For decades, a fundamental strategic question has echoed through boardrooms and IT departments across the global technology landscape: should an organization build its own solutions to address specific problems, or should it purchase existing software? This perennial "build vs. buy" conundrum, deeply ingrained in the fabric of traditional IT, now faces unprecedented transformation driven by the advent of generative artificial intelligence (AI) and sophisticated new integration protocols. What was once a largely binary choice, or at best a complex hybrid, is rapidly evolving into a dynamic, interconnected ecosystem where bespoke creation and off-the-shelf acquisition coexist and, crucially, collaborate in novel ways.
The decision-making process for this dilemma has historically hinged on a core principle: whether the problem to be solved lies within the company’s core competency. If the solution directly contributes to the unique value proposition or competitive advantage of the business, building has often been the preferred route, allowing for complete control, tailor-made functionality, and intellectual property retention. Conversely, for non-core functions—such as standard accounting, human resources, or customer relationship management—buying a commercial off-the-shelf (COTS) solution, typically in the form of Software-as-a-Service (SaaS), has been the default. This approach leverages vendor expertise, distributes development costs across multiple clients, and often provides faster deployment.
However, each path is fraught with its own complexities. Buying solutions, while offering speed and reduced upfront development, frequently comes with functional limitations, vendor lock-in, and the necessity for significant customization, particularly within large enterprises with intricate workflows. A 2023 survey by Flexera revealed that 89% of organizations use a multi-cloud strategy, often leading to challenges in integrating disparate SaaS solutions. Meanwhile, building in-house incurs substantial costs in development time, resource allocation, and, critically, ongoing maintenance and upgrades—a burden that can become a significant drag on IT budgets and innovation cycles. The average cost of an enterprise software development project can range from hundreds of thousands to several million dollars, with many exceeding initial estimates, according to data from various industry reports.
The Historical Arc of User Empowerment: From Spreadsheets to Low-Code
The landscape of software creation has never been static. The concept of "user programming"—enabling non-technical individuals to create functional applications—is not a recent phenomenon. Its genesis can be traced back to a pivotal moment in 1979 with the introduction of VisiCalc for the Apple II. This groundbreaking spreadsheet program was the first widespread "user-programming tool," allowing business users to manipulate data, perform calculations, and model financial scenarios without writing traditional code. VisiCalc’s profound impact democratized data analysis, shifting power from centralized IT departments to individual users and laying the groundwork for what would become an enduring trend. Its success quickly spurred competitors like Lotus 1-2-3 and Microsoft Excel, which today run countless millions of user-created programs, primarily in the form of formulas and macros, across virtually every company worldwide.
This wave of user empowerment continued to expand. In 1991, Microsoft released Visual Basic, a revolutionary development environment that provided a graphical interface for building applications. It abstracted much of the complexity of traditional programming, allowing developers to drag and drop elements and write code in an event-driven model. Visual Basic can be seen as an early progenitor of the "low-code" movement, enabling millions to create custom software solutions with significantly reduced coding effort.
The subsequent decades witnessed a steady acceleration of this trend with the rise of dedicated low-code and no-code (LCNC) platforms. These tools, such as Appian, OutSystems, and Salesforce’s Lightning Platform, offer visual development environments, pre-built modules, and drag-and-drop interfaces that allow business analysts, citizen developers, and even non-technical users to build sophisticated applications, automate workflows, and integrate systems without writing extensive code. Market research firm Gartner projects that by 2025, 70% of new applications developed by enterprises will use low-code or no-code technologies, a dramatic increase from less than 25% in 2020. This pre-Generative AI surge in LCNC adoption underscored a growing demand for agility and a desire to reduce reliance on overburdened IT departments.
Generative AI: The New Frontier of User Programming
Today, the evolution of user programming has reached an unprecedented inflection point with the advent of Generative AI. This new generation of tools, exemplified by platforms like Lovable and Bolt, is fundamentally redefining the interaction between humans and software creation. The most significant innovation is the shift in the "programming language" itself: it is now natural language, primarily English. This paradigm, sometimes referred to as "vibe coding," allows users to describe their desired application or functionality in plain text, and the AI generates the underlying code, interfaces, and logic.
This development drastically lowers the barrier to entry, potentially opening up application development to nearly anyone with a problem to solve and the ability to articulate it clearly. The types of applications that can be built are far less constrained than with previous user-programming tools, extending beyond simple data manipulation or form-based interfaces to potentially complex, multi-functional solutions. Industry experts are hailing this as a democratizing force, empowering an even wider cohort of "citizen developers" to address niche business needs and accelerate digital transformation. Data from Accenture’s 2024 Technology Vision report highlights that 98% of surveyed executives believe that generative AI will open up new pathways to enterprise innovation, many of which involve increased user-led solution development.
The Unyielding Complexity of Business Rules: Why SaaS Endures
Despite the undeniable power of these new user-programming tools, the widespread notion that "everyone will build, and nobody will buy," leading to the demise of SaaS vendors, is almost certainly an oversimplification. This perspective overlooks a critical, often underestimated, dimension of enterprise software: the intricate web of business rules and associated business logic that underpins virtually every core business function.
Consider mission-critical business software for procurement, invoicing, payments, budgeting, forecasting, payroll, staffing, sales force automation, customer relationship management (CRM), or customer service. Behind the seemingly straightforward user interfaces of these systems lie literally thousands of often-complex business rules and millions of lines of embedded business logic. These rules are not arbitrary; they are meticulously crafted to capture constraints and processes related to policy, regulatory compliance (e.g., GDPR, HIPAA, SOX), security protocols, legal frameworks, financial regulations, pricing models, supply chain dynamics, and more.
The discernment and codification of these rules represent an enormous undertaking, demanding considerable time, expertise, and cross-functional collaboration. The vast majority of non-technical people attempting to create business applications via user programming have little to no awareness of these deeply embedded rules. Even experienced technical developers often struggle to unravel them, as they are frequently intertwined within legacy codebases. The institutional knowledge behind these rules often resides with individuals who have long since moved on, and while documentation may occasionally exist, it rarely captures the nuanced rationale or the specific context behind each rule’s implementation. This crucial knowledge—the understanding of what constitutes a "viable solution" within a given business context—is precisely what product managers and business analysts spend years acquiring and refining. It is also why addressing technical debt in legacy systems is so challenging; it necessitates a forensic examination to identify which business rules remain applicable and which need to be updated or retired.
Therefore, building a robust, compliant, and secure system that accurately captures, manages, and enforces these thousands of essential business rules, ensuring transactions are handled precisely as required, remains a monumental task. This inherent complexity and the specialized knowledge required to navigate it are precisely why strong SaaS vendors, who have invested heavily in codifying and maintaining these rule sets, are unlikely to disappear. Instead, their value proposition is poised to evolve.
The Model Context Protocol: Enabling a New Era of Integration
While the core value of sophisticated SaaS solutions remains, significant changes are indeed on the horizon. The future of enterprise software is moving beyond solutions designed solely for human interaction. Increasingly, these platforms will need to be consumed and controlled not just by people, but also by AI agents and new custom solutions crafted (whether "vibe-coded" or hand-coded) to orchestrate and extend their capabilities.
The major enabler for this shift is a technological advancement that has been sorely needed since the early days of the internet but is only now gaining critical momentum: a widely accepted protocol for describing business services in a machine-readable and machine-understandable format. Anthropic’s proposal of "The Model Context Protocol (MCP)" about a year ago has rapidly gained traction because it addresses a crucial, long-standing architectural problem.
Traditional APIs allow systems to communicate, but they often require developers to read extensive documentation to understand the service’s capabilities, inputs, and outputs. A protocol like MCP aims to create a standardized, semantic description of business services, allowing AI agents and other software to autonomously understand how to interact with and utilize these services. This means that a financial system, for instance, could publish a machine-readable description of its invoicing service, including all relevant business rules, parameters, and expected outcomes. An AI agent could then, without explicit human programming for each integration, understand how to generate an invoice, apply specific payment terms, or check compliance against a set of policies described within the protocol.
The Hybrid Future: "Yes to Both"
The implication of this convergence—powerful user programming and intelligent, protocol-driven integration—is a future where the "build vs. buy" dichotomy largely dissolves into a sophisticated "yes to both" paradigm. Companies will continue to strategically buy complex and valuable component services from specialized SaaS vendors for critical parts of their business. These purchased components will, however, be fundamentally different in their design. They will be engineered to be accessed, controlled, and integrated seamlessly by both human users and sophisticated software agents.
Some of this orchestrating software will be AI agents developed by the SaaS vendors themselves, offering enhanced automation and intelligent features layered on their core offerings. Other agents will be created by systems integrators, building bespoke solutions for large enterprises that stitch together multiple vendor services into coherent, optimized workflows. Crucially, a growing portion will also be customer-defined workflows and applications, generated by non-technical users leveraging generative AI tools and "vibe coding" to tailor solutions precisely to their unique operational needs, all built on top of the robust, rule-bound foundation provided by the purchased SaaS components. This creates a flexible architecture where core functionalities are bought and maintained by experts, while customization and innovative extensions are built with unprecedented agility by internal teams.
Broader Implications and the Enduring Challenge of Discovery
This shift carries profound implications across the enterprise. For IT departments, their role will evolve from primary builders and maintainers of custom monolithic applications to orchestrators, integrators, and, critically, governors of this hybrid ecosystem. New skill sets will be required, focusing on API management, data governance, AI ethics, and ensuring the security and compliance of user-generated solutions interacting with core systems.
SaaS vendors, in turn, will face increasing pressure to adopt open, API-first, and protocol-driven architectures. Their competitive advantage will shift further towards the robustness of their core business logic, the intelligence of their embedded AI agents, and their ability to integrate seamlessly into complex enterprise landscapes. Those who cling to closed, monolithic systems risk being bypassed by more adaptable competitors.
For business users, this new era of user programming brings unprecedented empowerment. The ability to quickly prototype and deploy solutions for specific pain points will accelerate innovation and responsiveness. However, this empowerment comes with a significant responsibility. As more non-technical individuals aspire to create solutions beyond simple personal time-savers, they will invariably confront the same challenges that product professionals have grappled with for decades. The most important lesson, often learned the hard way, is that the technological act of "building and delivering the solution" is rarely the hardest part. The true challenge, the most critical skill, remains "discovering the right solution to build" in the first place. This requires deep empathy for user needs, a thorough understanding of underlying business processes and rules, rigorous validation, and an iterative approach to problem-solving. Generative AI simplifies creation, but it does not inherently simplify discovery.
In conclusion, the enduring "build vs. buy" debate is not nearing an end, but rather a sophisticated metamorphosis. Generative AI is democratizing creation, while advanced protocols like MCP are enabling unprecedented interoperability. The future is not one of exclusion, but of strategic symbiosis: companies will continue to invest in robust, rule-compliant SaaS components, while simultaneously leveraging AI-powered user programming to build agile, customized layers on top, orchestrated by intelligent agents. This hybrid model promises to unlock new levels of efficiency, innovation, and responsiveness, fundamentally reshaping how businesses acquire, develop, and integrate their critical software solutions.
