Why the "Frankenstein" Web is Dying in the Age of AI

This is the first in a series of articles about long-term survival and growth in an increasingly challenging short-term rental market. These articles are opinionated but based on significant management feedback, data, and market trends. They are designed to challenge the status quo, be educational, and question the industry and its practices at every level.

Major Shift

The short-term rental industry is undergoing a major shift in digital infrastructure. For over twenty years, platforms like WordPress have defined the “legacy web” standard.

However, a “new world” of web development is emerging, characterized by an “agentic web” — a digital environment where artificial intelligence (AI) agents actively find, evaluate, and book rentals on behalf of users. This shift is not just technical; it is a strategic response to the fragility of legacy architectures, the economic drain of OTA commissions, and the need for machine-readable data to increase conversions and understand friction points, trust issues, and more.

The Collapse of the “Legacy Commons”

For years, WordPress served as the default engine for the hospitality sector, but its reliability was deeply shaken by a high-stakes legal and personal conflict between late 2024 and 2025. The dispute between Automattic and WP Engine exposed the risks of an open-source framework subject to centralized governance decisions. For short-term rental operators, the blocking of access to essential plugin and theme repositories represented an immediate operational threat, potentially breaking the very update lifelines required for secure payment gateways and real-time availability.

Beyond governance risks, technical debt is a growing liability. Most traditional sites rely on a base theme with many third-party plugins, creating “plugin bloat” and a wide attack surface. By the end of 2025, WordPress vulnerabilities increased 42% year-over-year. Managing stacks of 20 or more plugins is no longer feasible for property managers.

The Tech-First Revolution

Traditional web models are struggling as “smart” platforms like Flataway emerge, operating as tech-first companies focused on data and LLM connectivity.

Instead of manual site builds, one example of the new direction is Flataway’s AI agent, “Staycy,” which creates PMS-integrated Next.js sites from prompts in minutes. This “vibe to code” approach enables property managers to rapidly launch direct booking sites.

Key technology advantages of this AI-native model include:

• Unified connectivity: Unlike traditional developers who “hook into” single APIs, Flataway uses a unified API layer to sync with over 30 property management system providers simultaneously.
• Seamless switching: While reconnecting a legacy site to a new PMS is often prohibitively expensive, these modern platforms enable hassle-free resyncing in minutes.
• Zero-maintenance infrastructure: These platforms are managed and updated automatically, offering worry-free security and performance compared to manual plugin management.

Designing for Living Data and “Answer Paths”

Traditional SEO, centered on “blue links,” is fading. As travel shifts toward AI-driven discovery with agents like ChatGPT or Gemini, the “front door” is a conversational interface that parses structured data for direct answers.

Legacy WordPress websites will struggle to survive this transition because their information is often buried in static or unstructured fields. To remain visible, properties must adopt “Generative Engine Optimisation” (GEO) by making every attribute machine-readable. In real estate, we’ve seen a shift toward “Q Data,” a standard that centralizes thousands of property data points, and short-term rentals could evolve similarly.

The “De-webbed” Future: MCP and A2A Transactions

The most radical perspective for 2026 and beyond is the replacement of traditional websites with the Model Context Protocol (MCP), an emerging digital standard that acts as a “universal translator” between AI agents and property backend systems. In this future, the visual frontend interface is replaced by an executable data layer accessible to large language models (LLMs).

This enables “Agent to Agent” (A2A) transactions, where a guest’s personal AI agent can negotiate terms and confirm a stay directly with the accommodation’s AI agent without a human visiting a website. Standards like WebMCP are already being previewed in browsers like Chrome to ensure direct booking channels remain robust even when travelers delegate their search to AI.

When choosing a partner for your next website, traditional criteria like aesthetic portfolio and hourly rates are obsolete. You now need a data architect, not just a “builder.”

To ensure you’re not buying a “legacy trap,” look for these three critical indicators during your evaluation:

1. Data Interoperability Over “Plugin Dexterity”

A traditional web company will tell you about which plugins they use to “connect” your site. A tech-first company will tell you about unified API layers. Ask them: “If I change my PMS in six months, how much of this website needs to be rebuilt?”

Red flag: They tell you they’ll have to recode the booking integration or charge for a new “theme setup or connectivity.”

Green flag: They use a decoupled architecture where the frontend is independent of the data source, allowing you to switch backend systems without downtime.

2. Machine Readability vs Human Readability

The goal is no longer just “rank on Google,” but “be recommended by LLMs.” Ask the developer how they handle Schema Markup and data standards.

Red flag: They focus exclusively on meta descriptions and keywords (1990s-style SEO).

Green flag: They provide a structured data map that translates every property attribute into a format that AI agents can ingest and “understand” instantly.

3. Ownership of the “Brain,” Not Just the “Body”

Avoid both a proprietary system where the vendor controls the code and hosting, and the WordPress open-source trap where you are solely responsible for security and maintenance.

The best middle ground: Look for companies offering managed Next.js or headless environments. Seek a partner with a living platform that stays updated against security threats and guarantees full export rights of your data and content.

Conclusion: The 2026 Mandate

The hospitality era is shifting from “project thinking” to “platform thinking.” Developers manually patching legacy themes are losing opportunities to faster AI-native alternatives. The path forward is clear for property managers and developers:

• Break the PMS lock-in: Move to platforms with unified API layers to ensure future flexibility.
• Adopt agentic orchestration: Leverage AI to accelerate development and reduce dependency on costly manual builds.
• Structure data for agents: Implement rich schemas and Q Data to ensure discoverability in an agent-mediated ecosystem.

The transition from legacy “Frankenstein” sites to AI-native and MCP-enabled ecosystems represents a unique opportunity for independent operators to reclaim their brand and guest relationships from the OTA-dominated past. Those who wait will find themselves invisible in a world where the human traveler no longer clicks on “ten blue links.”