How artificial intelligence can accelerate a brand’s eco-responsibility
Blog: Capgemini CTO Blog
A study conducted by the CSA in May 2020 showed that a brand’s positive contribution to societal issues is the main criterion of brand loyalty for 57% of people in France. At a time when consumers are particularly keen for brands to prove their usefulness and commitment to social, economic, and environmental issues, this figure lines up with the fact that consumers would be indifferent if 77% of brands disappeared altogether.
Although awareness is important today and has been increasing year after year (seven out of ten people in France consume organic foods at least once a month), when it comes to sustainability, the market and expectations are still difficult to gauge. Organic, carbon neutral, recycling, recycled, zero-plastic, circular economy, green, vegan, natural products, etc. – the list of characteristics is often long and complicates the notion of eco-responsibility.
Today, there are many measures that companies can take to support their eco-responsible approach both on their operations and design phase.
The good news is that a brand, by pursuing sustainable development, also improves its economic gains. Nearly 80% of brands say that doing so increases customer loyalty and 63% report that it directly contributes to an increase in revenues.
Artificial intelligence (AI) has already reduced greenhouse gases by about 13% among manufacturers and retailers and can help brands reach 45% of their carbon reduction target by 2030. Thus, how concretely can it help brands accelerate their eco-responsible initiatives?
1. Investing in the optimized use of resources: operational excellence in global logistics
AI is involved in many stages in the value chain, from supplying materials to shipping products. Once the units are produced, the right products must be sent to the right stores while avoiding error and optimizing distribution. Inventory management, logistics, and overall processes are improved through algorithms. At the warehouse level, AI solutions have helped Amazon reduce packaging requirements by 33%, saving of more than 915,000 tons of packaging materials, which is equivalent to the elimination of 1.6 billion shipping cartons. On the shipping side, Capgemini has deployed a tool that generates a complete cost-benefit analysis, which details estimated fuel consumption, fuel costs, and CO2 emissions and identifies different scenarios to optimize delivery strategies.
Manufacturing only the number of products sold is the dream of any organization. In addition to the economic gains associated with reducing the number of unsold items, it is also a great way to lower the carbon footprint (fewer raw materials used, less energy consumed, etc.). However, for several decades sales forecasts have been made based on legacy, for example historical projections or expected sales objectives. Today, new algorithms can achieve unprecedented accuracy that translates into strong gains in economic and environmental performance.
In early 2018, Swedish multinational clothing company H&M announced that it had a stock of unsold clothing worth more than USD4 billion, highlighting the increasing complexity of sales forecasting in a context of tight-flow logistics. Since ongoing investment in AI is making it possible to better predict in-store demands based not only on past sales but also on external data such as the weather forecasts or scheduled sports or cultural events, it is possible to make much more accurate predictions at a very reliable item and store level.
Carrefour, with the help of Capgemini, is using AI to optimize inventory management and reduce waste by integrating the SAS solution into its supply chain. By collecting and processing data from stores, warehouses, and e-commerce sites, Carrefour can better anticipate demand and refine orders from suppliers. Through smarter management of its supply chain, Carrefour has reduced the number of breaks and overstock in stores as well as warehouses.
The stakes for the agri-food sector are immense; reducing food waste through AI could generate nearly USD127 billion. Specific applications include using image recognition to determine when fruit is ripe, more effectively matching food supply and demand, and improving the value of food by-products.
Retailers and manufacturers now have the tools and data to optimize their operations to improve not only their economic but also their environmental benefits. By implementing these eco-responsible measures, brands can improve their carbon footprint; less transport saves fuel, optimized stock requires lower heating costs, optimized packaging decreases the carbon weight of products, the reduction of unsold items eliminates unnecessary production and lowers the manufacturing impact (shortage of resources, water, energy, etc.)
2. Creating new eco-responsible products for a circular economy
The development of sustainable products is a challenge for companies that pursue traditional development focused solely on cost optimization. The product design phase is a major driver to prevent cycles of reuse, repair, refurbishment, and recycling of materials. Because the tools to assess the environmental impact of eco-responsible products are complex and not very user friendly, designers and marketers often have to outsource this assessment.
Data mining makes it possible to automatically recover key lifecycle data in the decision-making process. With AI, it is possible to predict not only the cost of a product but also the carbon cost that must be incorporated into the design phase in order to develop optimized scenarios – for example, sourcing local products to reduce the carbon footprint associated with transport or substituting products during the manufacturing phase.
AI can improve and accelerate the development of new products, components, and materials that are suitable for a circular economy through machine-learning-assisted iterative design processes that enable rapid prototyping and continuous testing. It can also facilitate the implementation of new economic models based on the circular economy. For example, AI makes it possible to simplify the resale process in the second-hand market. Capgemini and its partners have developed a “circular” customer journey: customers bring in used clothing and photograph the items with a store camera; the solution scans and analyzes the clothing and automatically generates a product page, listing defects, holes, stains, or scratches; it then estimates the value of the item based on brand, authenticity, description, size, condition, and materials.
Decision makers are increasingly relying on the circular economy and sustainable development. Recently, Ikea announced that it was creating a store 100% dedicated to second-hand sales. In northern Europe, a shopping mall was built to focus solely on second-hand resale. In France, La Redoute launched its second-hand online store, “La Reboucle,” and major retailers are now putting second-hand departments in their stores.
The acceleration of these changes has spread to the regulators who strive to make the ecological impact of our consumption more transparent. Much like nutritional qualities, which are evaluated with a nutri-score, a “carbon score” informing consumers about the environmental impact of the products they buy and helping them in their purchasing behavior is becoming the new normal.
We can no longer afford to evaluate companies based solely on extra-financial criteria. Instead, the carbon standard must be integrated into the entire value chain to allow consumers to choose not only its brand but also the most respectful product on the planet.
The “Yuka of CO2” is becoming widespread. By not valuing environmental data to make the carbon footprint of its products transparent, companies risk being noted on non-affiliated channels, beyond the control of the company. It is therefore up to the brands to use the latest technologies to activate the right levers and assert themselves as resolutely eco-responsible entities.