Search engine chatbots are here to stay – here’s what you need to know for your website

In the demo, Microsoft demonstrated how AI-powered chatbots can enhance the search experience for users. Alongside the standard search results, Bing now displays a summary of the search term. This summary may include advanced search queries, such as planning a holiday. It is important to emphasise that the results are based on up-to-date information. Whilst ChatGPT in the OpenAI demo draws on knowledge from late 2021, the implementation in Bing also takes new information into account. Microsoft does not specify which version of the GPT language model is used in the Bing search bot. However, it can be assumed that Microsoft is using an optimised version of the GPT-3 model or even a very early version of the new GPT-4 model, which is due to be launched later this year.

The summary contains information from various sources. Microsoft cites these sources directly in the text, and users can visit them to find out more.

In addition to the new search experience, Microsoft is also introducing ChatGPT into its Edge browser. Users will be given an additional toolbar that allows them to search for information within the current tab. This feature can help users search within a page. For example, if a website displays a company’s annual reports, users can ask about the revenue for a specific month, and the chatbot will search for and display the exact figures. The chatbot is also capable of writing summaries. Users can ask the chatbot to write a brief summary of the content of a page or document.

These new features are a good indication of what the future of the internet might look like. Google has also announced that it will be enhancing its search engine with improved AI-powered search later this year. This is just the latest development in the field of conversational user interfaces. Since the first forays into this field with voice assistants such as Amazon’s Alexa or Apple’s Siri – which are only able to provide very limited information – OpenAI is now unveiling its latest development in the field of artificial intelligence, marking a new milestone. Chatbots are now capable of accessing virtually unlimited sources of information and aggregating this information for users. What sounded like science fiction just ten years ago is now set to become the new normal for millions of users worldwide.

However, this development also gives rise to new challenges, both for businesses and brands and for legislators and government bodies.

For legislators and government bodies, many legal questions are becoming important: when an artificial intelligence system aggregates information from various sources, how does it ensure that it filters out false or even fabricated information and distinguishes it from facts? For brands, a realistic scenario could be that opponents feed an AI system with fake news about the brand. Without filters, this information would spread and could become a new form of character assassination. This could even prove very dangerous for societies, as the same mechanism could be exploited by terrorists or regimes to destabilise markets and societies.

In ancient Rome, the satirist Juvenal asked sarcastically: “Quis custodiet ipsos custodes?”, which roughly translates as “Who will guard the guardians?”. In April 2022, the European Union adopted the Digital Services Act, which is aimed at companies that develop digital services that may contribute to the spread of disinformation. Technology companies such as Google, Facebook, Twitter and Apple must provide further information on how their algorithms work and how they tackle disinformation on their platforms. To this end, the European Union has also established the European Centre for Algorithmic Transparency, which is tasked with monitoring the major technology companies and ensuring that they comply with all European laws on data analysis and artificial intelligence. 

Apart from the legal perspective, there are also many factors that companies and brands need to take into account now. Ever since the early days of the internet, it has always been important to focus on search engine optimisation (SEO). Most website traffic still comes from search engines such as Google or Bing.

This paradigm will change with the success of conversational user experiences such as chatbots. At present, a user must visit a website to access all the information. In future, this will no longer be the case, as conversational user experiences are designed to present all relevant information and context directly. This means that brands and companies will see fewer users visiting their websites, as most of the information will already be provided by the chatbot. It also means that brands and companies will no longer be able to control their brand narrative to the same extent as they can today.

The big question is how brands and businesses can capitalise on this trend. In our view, websites will still exist in this new normal, but conversational experiences will be the most important source of leads for websites. It is therefore important to store the data in an accessible way. Monolithic content management systems such as WordPress Typo3 and Drupal store information on a page-by-page basis. In this paradigm, everything is displayed as a web page. Often, the content is even stored as pre-rendered HTML within the content management system. Of course, monolithic systems can be extended with plugins that can store data in an object-based manner, but in their default configuration, most monolithic content management systems store content in a pre-rendered form. This makes it difficult to recompile the information for other services that do not use a page-based approach.

This is where headless content management systems such as , Storyblok or Contentful come into their own. Headless content management systems enable developers to define a data model for specific data types and make the data available via an API. Here is a brief example to help illustrate this approach:

We want to build a large recipe platform. Our platform mainly contains recipes for various dishes. Each recipe includes a list of ingredients, quantities for the ingredients, cooking steps and kitchen utensils used to prepare that dish. All these different data types are stored in a custom model called ‘Recipe’. If I want to access all the ingredients, I can retrieve them directly via the API without loading all the other information. In a monolithic system, this information would have been rendered as a long block of HTML text. In a headless content management system, the information remains uncompiled and can be combined, modified and formatted by the requesting party itself.

This API is integrated into various front-end applications. For a website, frameworks such as Next.js or Gatsby.js can generate standard web pages from the data provided. The API can also be used for mobile applications, as well as serving as a data source for other data services such as chatbots. 

This is the most likely way in which data is fed into conversational user experiences such as ChatGPT. Marketing teams need a new skill set to determine what information needs to be fed into artificial intelligence systems. Alongside the role of (SEO ), a new role for Artificial Intelligence Optimisation (AIO ) will emerge. AIO will determine what information is made available to artificial intelligence. It will also be important to monitor the presence of information within these systems and track reporting across the various platforms. AIO teams will also be responsible for monitoring whether misinformation about a company and its products is being disseminated, and for reporting this to the platforms to have the misinformation removed.

All in all, we have an exciting future ahead of us. The internet is on the cusp of the next revolution in the way users access information. Nevertheless, there is much to be done to ensure that information is based on fundamental democratic principles and that misinformation cannot spread so easily. Brands and businesses will need entirely new skills to ensure that their content remains visible in the search engines of the future. Therefore, thinking about a holistic and object-based content management system is key to future success. It is also important to prepare marketing teams for the challenges of tomorrow and to develop strategies that ensure the company’s content remains relevant to these new interactive experiences, whilst providing users with valuable and important first-hand information.