Inventory management is the cornerstone of ecommerce. It can be the difference between providing an exceptional user experience and one that leaves customers disappointed or frustrated, something retailers can’t afford.
For example, inventory management ensures advertised products are in stock; too often customers will click on a promotion to discover an item is only available in the wrong model, size or colour. The reason is the promotion is disconnected from real-time inventory levels, leading to a poor customer experience.
A product’s availability is also linked to its exposure, just as in a physical store; the outfit in the shop window will be first to sell out. Therefore, retailers need to rethink how to manage product exposure and promotions in relation to what inventory is available.
This is where artificial intelligence (AI) can play a transformational role.
In a nutshell, AI can continually examine real-time inventory and product exposure, enabling the retailer to adjust which products it promotes based on fluctuating inventory levels. This means they never end up with too much inventory and can use their valuable real estate for promoting the right product at that time.
AI is also fundamental to the idea of connected commerce, or subscription selling, which sees merchants selling products on an auto-replenishment basis. Well-known examples include meal subscription service HelloFresh, beauty box Birchbox, and shaving and grooming products Harry’s Razors.
“These services benefited from a surge in popularity recently when, for the first time in generations, consumers experienced genuine scarcity of key products,” says Charles Nicholls, senior vice president of SAP Upscale Commerce and trustee at The Carbon Community.
“Connected commerce enables the retailer to increase inventory of highly profitable products and have forward visibility of demand that comes from auto-replenishment, so they can buy more efficiently.
“Meanwhile the consumer can be assured of a continuous, regular supply of their favourite products.”
AI and the dynamic environment
Traditionally, AI analysed as much historical data as possible to try to predict future patterns. Unfortunately, this is like driving while looking in the rear-view mirror; you’re looking back over several years trying to predict how your business will be in the future. This historical model doesn’t adapt well to any massive spikes or swings in demand. And all bets are off with any Black Swan event, such as the coronavirus pandemic.
There is, however, a way to use AI more effectively. Modern machine-learning models evolve on the job, using underlying streaming data to provide the best option for maximising sales and profit for the retailer, based on real-time data.
This dynamic approach means retailers can respond quickly to unexpected events. This could be any event that causes inventory to quickly fluctuate, a product going viral on social media for example.
“With AI, it’s possible to have the bulk of these activities handled by machine intelligence. This allows for human capital to be reinvested into new and higher value activities,” says Chris Hauca, general manager of SAP Commerce Cloud.
Elsewhere, AI can take the heavy lifting out of merchandising. A merchandiser’s job is to sell what they have in stock. They must understand which products to display relative to their inventory and adjust pricing based on stock levels.
The challenge with any large brand is there are too many products to manage this effectively. Merchandisers therefore end up focusing on the top sellers or the “problem child” products. Machine-learning can instead optimise the long tail, uncovering the hidden gem product for example, one that doesn’t get much exposure, but offers a high margin and of which there is high inventory.
It can also identify products that are traffic generators or “halo” products that draw customers in. They don’t necessarily buy these products, instead clicking through and making another purchase from the brand.
“Machine-learning can find these items automatically and surface them so the retailer is showing products that are both relevant to the customer and they need to sell. It’s not personalisation, it’s not merchandising; it’s a blend of the two,” says Nicholls.
Becoming more agile
COVID-19 has accelerated the need for more agile approaches in ecommerce.
Hauca says: “The pandemic has provided additional evidence that ecommerce is a strategic asset for businesses that had it in place. In many cases this was the only sales channel available.
The pandemic has provided additional evidence that ecommerce is a strategic asset for businesses that had it in place
“Retailers are accelerating their ecommerce strategies and hardening their operations. The consensus is their digital business is not going to go back to historical normal levels, but stay at the level previously considered holiday peak traffic.”
Retailers want to know how they can do things in a cheaper, faster, more agile way. This focus on flexibility and a lower cost of ownership means they are looking at implementing smaller, connected systems rather than huge, time-consuming systems integration projects.
So some retailers are changing their business models, perhaps selling directly for the first time.
“We have seen a food distributor that served exclusively corporate and educational customers in the UK see their business 100 per cent disrupted by the pandemic,” says Hauca. “They were able to quickly shift their commerce platform to go direct to consumer in five weeks, because of the agility of their SAP commerce implementation.”
With this move to greater agility, some basic building blocks must be in place. One of them is inventory. Retailers must understand where their product is and what they can sell.
Using AI can deliver a much better customer experience than has been possible in the past. As importantly, it enables the retailer to be more agile and make better use of data, which will be crucial as they look to navigate the new, disruptive landscape ahead.
For more information please visit www.sap.com/uk/products/crm.html or contact us at firstname.lastname@example.org