 7.

Figure 7: Brainstorming Data that Might be Better Predictors of Performance


See the blog “Data Science: Identifying Variables That Might Be Better Predictors” for more details on how to brainstorm data sources that might be better predictors of performance.

Step 6:  Implement Technology

The final step – not the first step – is now to identify the architecture, systems, and technology necessary to support the business initiative. Understanding in detail the business, data, and analytic requirements helps determine what technologies are needed – and what technologies are not yet needed – as IT builds out their big data architecture and infrastructure (see Figure 8).

Figure 8: Data Lake Components


The availability of scale out architectures and cloud environments ensures storage and compute can be expanded as needed, reducing the need to overspend on technology while achieving a compelling return on investment.


There is a logical workflow in successfully adopting and tapping into the potential of artificial intelligence. Decision makers all too often put their immediate attention to the technology, but there is a host of activities to complete beforehand. If you are serious about monetizing data to drive business value, it is imperative to begin with the business – the people, customers, and stakeholders – that have different roles and responsibilities in determining the success of a business initiative. Start small, work your way outward to identify the supporting decisions and financial value – then get to the technology.

Bill Schmarzo

About Bill Schmarzo

CTO, Dell EMC Services (aka “Dean of Big Data”)

Bill Schmarzo, author of “Big Data: Understanding How Data Powers Big Business” and “Big Data MBA: Driving Business Strategies with Data Science”, is responsible for setting strategy and defining the Big Data service offerings for Dell EMC’s Big Data Practice. As a CTO within Dell EMC’s 2,000+ person consulting organization, he works with organizations to identify where and how to start their big data journeys. He’s written white papers, is an avid blogger and is a frequent speaker on the use of Big Data and data science to power an organization’s key business initiatives. He is a University of San Francisco School of Management (SOM) Executive Fellow where he teaches the “Big Data MBA” course. Bill also just completed a research paper on “Determining The Economic Value of Data”. Onalytica recently ranked Bill as #4 Big Data Influencer worldwide.

Bill has over three decades of experience in data warehousing, BI and analytics. Bill authored the Vision Workshop methodology that links an organization’s strategic business initiatives with their supporting data and analytic requirements. Bill serves on the City of San Jose’s Technology Innovation Board, and on the faculties of The Data Warehouse Institute and Strata.

Previously, Bill was vice president of Analytics at Yahoo where he was responsible for the development of Yahoo’s Advertiser and Website analytics products, including the delivery of “actionable insights” through a holistic user experience. Before that, Bill oversaw the Analytic Applications business unit at Business Objects, including the development, marketing and sales of their industry-defining analytic applications.

Bill holds a Masters Business Administration from University of Iowa and a Bachelor of Science degree in Mathematics, Computer Science and Business Administration from Coe College.

Read More

Share this Story
Join the Conversation

Our Team becomes stronger with every person who adds to the conversation. So please join the conversation. Comment on our posts and share!

Leave a Reply

Your email address will not be published. Required fields are marked *