The Swedish town of Vasteras has an electricity plant powered mainly by burning about 15 tonnes of discarded clothes from a fast fashion brand each year. In Hong Kong, more than 340 tonnes of discarded clothes and textiles are landfilled daily! How did we get to the point where we produce clothes to burn as fuel or pour into landfills? A poor understanding of demand and supply and a less than optimal inventory management process is how. Some countries have started to take steps to reduce waste. France, for example, has banned fashion brands from destroying unsold items. Understanding the optimal method for inventory management is a complex area with a long history but we will review how a new generation of AI tools is automating the process of inventory management and reduce mistakes.
Inventory management can be broken down into forecasting what consumers want, which is also known as demand forecasting and counting what you have on hand. Most companies will have some combination of these methods intertwined into a process that can be fiendishly complex as an SAP integration or as simple as a pencil and paper.
When keeping track of the items you have on hand, a micro-retailer can rely on using the ‘pen and paper’ method or using a spreadsheet. As you scale, this becomes more difficult. POS companies like Lightspeed or EPOS have sophisticated inventory management systems that even allow retailers to upload products directly from vendor catalogues. Even larger retail corporations will opt to use an ERP tool, integrating it with a warehouse management system so that processes are more automated, giving a real time snapshot of which products are entering and leaving the warehouse. Any system you use will need to be attuned to the size of your company.
Although not AI, many companies can use this basic information for supply push selling. Any time you are shown the most popular items or items that are “also bought with”, you are viewing methods used to boost the sales of items that are selling well. This method can be used to give discounts to selling items (e.g. 10% off any item that has not sold in X days, in the hope it will reduce excess inventory).
Some companies have developed sophisticated services integrating these methods into full funnel remarketing campaigns to target consumers that have abandoned some items in their shopping carts in order to remind them to complete a purchase. These are very useful in any omni-channel campaign and are certainly important components of any website, but this is only half a solution.
At the core, these types of systems are counting machines and counting is not AI so it cannot deliver the other half; the demand forecasting that machine learning and deep learning can deliver. So how might AI help reduce excess inventory and at the same time help corporations reduce the impact they have on the environment? Well, demand forecasting and inventory management is mostly about optimization and optimized supply chains are efficient chains.
Long ago, Toyota taught the world that smaller inventory is less capital tied up with inefficient financing and if your company is more efficient, it will be less problematic for the environment and better for the bottom line. A win-win!
With Visual Recommendation tools like those supplied by Delvify, you gain valuable insight into your customer preferences. Firstly, we know that images can increase conversion 3x but this is usually necessary but insufficient in a world filled with crafted Instagram perfect products. So how can we take this further? Delvify combines powerful AI and machine learning to provide Visual Recommendations that boost the natural propensity to favor images. One nice image leads via an iterative search process to others and as customers make purchases, your business gains insight on which products are purchased less frequently.
Yes, the least frequently purchased items may be even more important to your bottom line! After all, an item you burn in a power plant is a total loss for the business and the environment. The goal of a recommendations system should not be to merely boost whatever it is selling, but rather to help manage the whole business.
Delvify’s machine learning helps you create a lean inventory process for your business as you decide which items to highlight and which to discount. Don’t discount your high demand items! In practice how might your super-charge this process? By working with our larger clients we can carefully craft systems that can, for example, make sure the recommendations are ranked by a sustainability score or take into account seasonal trends.
As we noted in the opening description of this blog, the environmental impact of choices made by corporations has a real-world effect. Fast fashion has created a more complicated planning process to predict consumer demand. Sustainable waste and material management is part of the 2030 Agenda for Sustainable Development. By shifting to a circular system, the fashion industry can unlock a USD 560 billion economic opportunity. Through helping retailers and brands create a more responsive and responsible business model we can do our part to be more gentle on the planet.