Using algorithms to generate new content is not new – it has also been used in retail and e-commerce for years. Despite its impressive capabilities, for the time being ChatGPT and other similar so-called Big Language Models are still a tool in development. Currently, in addition to generating marketing content or product descriptions, such a tool – thanks to its collaboration with other applications – is able to help book flights and hotels, search and compare prices in thousands of online shops; it also provides recommendations and reservations for restaurants, summarises specialised texts, helps with writing emails, etc.

However, the future possibilities for its use in the retail, food or catering industries will be much greater. In general, tools using language models can contribute to the efficiency and effectiveness of sales and marketing activities, as well as improving the customer experience. In the future, ChatGPT could power the chatbots or virtual assistants already in use and support them in personalising customer interactions – analysing preferences and behaviour, automatically generating offers, recommending products or, in the case of customer service, answering questions, complaints and giving feedback. In addition, it could help personalise service, reduce waiting times and improve communication.

For catering, it could be used as a chatbot to interact with customers and provide information about menu items, restaurant hours, reservations and other services. It could also answer common customer questions, such as dietary restrictions or payment options. It can be integrated into the ordering process to allow customers to place orders through the chatbot interfaces. Also analyse customer feedback from reviews and social media platforms to help restaurants improve their services and offerings. Such a chatbot would work well for any mass customer service task at a distance.

How do companies in the food production, sales and catering sectors use and can use this technology?

While grocery chains are now starting to talk about using ChatGPT, some companies are already using similar tools. Unilever has developed the GPT API ‘Alex’, which filters consumer service centre emails, sifts out spam messages and prompts the company’s employees to respond to emails. In contrast, the same company’s GPT API ‘Homer’ uses an AI bot to generate product descriptions. Taking a few product details, it generates a list of similar products offered on Amazon’s platform, along with short and long descriptions that match the tone of the brand’s communications. It does this to differentiate the content of Unilever’s product communications from its competitors. Another tool reportedly supports the Hellmann’s mayonnaise brand in its efforts towards reducing food waste.

In turn, Tastewise – an AI-powered research platform for the food and beverage and catering industries – has launched TasteGPT, an AI chatbot that helps search for popular flavours, dishes, recipes, for example, and generates ready-to-go reports using conversational AI. Smaller, local food companies are also supported by ChatGPT in marketing decisions, such as names, prices, packaging and recipes for new products. In contrast, e-commerce companies looking to leverage the capabilities of ChatGPT or develop similar tools themselves are primarily large, technology-driven corporations such as Amazon and Shopify. Among retail chains, Carrefour is experimenting with ChatGPT and generative AI to answer frequently asked questions by customers, for example: how to eat healthier for less. The aforementioned Shopify, on the other hand, automatically generates product descriptions from a list of keywords in a tone chosen by retailers.

What opportunities and risks does this raise? 

ChatGPT and similar tools will enable efficient and personalised customer service, even more personalised recommendations, promotions and offers, confidently increasing customer satisfaction and loyalty.

Unfortunately, by automating repetitive and time-consuming tasks and processes with tools such as ChatGPT, more jobs will be cut, especially in customer service and marketing departments. It is true that new jobs will also be created, but there will be a problem of a structural mismatch between the skills of employees and demand. This will be a challenge for both companies and society, as the costs will of course be borne by society as a whole.

A major concern with ChatGPT is data privacy and security. Particularly when it comes to the processing of customer data, such as personal details, credit card details, booking information. There are also many unknowns regarding intellectual property rights and ethical issues, such as plagiarism, which cannot be completely eliminated. Companies are particularly concerned about the leakage of confidential competitive information. It is also important to remember that tools relying on algorithms are not perfect, which can hinder personalisation and accuracy. In addition, customers may have reservations about interacting with a machine rather than a human, especially if bots are not able to provide adequate service quality and reliability. We are already experiencing this today.

Another challenge is also the cost of such solutions and the integration of ChatGPT chatbots with existing systems and databases to provide relevant and accurate information to customers. This requires significant IT resources and expertise, as well as system compatibility.

Is this the future of the retail, food and catering industries?

There is unlikely to be any escape from any tools and processes that use artificial intelligence. Leaders will soon emerge who will also introduce new ways of working in the retail, food and catering industries, including improved customer service, new marketing processes and new functionalities. Those who do not keep up will fall behind the competition. The important issue of regulating the development and use of these tools to keep everyone safe remains.

On the other hand, it is important to remember that many companies need to ensure that they have the right quality of data with which to power such solutions and the right computing power before ‘throwing themselves’ into AI or ChatGPT. There is a lot that needs to happen both on the technology development side and the internal processes within companies to experience the true fruits of GPT and other generative AI models.