The Next Generation of Websites: AI, Data, and Structured Corporate Knowledge at the Core

How corporate knowledge evolves from website content into a central, independently accessible source of information.

For decades, the primary focus in developing a corporate website was on how to present information in an engaging way. With the increasing prevalence of artificial intelligence, this perspective is changing. In the future, it will no longer be solely decisive what visitors see on a website, but also how reliably the underlying information can be found, understood, and used by other digital systems. This brings to the forefront a corporate asset that has often been underestimated until now: the company’s own data.

Anyone visiting a company website today will typically find a familiar layout: information about the company, its services, products, employees, and locations, supplemented by images, contact information, and possibly an online store or a customer portal. The content is structured so that visitors can navigate through the various pages and find the information they’re looking for. Consequently, many website projects focus on design, user navigation, search engine optimization, and editorial content. This approach made sense for many years and will continue to play an important role in the future. However, it is increasingly insufficient to view a digital presence exclusively from the perspective of its human visitors.

The reason for this lies in a fundamental shift in how the Internet is used. With the development of powerful AI systems, another way to obtain information and use services has emerged alongside directly visiting a website. People can already use an AI assistant to research offers from various companies, summarize information, or suggest suitable services. In the coming years, such systems are likely to increasingly take on more complex tasks as well, such as checking availability, preparing inquiries, or initiating specific business transactions after receiving explicit authorization. This does not mean that traditional websites will disappear. Rather, it means that in the future, companies will no longer communicate with their digital environment exclusively through their visible website.

This trend is already evident in today’s Internet search. In its documentation on AI-powered search functions, Google describes how information from the existing web index is used to generate answers. While the familiar fundamentals of search engine optimization remain relevant, a system’s ability to correctly categorize information and link it to other sources is becoming increasingly important. This presents a new challenge for companies: Their information must not only be easy to read and find, but also sufficiently precise and traceable so that automated systems can interpret it correctly.

Structured corporate knowledge is derived from individual web pages

In many companies, valuable information is scattered across numerous systems and documents. Some of it is on the website, while other information is stored in product databases, membership management systems, booking systems, PDF documents, Excel spreadsheets, or internal applications. Often, the same information is maintained in multiple locations—for example, business hours, prices, areas of responsibility, or service descriptions. As long as this information is primarily searched for and read by people, certain inconsistencies can be resolved through experience or by asking for clarification. However, if the same information is increasingly processed automatically, contradictory or incomplete data can lead to significant problems.

Let’s take as an example a Swiss tourism organization that features several hundred accommodations, attractions, events, and recreational activities on its website. Today, this information is often embedded in individual web pages, each of which contains a description, images, and perhaps a link to a booking option. For a human visitor, this may already be sufficient. However, a digital travel assistant designed to find accommodations with specific characteristics and combine them with suitable activities requires much clearer information. It must be able to identify which listing belongs to which provider, where it is located, what features it offers, and whether the data is up to date. If it is also supposed to check availability or prepare a reservation, it needs access to appropriately integrated, reliable systems.

This is precisely where the difference between website content and a standalone source of information becomes apparent. While a traditional website summarizes various details on a topic in a single text, structured data management organizes individual pieces of information as clearly defined components. An accommodation listing is not simply a text with images, but a standalone data record with specific attributes and relationships to other information. An event has a uniquely identified venue, a start time, an end time, and, if applicable, an organizer. A service is linked to an organization, a location, and possibly specific requirements. As a result, a collection of web pages gradually evolves into a cohesive information system.

This idea by no means originated with today’s artificial intelligence. As early as 2009, scientists Christian Bizer, Tom Heath, and Tim Berners-Lee described in their paper “Linked Data – The Story So Far” how structured and linked data can expand the Internet beyond the mere presentation of documents. Subsequent research on so-called knowledge graphs—including work by Natasha Noy and her co-authors, as well as Aidan Hogan and his research team—demonstrates how information can be made usable for various applications through clearly defined relationships. Put simply, a knowledge graph is a representation of knowledge in which not only individual pieces of information are stored, but also the connections between them. For many companies, a complex graph database will not be necessary. However, the underlying principle—describing information unambiguously and linking it together in a meaningful way—is of great importance.

The internationally recognized FAIR principles from the field of scientific data management are also insightful in this context. They state that data should be discoverable, accessible, interoperable, and reusable. Originally developed for research data, these principles also provide valuable guidance for businesses. Accessibility by no means implies that all information must be made public. Rather, it is about organizing data in such a way that authorized individuals and systems can reliably find it and use it under clearly defined conditions.

The visible website is becoming the interface of a digital platform

. An important consequence of this development concerns the technical architecture of corporate websites. Today, content, design, and functionality are closely intertwined on many websites. Therefore, if the visible website undergoes fundamental changes or the underlying system is replaced, existing content, data structures, and interfaces often must be adapted or completely rebuilt. This requires significant effort and can lead to investments already made being called into question again after just a few years.

A forward-looking approach therefore aims to manage the actual corporate information more independently of how it is presented. The website becomes one of several possible interfaces through which the same data and services can be accessed. Behind this interface lies a central information and service layer that consolidates data from various sources or makes it available via defined interfaces. Instead of preparing information exclusively for individual web pages, it can also be used by a mobile app, a customer portal, a search function, or an AI-powered assistant.

In technical circles, this concept is often referred to as a decoupled or headless architecture. This refers to the separation between content management and content presentation. For example, a company can continue to use an established content management system while making the stored information accessible to various applications via interfaces. Both WordPress and Drupal already provide the technical foundation for this. Therefore, the key consideration is not so much the choice of a specific product as whether the information is neatly structured, maintainable over the long term, and usable independently of any single user interface.

The benefits of such an architecture are particularly evident when requirements change. For example, an organization could completely redesign its website, develop an additional mobile app, or offer its members personalized access without having to re-enter all the information. A multilingual Swiss association could centrally manage its events, publications, and services and make them available on a German-language website as well as on French- and Italian-language sites. If an event date or contact information changes, the update is made in the relevant data source and is then reflected in all connected applications.

This model is particularly interesting in light of the ongoing development of artificial intelligence. For several years now, research on so-called Retrieval-Augmented Generation (RAG) has been investigating how language models can draw on additional information sources when answering questions, rather than relying exclusively on the knowledge they acquired during training. If corporate information is reliably maintained and made accessible to relevant applications, it can also serve as a knowledge base for such systems. For example, one internal assistant could answer questions about services or club bylaws, while another explains the company’s publicly available offerings to potential customers.

However, it is important to make a key distinction: Structured data and technical interfaces do not automatically ensure that AI systems will give a company preferential treatment or reproduce its information accurately. Nor will every organization need a completely decoupled website or a complex data platform. Rather, the goal is to create a robust technical foundation tailored to the company’s specific needs—one that can be expanded as needed. The technical architecture should serve the company, not the other way around.

What This Development Means for Swiss Companies and Organizations

For Swiss SMEs, associations, clubs, and foundations, this transformation opens up interesting opportunities, but it also places new demands on the maintenance of and responsibility for information. Smaller organizations, in particular, often possess extensive expertise that has been accumulated over the past few years on websites, in documents, and in various administrative applications. This knowledge is of considerable value because it describes services, documents experiences, and forms the basis for communication with customers, members, or other stakeholders. However, it is often unclear which information is officially maintained where, how up-to-date it is, and which parts of it may be used for other purposes.

Anyone looking to further develop their digital presence should therefore start by examining their own information assets. What details are particularly important to customers or members? Where is this information currently managed? What data is being entered multiple times? Who is responsible for ensuring its accuracy, and what information could also be useful outside the website in the future? Simply answering these questions can reveal where processes can be streamlined and existing information can be put to better use. The first step toward a future-proof corporate website therefore doesn’t necessarily begin with new technology, but with a better understanding of your own data.

At the same time, the quality of information is becoming increasingly important. In a scientific review published in 2024, Maroua Masmoudi and her co-authors describe how data sets with different structures and that are isolated from one another make it difficult to use them together. The research thus highlights a problem that is all too familiar in everyday business life: While information may be available, it is not readily usable for other tasks. Those who structure their data consistently, define responsibilities, and establish clear rules for maintaining it not only improve the conditions for future AI applications but often enhance today’s workflows as well.

Of course, responsible use also includes data protection. Not all information held by a company may or should be freely accessible. Particularly when it comes to membership directories, customer data, internal documents, or personal information, a clear distinction must be made between public and protected information. The Swiss Data Protection Act requires appropriate protective measures when processing personal data; the legal requirements must also be observed in the event of any transfer of data abroad. A modern data architecture must therefore not only be able to provide information but also reliably regulate who is permitted to access it and under what conditions.

For executives and marketing managers, this leads to an important insight: When planning a new website, the discussion should no longer focus exclusively on the visible design, navigation, or individual features. Equally important is the question of how sustainably the underlying information is structured. After all, the lifespan of a specific user interface is limited, whereas well-organized corporate data can remain valuable for various applications over many years.

What Happens Behind the Scenes Is Becoming More Important

The website of the future may not look all that different to its visitors as one might initially assume. It will continue to present information, build trust, provide guidance, and make services accessible. In the best-case scenario, it will even be easier, faster, and more intuitive to use than many of today’s websites.

The key difference will increasingly lie in this. Corporate information will be less tightly tied to individual pages and will instead be available as independent, reliable, and reusable data sets. The visible website will remain the direct point of access for people, while additional interfaces will enable communication with other applications and, increasingly in the future, with intelligent AI agents.

This also changes the long-term perspective on investing in a company’s own website. It is no longer just about creating a modern digital presence, but about building a foundation of information that can evolve alongside the company and technological developments.

A good company website will remain important in the future. However, its long-term value increasingly stems from the knowledge behind it—and from the ability to make meaningful use of that knowledge, regardless of how it is currently presented.

You might also be interested in

This post is part of our article series“The Internet Is Changing. Is Your Company Ready?”
Here you’ll find our guide on the topic of corporate websites in transition: How AI is changing things and what matters most right now
 

Sources and further reading

Scientific Publications

7. Bizer, C., Heath, T., & Berners-Lee, T. (2009). Linked Data – The Story So Far. International Journal on Semantic Web and Information Systems, 5(3), 1–22. https://doi.org/10.4018/jswis.2009081901
8. Wilkinson, M. D. et al. (2016). The FAIR Guiding Principles for Scientific Data Management and Stewardship. Scientific Data, 3, 160018. https://doi.org/10.1038/sdata.2016.18
9. Noy, N. et al. (2019). Industry-scale Knowledge Graphs: Lessons and Challenges. Communications of the ACM, 62(8), 36–43. https://research.google/pubs/industry-scale-knowledge-graphs-lessons-and-challenges/
10. Lewis, P. et al. (2020). Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. Advances in Neural Information Processing Systems 33. https://arxiv.org/abs/2005.11401
11. Hogan, A. et al. (2021). Knowledge Graphs. ACM Computing Surveys, 54(4), Article 71. https://doi.org/10.1145/3447772
12. Masmoudi, M. et al. (2024). Semantic Data Integration and Querying: A Survey and Challenges. ACM Computing Surveys, 56(8), Article 209. https://doi.org/10.1145/3653317
13. World Wide Web Consortium (W3C). (2017). Data on the Web Best Practices. https://www.w3.org/TR/dwbp/
14. Google Search Central. Optimizing Your Website for Generative AI Features on Google Search. https://developers.google.com/search/docs/fundamentals/ai-optimization-guide
15. Google Search Central. Introduction to Structured Data Markup in Google Search. https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data
16. WordPress.org. REST API Handbook. https://developer.wordpress.org/rest-api/
17. Drupal.org. Getting Started with JSON API and REST – Decoupled Drupal. https://www.drupal.org/docs/develop/decoupled-drupal/getting-started-with-json-api-and-rest
18. Swiss Federal Data Protection and Information Commissioner (FDPIC). Transfer of Personal Data Abroad. https://www.edoeb.admin.ch/de/bekanntgabe-von-personendaten-ins-ausland