For decades, the fundamental question of whether to "build" custom software solutions or "buy" off-the-shelf products has been a cornerstone of strategic decision-making within the technology sector. This enduring dilemma permeates all levels, from traditional IT departments managing vast internal systems to agile product teams crafting consumer-facing applications. Historically, the choice has been a complex calculus involving development costs, implementation timelines, functional limitations, ongoing maintenance, and alignment with an organization’s core competencies.
The Enduring Dilemma: Build vs. Buy in Enterprise Technology
At its heart, the build versus buy debate centers on control, cost, and strategic alignment. Opting to build a solution offers unparalleled control over functionality, integration, and future development, allowing for precise tailoring to unique business processes. However, this path demands significant upfront investment in time, resources, and expertise for creation, followed by the sustained commitment of maintenance, updates, and bug fixes. A recent report by McKinsey & Company highlighted that custom software projects often exceed initial budget and timeline estimates by an average of 30% and 20% respectively, underscoring the inherent challenges.
Conversely, buying a commercial off-the-shelf (COTS) solution, particularly from Software-as-a-Service (SaaS) providers, promises faster deployment, lower initial costs, and reduced maintenance burdens as these responsibilities fall to the vendor. Market data from Gartner indicates that the global SaaS market alone is projected to reach nearly $232 billion in 2024, reflecting the widespread adoption of these pre-built solutions. Yet, purchased solutions frequently come with inherent limitations in functionality and customization options. While vendors often offer configuration settings, deep customization to align with highly specialized workflows, especially within large enterprises, can be challenging and costly, sometimes leading to a hybrid scenario where a purchased product requires extensive custom integration or development to fill critical gaps.
The strategic imperative often dictates the direction: if a problem directly pertains to a company’s core competency – what makes it unique and competitive – the inclination is typically to build. This ensures proprietary advantage and full ownership of the intellectual property. For non-core functions, such as human resources, accounting, or basic customer relationship management, buying a proven solution is generally preferred. Nevertheless, exceptions abound, and highly specialized problems may lack viable commercial alternatives, forcing an organization to build regardless of core competency.
From IT Requests to User Empowerment: A Historical Arc
The historical landscape of software development often saw a stark division: dedicated product teams had the luxury of building bespoke solutions, while the broader business relied heavily on internal IT departments to fulfill an ever-growing list of requests. This bottleneck frequently led to long project queues and a perception of IT as a cost center rather than an enabler.
This dynamic began to shift dramatically with the advent of "user programming" – the ability for non-technical individuals to create functional applications or automate tasks without traditional coding expertise. The seminal moment arrived in 1979 with the introduction of VisiCalc, the first electronic spreadsheet program, designed for the Apple II personal computer. VisiCalc was revolutionary. For the first time, business professionals could manipulate data, perform complex calculations, and model financial scenarios by entering formulas into cells, effectively "programming" without writing a single line of traditional code. This enormously empowering tool democratized computing for a generation of non-developers, moving problem-solving capabilities closer to the business users who understood the problems best.
The legacy of VisiCalc blossomed with the rise of Microsoft Excel and similar spreadsheet applications. Today, countless millions of user-created programs, primarily in the form of formulas, macros, and basic scripting, underpin critical operations across virtually every industry worldwide. These "shadow IT" solutions, while sometimes posing governance challenges, demonstrably demonstrate the immense demand for accessible programming tools.
The evolution continued in 1991 with the release of Visual Basic (VB) by Microsoft. VB provided a graphical development environment that allowed users to build more sophisticated applications with less code, bridging the gap between simple spreadsheets and complex compiled software. It offered drag-and-drop components and an event-driven programming model, significantly lowering the barrier to entry for developing desktop applications. Many industry observers consider Visual Basic to be a foundational precursor to the modern low-code movement, enabling millions to create custom software solutions who might never have touched a traditional programming language.
In the decades leading up to the generative AI revolution, a distinct wave of low-code and no-code platforms emerged. These platforms were specifically designed to further empower non-technical users and accelerate development for professional developers alike. Tools from companies like OutSystems, Mendix, Appian, and Salesforce’s Lightning platform offered visual development environments, pre-built modules, and intuitive interfaces to construct everything from simple departmental applications to complex enterprise workflows. The primary drivers behind this wave were the increasing demand for digital transformation, the persistent shortage of skilled developers, and the need for greater agility in responding to market changes. While these tools significantly expanded the scope of what non-technical users could build, they still required an understanding of logical flows, data structures, and often, some degree of platform-specific knowledge.
Generative AI: Reshaping User Programming with Natural Language
The current era marks a profound transformation in user programming, largely driven by the advancements in Generative Artificial Intelligence (Gen AI). This new generation of tools, exemplified by platforms such as Lovable and Bolt, is shifting the paradigm once more. What distinguishes these tools is their ability to interpret and translate natural language – specifically, English – into functional code or application components. This effectively positions natural language as the new "programming language," dramatically expanding the accessibility of software creation to virtually anyone who can articulate a problem.
The implications are far-reaching. Imagine a business analyst describing a desired workflow or a marketing professional specifying a data visualization requirement in plain English, and the Gen AI tool generating the underlying application logic or interface. This leap in intuitive interaction lowers the cognitive load associated with traditional programming or even earlier low-code/no-code platforms, potentially unleashing an unprecedented wave of innovation from the non-technical workforce. While the underlying principle of "build vs. buy" remains, the "build" side is becoming significantly more accessible, blurring the lines between user and developer.
Beyond Simplification: The Unseen Complexity of Business Rules
Despite the revolutionary potential of natural language user programming, a critical misunderstanding often arises: the notion that these new tools will render traditional SaaS vendors obsolete and lead to a universal shift towards internal building. This perspective, while understandable given the excitement around Gen AI, oversimplifies the profound complexity embedded within enterprise business solutions. The reason most sophisticated business software – encompassing procurement, invoicing, payments, budgeting, forecasting, payroll, staffing, sales force automation, customer relationship management, and customer service – will not be easily replaced by user-programmed solutions lies in the intricate web of business rules and their associated business logic.
What many outside observers fail to grasp is that enterprise applications are underpinned by literally thousands of often highly complex business rules. These rules translate into millions of lines of code, forming the bedrock of how a business operates. They codify critical constraints and processes related to:
- Policy: Internal company guidelines, approval hierarchies, expenditure limits.
- Compliance: Regulatory requirements (e.g., GDPR, HIPAA, SOX, industry-specific standards), tax laws, anti-money laundering protocols.
- Security: Access controls, data encryption standards, fraud detection algorithms.
- Legal: Contractual obligations, intellectual property rights, regional legal frameworks.
- Financial: Accounting principles (GAAP, IFRS), revenue recognition rules, pricing models, discount structures, payment processing logic.
These rules are not merely suggestions; they are essential operational guardrails that ensure transactions are handled correctly, legally, and in alignment with organizational objectives.
The Crucial Role of Business Logic and Product Discovery
The challenge extends beyond merely knowing these rules; it involves the painstaking process of discerning, documenting, and accurately codifying them into executable business logic. This is an immense undertaking, even for experienced technical teams. Often, the individuals who originally defined and implemented these rules have long since moved on, and the rationale behind specific nuances can be lost to time. While some companies maintain documentation of business rules, these documents rarely capture the subtle context, historical reasons, or intricate dependencies that shaped their implementation.
This is precisely where the expertise of product managers (and their predecessors, business analysts) becomes indispensable. Their role is to engage deeply with stakeholders, understand the business domain, and meticulously uncover these hidden rules and their implications. This process of "product discovery" is paramount to defining viable solutions that not only address user needs but also adhere to the myriad operational, legal, and financial constraints. Furthermore, addressing technical debt in legacy systems often proves challenging precisely because it requires an archeological effort to extract and re-evaluate embedded business rules, determining which remain relevant.
Therefore, building a system capable of capturing, managing, and enforcing thousands of essential business rules, ensuring accurate and compliant transaction processing, represents a monumental development effort. This inherent value and complexity explain why robust, commercially available business solutions, refined over years by dedicated teams, are not easily replicated by even the most advanced user programming tools.
A New Paradigm: Integrated Solutions and AI Agents
However, the enduring value of these complex SaaS solutions does not imply stagnation. On the contrary, the advent of sophisticated user programming tools, particularly those leveraging Gen AI, signals a significant evolution rather than an outright replacement. Strong SaaS vendors are not doomed; they are, however, facing an imperative to adapt.
Historically, business solutions were designed primarily for human interaction. The future, however, envisions these solutions interacting seamlessly not only with humans but also with AI agents and new custom solutions crafted (whether through "vibe-coding" or traditional hand-coding) on top of these component services. This marks a shift towards an integrated ecosystem where specialized SaaS platforms serve as intelligent, accessible components within a larger, more flexible architecture.
Imagine an enterprise using a leading procurement SaaS platform. In the past, human users would navigate its interface to manage requisitions. In the future, an AI agent, perhaps programmed by a departmental user through natural language, could autonomously monitor inventory levels, generate purchase orders based on predefined rules, and even negotiate with preferred vendors through the procurement platform’s APIs. Simultaneously, a system integrator might build a custom dashboard using a Gen AI tool that pulls data from the procurement system, an ERP, and a supply chain management platform, providing a holistic view tailored to executive decision-making.
Enabling the Future: The Model Context Protocol
This vision of deeply integrated, intelligent, and highly customizable enterprise solutions hinges on a crucial technological enabler: a widely accepted protocol for describing business services in a way that is understandable by both humans and machines. The industry has long sought such a standard, particularly since the early days of the internet.
A significant step towards realizing this has been Anthropic’s proposal of "The Model Context Protocol (MCP)." Introduced approximately a year ago, the MCP has rapidly gained traction because it addresses a critical, long-standing architectural challenge. It aims to provide a standardized, machine-readable format for describing the capabilities, inputs, outputs, and constraints of software services. This allows AI models and other automated systems to intelligently discover, understand, and interact with various business applications without needing human intervention to interpret complex API documentation or service definitions. By standardizing how services communicate their context, the MCP facilitates the creation of sophisticated AI agents and custom workflows that can orchestrate actions across disparate enterprise systems with unprecedented efficiency and reliability.
The "Yes to Both" Future: A Hybrid Ecosystem
If these trends continue, the future of the build vs. buy dilemma will converge on a "yes to both" paradigm. Companies will continue to strategically invest in and "buy" complex, high-value component services for critical business functions from specialized SaaS vendors. These purchased components, however, will be designed with an API-first approach, built to be accessed and controlled seamlessly by both human users and intelligent software.
Some of this controlling software will manifest as AI agents, often developed by the SaaS vendors themselves to enhance their offerings, by expert systems integrators tailoring solutions for specific clients, or even by end customers leveraging user programming tools to create bespoke automations. Other custom software will take the form of highly specific workflows and applications, generated either through natural language "vibe-coding" or traditional development, that sit on top of and orchestrate actions across these purchased component services. This hybrid model promises to unlock significant organizational agility and innovation, allowing businesses to leverage the best of breed in commercial solutions while retaining the flexibility to customize and automate their unique processes.
Navigating the Age of Mainstream User Programming
While user programming has existed for decades, its impact has largely remained at the margins, often confined to individual productivity hacks or departmental "shadow IT" projects. The current generation of Gen AI-powered tools, with their natural language interfaces, is poised to bring these capabilities squarely into the mainstream. This represents a tremendously positive development for countless individuals and teams who have historically been constrained by limited IT resources and slow development cycles. It empowers business users to directly address their pain points and automate routine tasks, fostering a culture of innovation and self-sufficiency.
However, as more non-technical individuals venture beyond simple personal time-savers to create more complex business solutions, they will inevitably encounter the same challenges and learn many of the hard-won lessons of the professional product development world. The most crucial of these lessons is that the truly difficult part of software creation is rarely the act of building and delivering the solution itself. Instead, the profound challenge lies in discovering the right solution to build – understanding the problem deeply, identifying user needs, navigating complex business rules, and ensuring the proposed solution delivers genuine value while remaining viable and sustainable.
Implications and The Road Ahead
The implications of this evolving landscape are significant. For SaaS providers, it necessitates a pivot towards more open, API-driven architectures and a focus on how their platforms can serve as intelligent, composable components within a broader ecosystem, rather than monolithic, closed systems. For system integrators, it means adapting to new tools and methodologies, potentially shifting from purely custom code development to orchestrating and enhancing solutions built with AI and low-code platforms. For businesses, it opens up new avenues for innovation, potentially reducing reliance on traditional IT for certain types of solutions, but simultaneously introduces new challenges related to governance, data security, scalability, and the long-term maintenance of user-generated applications.
Ultimately, the future of enterprise software is not a simple dichotomy of build or buy. It is a sophisticated symphony where robust, commercially developed components are harmonized with intelligent AI agents and agile, custom-built solutions, all orchestrated to create a more responsive, efficient, and innovative business environment. The age of mainstream user programming is here, but its success will hinge not just on the ability to build, but on the wisdom to discover what truly needs to be built.