?Prioritizing Business Value Over AI/ML Implementation

podcast link

In today’s technology-driven era, it's all too easy for companies to dive headfirst into implementing advanced machine learning (ML) solutions without first understanding the real business problems at hand. Rather than chasing the allure of cutting-edge technology, we must start by framing our challenges correctly and asking the tough, thought-provoking questions that lead to genuine value creation.

The Technology-Value Disconnect

Many organizations fall prey to the trap of overengineering solutions. Instead of asking, “What critical business problem are we trying to solve?” Companies often begin with the technology—only to realize later that a sophisticated ML model might have been overkill. Dr. Robin Hanson of George Mason University wisely noted,

“Most firms that think they want advanced AI/ML really just need linear regression on cleaned-up data.”

This misalignment between technology and business value not only wastes resources but also sets the stage for disappointment when expected returns fail to materialize.

Framing Business Problems Effectively

Before venturing into technological solutions, it’s crucial to frame the business problem in a way that unearths its true roots. This involves:

• Digging Deeper: Look past symptoms to identify the underlying causes.
• Challenging beliefs: question the unshakable industry assumptions.
• Seeking Direct Paths: Explore the shortest path to customer value.

A method championed by Danny Lange at Uber asks, “If we only knew ____.” This approach pushes teams to identify what key information is missing—information that could unlock new opportunities and drive true innovation.

When Does a Problem Warrant ML/AI?

Not every challenge demands a complex ML solution. Before investing in AI/ML, consider:

• Complexity Check: Is the prediction or decision sufficiently complex to justify an ML approach?
• Simplicity First: Could a structured, rule-based method or even a simple linear regression suffice?
• Successful Precedents: Are there proven cases where similar problems were effectively solved with ML?

Experts like Dr. Ronen Meiri remind us that many business challenges boil down to two primary ML methods—classification and regression—underscoring that even these should only be used when absolutely necessary.

Data Quality Considerations

The foundation of any ML solution is high-quality data. Inadequate or messy data can lead to misleading models and poor decisions. As Dr. Chengwei Wang puts it:

“Finding a good ML solution is an iterative process that involves research, trials and errors, experimenting, and talking to business experts.”

Before diving into ML, ensure that your data collection and cleaning processes are robust enough to support the intended analytical efforts.

Value-Based Approaches to Business Problems

Instead of letting technology drive your strategy, anchor your approach in value creation. Consider the “value stick” framework, which offers four strategic levers:

• Raise Prices
• Raise Customers' Willingness to Pay
• Lower Costs
• Lower Suppliers' Willingness to Sell

Focusing on these levers helps you pinpoint where the real opportunities lie, ensuring that any technological solution enhances—not replaces—the pursuit of genuine business value.

Disruptive Questions for Business Innovation

The journey to real innovation begins with asking the right questions. Rather than seeking incremental improvements, challenge the status quo with disruptive questions that spark fresh thinking. For example:

• Business Focus:
• Competitive Insight:
• Process Evaluation:
• Industry Assumptions:

These questions help illuminate the unseen opportunities and potential pitfalls that might otherwise go unnoticed.

Thought-Provoking Questions for Problem Exploration

Here’s a curated list of engaging questions to further explore and define your business challenges:

• Understanding Core Business Value:
• Clarifying Problem Definition:
• Evaluating Solution Approaches:
• Challenging implementation assumptions:

By posing these questions, you ensure that every angle is examined—making sure that any subsequent technology implementation is both justified and targeted.Prioritizing Outcomes Over Technology

The rush to adopt new technologies can lead to massive investments in infrastructure—think Hadoop and Spark—without delivering tangible returns. As Dr. Charles Martin cautions:

“Setting up massive infrastructure until you have a handle on what you want to do can lead to 6 months to a year of work without seeing any ROI.”

Starting with small, focused experiments can help validate your problem framing, identify the necessary data sources, and secure stakeholder buy-in. Only once the business impact is proven should you consider scaling up your technology initiatives.

Prioritizing Outcomes Over Technology

The rush to adopt new technologies can lead to massive investments in infrastructure—think Hadoop and Spark—without delivering tangible returns. As Dr. Charles Martin cautions:

“Setting up massive infrastructure until you have a handle on what you want to do can lead to 6 months to a year of work without seeing any ROI.”

Starting with small, focused experiments can help validate your problem framing, identify the necessary data sources, and secure stakeholder buy-in. Only once the business impact is proven should you consider scaling up your technology initiatives.

Conclusion

In the race toward digital transformation, it’s essential to remember that technology should serve business needs—not the other way around. By framing business problems effectively, prioritizing value creation, and asking disruptive questions that uncover real challenges, organizations can avoid costly missteps and unlock significant opportunities for innovation.

Rather than asking, “How can we implement ML?” Start with, “What critical business problem are we trying to solve, and how can we solve it most effectively?” This mindset not only ensures that technology investments yield meaningful returns but also paves the way for sustainable growth and long-term success.

If you found these insights valuable, consider subscribing to our newsletter for more deep dives into the intersection of business strategy and technology. Let’s start a conversation—share your thoughts and join the community of forward-thinking business leaders!

follow/connect and do share your experience in the comment