The Hidden Problem Inside the Modern Software Stack
- 3 days ago
- 7 min read
Updated: 2 days ago
Mike Falkow is the CEO of Meritus Media, a PR and digital marketing agency based in Los Angeles. He is also known for his work as a creative director at Falkow Creative and Rogue Magazine. He's the author of the novel Desert Storm, released in 2025, and the host of the ProActive Podcast.
Most companies have more software than ever. They have systems for sales, finance, operations, production, projects, customers, analytics, messaging, and automation. On paper, this should create a more connected business. In practice, it often creates something else–a company that has to piece together its own reality from too many disconnected sources.
The CRM says one thing, the ERP says another, and the dashboard is close but not always current. The spreadsheet has the latest exception, the email thread explains what actually happened, and someone on the team knows the answer because they have learned how the business really works around the system. This is one of the most important and under-discussed problems in modern business technology. The software stack was supposed to create visibility. For many companies, it has created a new kind of uncertainty.
The problem is not a lack of software
Most growing companies do not have a shortage of tools. They have too many tools doing too many isolated jobs. A sales team adopts a CRM. Finance uses accounting or ERP software. Operations adds scheduling tools. Leadership asks for dashboards. Teams create spreadsheets to manage exceptions. Someone builds a custom app to fill a gap. Integrations are added so all of these systems can exchange information.
Each decision may make sense at the time. The problem appears later, when the company realizes its tools don't share one live understanding of the business. Customer data is updated in one place. Order status changes somewhere else. A production issue is tracked manually. A delivery date changes, but the report does not reflect it yet. A manager has to ask three people which number is right.
The result is not just inefficiency. It weakens operational confidence. Leaders are not only asking, “What does the report say?” They are also asking, “Can we trust it?”
The hidden cost of unreliable operational data
Unreliable business data rarely announces itself dramatically. It shows up as friction. Reports need to be checked before meetings. Teams spend time reconciling numbers. Employees re-enter the same information into different systems. Customers wait while someone manually checks the status. Departments disagree about the same order, job, shipment, margin, or customer issue.
Over time, this becomes normal. People stop calling it a data problem. They call it “how things work here.” But the cost is real. When a company cannot trust its operational data, decisions slow down. Accountability becomes harder. Growth creates more complexity instead of more leverage. Reporting becomes a negotiation. Automation becomes risky. AI becomes harder to deploy meaningfully because the underlying business record isn't reliable enough. This is where software stacking becomes more than an IT issue. It becomes a leadership issue.
How the stack creates several versions of the same business
The modern software stack often creates what can be called assembled software, one business represented across many systems, databases, reports, spreadsheets, and workarounds. The company is one operating entity, but the software environment behaves as if it is several. Sales has its version. Finance has its version. Operations has its version. Reporting has a delayed version. The spreadsheet is the version people trust when the main system doesn't fit the real process.
This happens because most software is built around standard workflows. But real businesses are specific. Every company has its own way of quoting, approving, scheduling, handling exceptions, serving customers, managing margins, and getting work out the door. When the software does not match that reality, the business has to choose. It can change the way it works to fit the software, wait for the vendor, build around the system, create a spreadsheet, add an integration, or commission custom software.
Each workaround may solve the immediate problem. But it also creates another place where the business can drift away from a single reliable record. That drift is where operational data breaks down.
The business should not have to translate itself
One of the clearest signs of a broken software foundation is when the business constantly has to translate itself. A manager translates the report into what is actually happening. An employee translates the system workflow into the real workflow. Finance translates operational updates into numbers it can trust. A customer service person translates between what the system says and what the customer needs to know. Leadership translates dashboards into decisions, often with caveats, follow-ups, and manual checks.
This translation layer is usually invisible in software budgets, but it is everywhere in daily operations. It lives in meetings, messages, spreadsheets, side conversations, manual fixes, and experienced employees who know where the official system falls short. The danger is that the company begins to depend on this translation layer without realizing it. The official system says one thing. The real business runs somewhere between the systems.
Artificial intelligence raises the stakes
AI is making this problem more urgent. Many companies are now asking AI to summarize, recommend, answer, automate, and act. But AI depends on the business context it can access. If the operating data is fragmented, AI inherits the fragmentation. If the dashboard is delayed, AI may answer from outdated information. If the latest exception lives in a spreadsheet, AI may miss it. If two systems disagree, AI may not know which one should govern the decision. If the real process lives outside the official system, AI may produce a technically plausible answer that does not reflect how the business actually operates.
This is why AI readiness is not just a model selection problem. It is not only about choosing the best chatbot, copilot, or agent platform. It is about the foundation underneath those tools. A business that cannot maintain reliable operational data will struggle to make AI useful beyond narrow productivity tasks. AI can accelerate work, but it can also accelerate confusion when it acts on incomplete or inconsistent information. Before leaders ask what AI can do, they should ask what the AI is being asked to stand on.
The next shift: From stacked tools to defined systems
For years, the default response to operational complexity was to add another tool. That era is reaching its limits. The next phase of business technology will likely be less about adding more software and more about creating systems that reflect how the business actually works.
This is the idea behind Business-Defined Systems. A Business-Defined System lets the business define how work needs to move, while the underlying foundation remains standardized and managed. It is not a giant application that forces every department into the same interface. It can include different applications, workflows, permissions, reports, and AI support for different roles.
The key difference is that these parts operate from one live operational foundation rather than from separate databases and reporting copies. In plain language, the business defines the work, and the system keeps the work coherent.
Where Yolm fits
Yolm is an example of this approach. Rather than adding another disconnected application to the stack, Yolm helps companies bring standard workflows, company-specific processes, reporting, permissions, and AI onto one live operational foundation.
The goal is not to force a business into a vendor-defined process. It is to let the business define how work should move, while avoiding the familiar trap of creating another separate database, another reporting copy, or another place where operational data can drift. That distinction matters because many companies need software that fits how they actually operate. But they don't want the burden of maintaining a fully custom software estate, and they don't want to keep stacking packaged tools that solve only isolated pieces of the operation. A Business-Defined System offers a different path: standardize the foundation, not the business.
What leaders should be asking
The warning signs are often easy to spot. Do reports need to be manually checked before they are trusted? Do different teams use different numbers for the same issue? Do important workflows still depend on spreadsheets? Do people re-enter the same information into multiple systems? Do customers wait while employees check several places for an answer? Does the business depend on a few key people who know what the software does not? Are AI or automation projects stalling because the systems cannot provide reliable context?
These questions are not only technical. They are strategic. They reveal whether the business is operating from a reliable foundation or from a collection of partial records held together by people, workarounds, and reconciliation.
The bottom line
The modern software stack gave companies flexibility, speed, and choice. But it also created a new challenge–too many businesses now run across disconnected tools that cannot fully agree with each other. That creates a deeper problem than software frustration. It creates uncertainty at the operating level.
When the business has to keep checking, translating, reconciling, and explaining its own data, something important has been lost. The next advantage will not belong only to companies with the most software or the newest AI tools. It will belong to companies that can build a more reliable foundation underneath the work.
A business cannot move with confidence if its systems cannot agree on what is current and reliable. AI cannot create clarity from a business foundation that is already fragmented. The future belongs to companies that stop managing around the gaps and start designing systems around the way the business actually runs.
Follow me on Instagram, LinkedIn, and visit my website for more info!
Read more from Mike Falkow
Mike Falkow, Strategist, Creative Director, and Writer
Mike Falkow is the CEO of Meritus Media, a PR and digital marketing agency in Los Angeles. He helps founders and brands turn expertise into coverage, thought leadership, and measurable growth. Previously a creative director at Falkow Creative and Rogue Magazine, he is the author of the 2025 novel Desert Storm and host of the ProActive Podcast.
