Case studies: artificial intelligence in the consumer goods industry

Tastewise is an AI-powered data platform that collects data from consumers' social media posts and online activity for the consumer goods and foodservice industries. In March 2023, Tastewise released TasteGPT, a generative AI conversational chatbot designed to help clients interpret and summarise insights. The new tool can make new product recommendations, advise how to target specific demographics, and recommend marketing strategies based on Tastewise's existing dataset. More importantly, TasteGPT simplifies data-driven insights by giving actionable insights, adding a layer of efficiency and speed to understanding consumer behaviour for clients that lack data literacy skills.

Adopting generative AI tools in R&D is a relatively safe investment for consumer goods and foodservice companies. Large language models (LLMs) can be pre-trained on company datasets, making them highly customisable. These models can assist researchers as co-pilots, processing data at a much faster speed than humans. They lack accuracy and fact-checking abilities, so human intelligence is still needed to filter through their suggestions and predictions. Generative AI tools should be used alongside humans for optimal results. 

Singapore-based startup Ai Palette has launched a comprehensive product innovation solution aimed at expediting the development process for food and beverage companies. The product feature called Concept Genie leverages generative AI to generate new product concepts based on prompts provided by the user.

Concept Genie aims to bridge the gap between trend discovery and concept testing in the product innovation process. By streamlining this process, it empowers brands to capitalise on growth opportunities. Ai Palette claims to accelerate the innovation process by integrating with Ai Palette's AI-powered insights platform (Foresight Engine) and its concept screening tool (Screen Winner).

Brands can leverage the platform to research new product opportunities, generate multiple concepts based on their research, and virtually screen-test their concepts using real-time consumer data. Ai Palette's technology enables companies to identify trends and needs in specific target markets, generate tailored product concepts, and save time and resources compared to traditional market research methods.  

The use of generative AI for concept generation and virtual screening of product concepts against real-time consumer data represents a growing trend in leveraging advanced technologies for efficient and effective product development in the modern business landscape.

Ai Palette's Foresight Engine boasts to have empowered packaged food companies like Nestle and Danone to identify unmet consumer needs and emerging trends by analysing images and online data in multiple languages from various sources. The sources include retail e-commerce platforms, menus, recipes, social media posts, and search engines in 24 countries.

During the Covid-19 pandemic, Kellogg utilised Ai Palette to scrutinise online content from Malaysia, the Philippines, Singapore, and Thailand in four languages, including Bahasa Melayu, Thai, Tagalog, and English, to identify trending recipes that incorporated breakfast cereal, such as crispy calamari. This data was then utilised to launch a social media campaign on Kellogg's channels, featuring recipes that incorporated their branded cereals.

The startup raised $4.4 million in Series A co-led by pi Ventures and Exfinity Venture Partners in August 2021. The funding was utilised to support its expansion into new markets and recruit talent for engineering and data science positions.

US-based restaurant tech startup Lunchbox has collaborated with domestic AI startup OpenAI to develop an AI-powered food photo creator AI to create menu photos that could convince customers to order more. It uses OpenAI's DALL-E text-to-image technology and is aimed at restaurants that need to quickly create images of food for their social media pages, websites, and menus.

Using OpenAI's DALL-E generative AI, Lunchbox aims to assist restaurants in avoiding instances where they lack updated food photographs. In the "Search" field, the user can type the dish they want to see, such as a cheeseburger with fries or a Caesar salad. Also, the user can ask Lunchbox to produce a picture of a person eating a salad. 

The creation of a food image on Lunchbox requires only this one mandatory category. The user can select a background for their food image in the optional background field that includes options like "yellow," "wood," "a cafe in Los Angeles," or "a bar in Paris." Additionally, the users can optionally choose the photographic style of their food image, such as a Polaroid, iPhone photo, or film camera. Instead of a photo, users can also create their photos as animations or paintings.

It has been discovered that food items with images receive more orders and generate higher sales in comparison to restaurants that don't. However, hiring a professional photographer to shoot food images may be a time-consuming and challenging procedure, especially for menu specials or if a company needs to test a product.

Several restaurants are already employing generative AI to create photos of their dishes for web menus, which could help develop content quickly. Lunchbox seeks to eliminate frequent stumbling blocks for digital sites, such as "coming soon" photos, photos for social media, modifier photos, and other issues. The startup intends to continue to add lessons as it develops. For instance, it will be able to identify visually appealing and clickable images so it can generate more like them.