Executive Summary: Most conversations these days lead to “AI,” driven by FOMO and searching for a silver bullet. Yet, solutions without a clear problem create more chaos than value. Crucially, technology doesn’t move dollars directly; it matures intermediate building blocks that might move dollars. Furthermore, software psychology from low-risk desk work cannot simply translate to high-risk ecosystems. To derive tangible value, always start with real-world problems and solve them independent of technology.
Necessity is the mother of invention. Or at least it used to be. Our more recent inventions have come about because we can create them and not because we know what to do with them. Personally, I put social media in the same category as GenAI. By itself, social media was not an invention per se. It used other groundbreaking inventions to make access to other people’s interest mainstream. Model building and automations have been around for decades. The “free” chat window that everyone can access made it mainstream and we are all clamoring to monetize it somehow.
I have been on every side of technology throughout my career, starting from enterprise-scale implementations through the SaaS wave, and I am a personal adopter of GenAI capabilities, including building a client-facing autonomous diagnostic. But I also have consistently noted that solutions without a clear definition of a problem often create more chaos than value. I bet we all think about social media differently today than we did 15+ years ago.
One of the big issues is lumping a very specific novel product category of generative AI with all manner of approaches to solving problems using models or even simple business rules as we have done for decades.
As I speak with people who have decades of experience with hands-on modeling and machine learning, most of them find it frustrating that mainstream conversations bucket almost everything under “AI.” Below are two simple but foundational facts that focus on driving value from technology (including “AI”) as opposed to falling into the FOMO trap that appears prevalent in the market.
Everyone has been told that our resumes should be stacked with dollar metrics and our slide decks should cover ROI impact. As a result, companies feel an intense need to show immediate impact, specifically in dollars. When software companies build and pitch products today, there is a heavy tendency to focus almost exclusively on high-level dollar metrics.
The dollar-ROI focus, coupled with a systemic feeling that "Time is Money," is precisely why GenAI tools are pitched with a "cut unnecessary work" or "make work easier" flavor so that employees can do "more important things." But what exactly are those important things? I have yet to hear what they are.
Here is a simple thought exercise: What is the point of taking "notes" during meetings?
I have spent years doing customer and employee interviews while taking notes, and manually re-listening to those recorded conversations over and over to ensure I understand the nuances that lead to the recommendations I make. Asking disengaged questions off a pre-built interview guide doesn't lead to useful insights.
When we are building or deploying a brand-new product that is supposed to disrupt a market, we cannot simply jump straight to dollars and get there by promising to cut specific activities that may have a genuine purpose. Technology has to mature relevant building blocks first, and often, that building block is several layers removed from dollar measurements. The key word there is relevant. Not all unpleasant or time-consuming activities are unnecessary.
Remember: Technology doesn’t move dollars. Tech moves something else that might move dollars. The question we have to ask is: what is that "something else" in the middle?
Taking out actual value-creating activities (like taking notes so people aren’t sitting in meetings without paying attention) is not the goal of technology. It is easy to sell or buy software that claims "you don't need analysts to do research anymore." But what is the quality of decisions made based on "research" coming out of an undifferentiated box sold by someone else?
Over the last 20 years, most technology investment, development, and deployment happened in low-risk scenarios, specifically in desk work settings. Officework centric software technology was easier to scale and exit. Hardtech took a back seat because it required more capital investment and, frankly, impact is harder to demonstrate.
Office work is by design low risk. People think, talk, and decide on things while real-world impact happens elsewhere. The direct impact on people using such technology is never catastrophic. Mistakes are often hard to discern and thus easily overlooked.
During this officework-centric ERP-era and SaaS era, total technology spend went up by almost 10% every single year. Yet, total factor productivity in the broader economy was only up by 0.5% over that same period. We spent massive amounts of capital on software, but it didn't show up in productivity metrics. The Solow Paradox continues to plague us.
Operational effectiveness in office work settings has always been a lower priority because it is easier to ignore reams of slides and spreadsheets that sit in virtual storage compared to a room full of unsold inventory or wasted raw materials. The true value realized from office work technology has always been spurious.
The psychology born in low-risk office environments where we “start with a nifty solution and find a use for it” doesn't apply to high-risk environments like industrials, energy, or healthcare where impact is far more noticeable. Thinking in adding more features that may not create value and perpetual releases with bugs only work in low-risk environments.
For instance, an “AI Disease Identifier” in a healthcare provider setting could go-to-market with "Look at this product. Your doctors won't have to do as much work, so you can cut the number of doctors on staff, and make the hospital more profitable."
It’s a great ROI story with office-work thinking. Is an AI Disease Identifier actually better at identifying the disease or telling the doctor they don’t have to read the charts? If the latter, the likely reality is that the doctor probably will end up running to the patient’s bedsides for all manner of alerts, without having read the charts and that is operational chaos.
Always start with tangible real-world problems and solve them independent of technology. Then, we will have a better understanding of what speed, quality, or flexibility gaps a software or hardware solution can address. Regardless of setting, it’s the only path to derive value from technology; value realization just happens to be far more discernible in high-risk settings.
John Oommen is an electrical engineer turned company-level evolution architect who partners with private company investors and middle-market CEOs to mature and transform portfolio companies. He is the author of Congruence: The Playmaker's Guide to High Value Companies, creator of the underlying proprietary Congruence mindset, model, and method, and pioneer of the Playmaker Capability, which is a must-have value injection mechanism that dovetails purpose, strategy, and operations.
Over more than two decades, John has led company-level transformations, M&A integrations, and operational turnarounds across dozens of mid-market, VC and PE-backed, and Fortune 100 companies, including General Electric, Accenture, Blue Ridge Partners, and Capital One, as well as enterprise clients like Emerson, Ferguson, Cox, Avaya, and Nortek, alongside numerous mid-market companies. He specializes in addressing business-to-business market problems spanning industrial manufacturing, distribution, technology, and tech-enabled services. John holds an MBA from Northwestern University’s Kellogg School of Management, is a certified Six Sigma Black Belt, and is based in Chicago.
