Most enterprises didn’t set out to collect a pile of disconnected apps. It just happened, one purchase at a time, until the IT portfolio became more of a liability than an asset. Every new tool solves part of a problem for one department while creating proprietary data and unintended consequences for another. Yet the organizations pulling ahead, whether small, fast companies or giants like Amazon and Walmart, share a deliberate investment in how their systems work together. No single “killer” app will change that. To drive actual ROI, you have to stop buying the next “killer” app and start building IT as an economic engine that turns data into decisions and decisions into dollars. That is IT business value.
This article shows you how. I’ll first look at why the killer-app habit isn’t paying off and why our systems of record were never designed to be rapid, decision engines. Then I’ll lay out what an IT economic engine looks like and why multi-agent AI makes it urgent now. I’ll also show how to measure IT by business outcomes and point to examples worth studying. I close with a six-question self-check so you can see where you stand. If your answers bother you, that’s a good sign. Closing the gap starts with how you think about IT business value, not with the next app purchase.
1. The Killer-App Habit: Why More IT Tools Aren’t Adding Up
Most enterprises don’t set out to collect apps. It just happens one purchase at a time. Today, I constantly see organizations buying every new automation and AI tool that they perceive meets their needs, expecting these disjointed purchases to produce a cohesive IT powerhouse. Instead, they end up with a digital Frankenstein’s monster: a mix of legacy bots, siloed enterprise software and rogue AI experiments that barely talk to each other. Each tool creates more chaos, solving a problem for one department and unintended consequences for another. Moreover, each app creates its own proprietary data with its own version of the truth. At the same time, vendors keep adding AI, analytics and decision tools to their software suites. The result: a business owns ten or more disjointed application silos, each claiming to offer insights, and still cannot generate a straight answer for a cross-functional question.
No single “killer” app is going to transform a supply chain, or any other industry for that matter. The organizations pulling ahead are either small, fast companies or giants like Amazon and Walmart. What they share is a deliberate investment in how their systems work together. To drive actual ROI, you must stop buying the next “killer” app and start building a business architecture. Information technology is complex, and this IT patchwork will no doubt continue to compound as we incorporate AI into our business workflows. From my perspective, enterprise automation consists of five layers: personal assistants, function-specific systems like ERP and CRM, an integration layer that connects them, task-based bots, and now, multi-agentic AI for business workflows. Each layer has a job within the tech stack, and there will never be one monolithic “killer” app that will buy your way into a winning IT strategy at any level, much less the whole. For a better understanding of these layers of enterprise automation, see my article, The Five Layers of Modern Enterprise Automation: From Back-Office Cost-Cutting to the Digital Workforce.
2. The Truth About Enterprise Software: Systems of Record, Not Decision Engines
We also need to understand why more apps and more features don’t necessarily add up to more business value. This is especially true about enterprise software like ERPs, CRMs, TMS and WMS. At their best, they are a system of record for businesses, excelling at tracking what happened, storing master data and maintaining operational consistency. At their worst, they are bloated with “features” and short on insights. Moreover, these systems do not share data with each other well. Can any of these systems really handle both day-to-day transaction processing and complex analytics? In my experience, the answer is no. It’s like asking a sumo wrestler to win a marathon.
The biggest problem is that each system sits in its own functional silo. In my research, I group enterprise software into six types: execution, visibility, planning, analytics, backend business systems and automation. Each holds valuable data, but CEOs, COOs and CFOs need a cross-functional view of the business. A warehouse system optimizes space and a transportation system tracks shipments, yet neither was built to tell the COO or CFO what a sudden demand shift will do to service levels and margins next week. These systems are needed, but over the decades, relying on them for decision-making, businesses have slipped into siloed, application-centric thinking. As a matter of habit, we treat the killer app as our most critical asset and its data as a by-product, not a strategic asset. Worse, when we seek to upgrade our IT, instead of fixing the root problem, we go searching for the next “killer” app. No wonder we receive so few insights from our enterprise systems and their siloed data.
For a full breakdown of the six system types and where each falls short, see my article, Agile Supply Chain Decision-Making: First You Need to Know The Truth About Enterprise Software.
3. IT as an Economic Engine: Sync Data, Process Information, Act
So what does the alternative look like? Instead of a collection of separate applications, IT acts as one continuous engine that syncs data, processes information and accelerates action. I always tell my clients that data sitting idle in a silo is a liability, not an asset. The best model I know for this IT-powered economic engine is Colonel John Boyd’s OODA loop (Observe, Orient, Decide, Act), a high-velocity decision framework forged in aerial combat. Syncing data feeds Observe and Orient, giving leaders and AI agents timely, relevant information on demand, not on a reporting schedule. Processing information powers Orient and Decide, moving from diagnostic to predictive and prescriptive analytics fast enough to answer an executive’s question while it still matters. Acting closes the loop: when a supplier misses a ship date, the engine doesn’t just flag it on a dashboard. It identifies the affected orders, weighs the options against cost and service, and either executes the best one or puts it in front of the right person to approve.
Each pass through the loop makes the organization faster and smarter, and this is where the IT economic engine earns its keep. In many companies, executives from planning, operations and finance argue over whose numbers are correct and fall back on Excel to reconcile the data. An IT-powered engine that runs the loop on shared data fixes that at the source. It delivers the right insight at the moment of decision, on the decision-maker’s schedule, not the IT department’s. This is where IT business value becomes visible to the CFO, in numbers everyone already agrees on. For more on each phase of the loop, see my article, The OODA Loop for Agile Businesses: Delivering High-Velocity Decisions Through Continuous Intelligence.
4. Why Now: Multi-Agent AI Turns IT into a Business Driver
Why does this matter more now? The answer is agentic AI. For too long, IT has been the department you call when a laptop crashes. That era is over. IT is moving beyond basic software support to deploying multi-agent AI workflows, where specialized agents work together to carry out complex, multi-step work. I call this the cognitive assembly line, where humans act as line managers overseeing AI-powered workflows. IT is no longer just a cost center. It is becoming a driver of business value, an economic engine for the enterprise.
Here is the catch. AI agent insights are only as good as the data they are fed. If you add AI agents to your collection of enterprise apps, you get faster, disjointed information silos. AI will do no better than your existing software without access to the right data at the right time. An agent with only access to transportation data can reroute a truck, but it can’t weigh that choice against inventory and cost unless the integration layer underneath gives it that view. For more on this shift, from break-fix support to the new IT economic engine, see my article, The IT Shift to Multi-Agent AI Workflows: From Tech Support to AI-Powered Economic Driver.
5. Measure IT Business Value by Outcomes, Not Features or Uptime
If IT is going to shift from a break-fix operation to an economic engine, it has to be measured against the business value it generates. In the past, IT executives focused on server uptime and ticket resolution times. Those numbers still matter, but they say nothing about IT’s business value. A collection of apps is judged by its features. An economic engine is judged by what it returns.
a. From Uptime to Outcome-Based IT Metrics
I advocate measuring IT by the tangible business value its systems generate, such as revenue accelerated, operational costs cut or customer satisfaction improved. When you judge IT investments by the outcomes of the workflows they power, the conversation with executive leadership shifts from costs to outcomes. The table below shows the contrast.
Comparing Mindsets: Killer Apps vs IT Economic Engine
| Killer-App Thinking | Economic-Engine Thinking | |
| What you buy | The next tool with the best demo | Capabilities that connect data, decisions and action |
| Critical Data | A by-product locked in each app | A strategic asset, ready on demand |
| Who it serves | One function, such as warehouse, transportation or planning | Planning, operations and finance together, up to the CEO, COO and CFO |
| What IT is | Support and break-fix, a cost center | A business driver that runs the decision workflows |
| How you measure it | Features, uptime, tickets closed | Time-to-value, cost per outcome, revenue attributed |
| AI’s role | A chat box on top of a silo | Agents working end to end on clean, shared data |
b. Measuring IT as an Economic Engine
Now, let’s look at how we can measure IT as an economic engine. Here are five metrics for measuring IT business value.
Five Metrics to Measure IT Business Value
- Time-to-Value (TTV). Measure how quickly a workflow turns a signal into a decision, such as the time from a supplier’s late-shipment notice to an approved recovery plan.
- Task Automation Rate. Instead of counting closed tickets, track the share of workflows executed without human intervention, such as routine replenishment orders released without anyone touching them.
- Error Remediated Value. Put a dollar figure on problems caught before they hit operations, such as freight invoice errors or duplicate purchase orders stopped before payment.
- Cost-Effectiveness Rate (Computing, Financial, Human). Compare the cost of compute, subscriptions and oversight against the labor and overhead previously needed for the same output, such as the cost per carrier booking before and after automation.
- Revenue Growth Attribution. Track how directly a workflow drives top-line results, such as orders saved when an agent finds stock at another location and re-promises delivery instead of letting the order cancel.
For more on measuring IT as an economic engine, see section 4 of my article, The IT Shift to Multi-Agent AI Workflows. Also, the same outcome-based standard should apply to your technology partners: look for those who own outcomes instead of billing hours. For more on this, see my article, The AI Value Multiplier: The Winning Traits of IT Vendors for Scaling AI.
6. What It Takes to Deliver IT Business Value: Examples and Attributes.
It helps to see what good looks like, both in companies that have done it and in the attributes that make it work.
a. Amazon and Walmart: Linked Decisions on Shared Data
Not many companies, from public information, have shown that they’ve turned their IT into an economic engine. Amazon and Walmart are two that have. According to Amazon Science, Amazon’s Supply Chain Optimization Technologies (SCOT) team “has built one of the largest and most sophisticated automated decision-making systems in the world.” It forecasts demand, decides which products to stock and in what quantities, allocates stock to fulfillment centers and computes the delivery promises customers see. These aren’t separate apps making separate decisions. They are linked decisions running on shared data, and the same article credits the move to full automation with “improving customer experience outcomes like in-stock rates while reducing costs.”
Walmart tells a similar story. In its July 2025 update, Walmart’s U.S. Supply Chain Playbook Goes Global, Walmart describes its supply chain as “increasingly unified — not just by mission, but by AI and machine learning (ML).” Its international CTO talks about “creating systems that turn real-time signals into real-time action,” and its Enterprise Inventory view “offers a single, unified view of what’s in stock across stores, fulfillment centers, and online.” What Walmart describes is an IT economic engine that syncs information, processes it and acts at scale. Remarkably, to date too few companies have truly turned their IT into an economic engine, and therein lies the opportunity.
b. The Attributes Behind IT Business Value
What separates an organization that uses IT as an economic engine from one that just owns a collection of “killer” apps? I have written on this topic several times from a supply chain perspective, covering two key areas:
- Executive Decision Platforms: Attributes for High-Velocity Decision-Making. Here, I look at five attributes: targeted data collection, agile decision intelligence, an executive user interface (not just for planners and analysts), rapid decision-making communications and decision traceability. For a full breakdown on these attributes, see my article, An Agile Decision Platform to Empower Executives For Superior Supply Chain Performance.
- Scaling Supply Chain AI: Organizational Traits. To move beyond AI pilot projects, there are four organization-level traits that decide whether AI scales and delivers unprecedented business value. I call these traits the AI multiplier: continuous data readiness, guarded flow architecture, high-velocity decision-making and operational resilience. For more, see my article, The AI Multiplier: 4 Essential Traits for Scaling Supply Chain AI.
7. Is Your IT an Economic Engine? A Six-Question Self-Check
Now it’s your turn. Answer these six questions honestly with your leadership team. If you find that most of your answers point to siloed apps or disjointed data, you are most likely still collecting ‘killer’ apps rather than building an IT economic engine.
- Do your executive decision meetings routinely stall over whose numbers are right between planning, operations and finance?
- When a disruption hits, do your people act inside your systems, or export to Excel first?
- Can you quantify the dollar result, such as costs cut, revenue gained or errors avoided, from your last three IT investments?
- Is your critical data ready on demand for both people and AI agents, or is it stuck inside multiple enterprise applications?
- Do you measure IT by business outcomes, or by uptime and tickets closed?
- Are you adding AI agents to optimize one functional silo, or incorporating agentic AI into workflows that run across planning, operations and finance?
If your answers bother you, that’s a good sign. It means you see the gap, and closing it starts with how you think about IT business value, not with the next app purchase.
More References
- Beyond the Forecast: Forward-Thinking Analytics for Supply Chain Leaders
- Data Capitalization: Fueling AI-Powered Cognitive Workflows
- Beyond Decision Intelligence: Scaling High-Velocity AI for On-Demand Insights and Autonomous Workflows
Greetings! As a supply chain tech advisor with 30+ years of hands-on experience, I take great pleasure in providing actionable insights and solutions to industry leaders. My focus is on supply chains leveraging emerging LogTech. I zero in on tech opportunities and those critical issues that are solvable, but not well addressed, offering industry executives clear paths to resolution. I have a wide range of experience to include successfully leading the development of 100s of innovative software solutions across supply chains and delivering business intelligence (BI) solutions to 1,000s of shippers. Click here for more info.