![\"Retail](\"https:\/\/dmsretail.com\/RetailNews\/wp-content\/uploads\/2022\/05\/RETAIL-ONLINE-TRAINING-728-X-90.png\")
<\/a><\/p>\n<\/p>\n
The requirements for retail success don\u2019t get much more basic than the ability to accurately forecast customer demand. Even a mom-and-pop bodega has to have a pretty good sense of how many people will order a breakfast sandwich and coffee each day. And for larger-scale retailers, that need for an accurate forecast gets complicated, and re-complicated, by the need to correctly place and price inventory in multiple locations in order to achieve maximum sell-through with minimum discounting.<\/p>\n
Can artificial intelligence (AI) give retailers and their supply chain partners the guidance they need to improve demand forecasts and inventory optimization? It\u2019s already doing so, and generating measurable results in a range of use cases. One of the most promising is a challenge that has bedeviled merchandisers, marketers and inventory managers for decades, if not centuries \u2014 demand forecasting.<\/p>\n
\u201cDemand is typically the most important piece of input that goes into the operations of a company,\u201d said Rupal Deshmukh, a Partner in the Strategic Operations practice at Kearney in an interview with Retail TouchPoints<\/em>. \u201cPoor demand forecast accuracy equals cash out the door. For example, in the chemicals industry, if you\u2019re at 40% to 50%<\/strong> forecast accuracy, you\u2019re probably holding two to three times as much inventory as your peers.<\/strong>\u201d<\/p>\n And therein lies the opportunity.<\/p>\n Dive deeper into how retailers are unlocking new efficiencies in their inventory practices with AI and see it in action at Floor & D\u00e9cor<\/strong> in this <\/em>free Retail TouchPoints special report<\/em>.<\/em><\/p>\n Even when a retailer\u2019s overall demand forecast is optimal, inventory placement represents a big challenge. . Retailers are well aware that having too much inventory in the wrong store, or not enough in the right one, affects not just sell-through but labor costs and store operations. These mismatches also create the need for either markdowns or additional shipping costs to get the products to where they are more likely to be sold.<\/p>\n Compared to demand forecasting, \u201cinventory is more of a math problem that\u2019s tough to get right,\u201d said Deshmukh. \u201cThe equation is simple: there\u2019s this much variability in demand, this is how much safety stock you\u2019ll need \u2014 but the inputs going in are often wrong, and that\u2019s when you get ripple effects.\u201d<\/strong><\/p>\n Deshmukh noted that retailers\u2019 forecast inaccuracies often stem from overreliance on their own internal data: \u201cThey often have a lot of POS data, but they don\u2019t use [the data that] product companies have or even what consumers have,\u201d she said. \u201cThey don\u2019t listen to consumers or market signals; we find over and over again that they don\u2019t do this right. Companies that are doing well have a true pulse on market signals<\/strong>, for example what consumers are seeing on social media<\/strong>, as well as weather trends<\/strong> and the impact of geopolitical events<\/strong>.\u201d<\/p>\n This is an area where AI can shine, she explained: \u201cAI allows you to evaluate unstructured data in a much more structured way,\u201d said Deshmukh. \u201cUse of large language models (LLMs) and generative AI can help in the area of customer inputs. <\/strong>A lot of companies find that their sales teams get inputs through emails and calls they receive, but they can\u2019t convert that into a data point. Gen AI could be a game-changer in this area.\u201d<\/p>\n AI\u2019s ability to gather and synthesize data from so many sources also makes it valuable for retailers managing another part of their inventory: retail media ad placements<\/strong>. With a better understanding of not just how many shoppers are in a store (or on a website) at a particular time but also their demographics, lifetime value and purchasing intent, retailers can both improve the performance of advertisers\u2019 campaigns and quantify the incremental benefits.<\/p>\n Many of AI\u2019s enhancements to demand forecasting and inventory optimization already are being recognized, but industry experts agree that there\u2019s still plenty of white space for additional benefits. However, tapping these benefits will require more than just technological advances; user mindsets also need to change.<\/p>\n \u201cWe\u2019re still missing people who have the vision to understand what is possible [with AI], and the ability to connect that vision to the people who can ask the right questions,\u201d said Fabrizio Fantini, PhD, VP of Product Strategy at ToolsGroup in an interview with Retail TouchPoints<\/em>. \u201cThe number one piece of feedback we get is that before working with these [AI tools], we didn\u2019t even know what was possible. Well, now that you know it is possible, you need to acquire the capability to systematically get these [tools] and embed them into day-to-day operations<\/strong>.\u201d<\/p>\n When this happens, it opens up opportunities for a highly integrated set of systems that are focused on the customer. \u201cUltimately, any kind of data is about the customer, directly or indirectly,\u201d said Fantini. \u201cWhen you suddenly have the ability to cost-effectively and systematically infuse data into your decisions, your ability to create an outcome [you want] will be much more integrated. Supply chain decisions will be integrated with pricing and marketing decisions as well as financial and budgeting decisions; it becomes part of a continuum<\/strong>.\u201d<\/p>\n Retailers need to build on the \u201cpockets of goodness that AI is creating across a large number of businesses,\u201d added Deshmukh. \u201cWe\u2019ve seen AI applications for demand sensing drive tremendous value; telecom and home goods retailers using external market signals to augment forecast accuracy has led to improvements by as much as 10 to 20<\/strong> percentage points. Additionally, companies that have started using AI in their procurement and supply spaces, to test categories or quantify massive amounts of spend data, are doing that faster and with higher accuracy.\u201d<\/p>\n Fantini sees AI\u2019s effects in this area becoming even more widespread as the tech advances. \u201cThese AI models are unlocking supply chains\u2019 capabilities for doing good, and everybody is a winner,\u201d he said. \u201cThe consumer gets better service at potentially lower costs, the company gets more efficiency and profit. They reduce waste, improve environmental issues and boost economic performance by making the market work better.<\/p>\n \u201cThe supply chain is something you only know exists when it doesn\u2019t work, but [it can fail] in many different ways<\/strong>,\u201d Fantini added. \u201cHowever, if you\u2019re making it more efficient, everyone wins.\u201d<\/p>\n Dive deeper into how retailers are unlocking new efficiencies in their inventory practices with AI and see it in action at Floor & D\u00e9cor<\/strong> in this <\/em>free Retail TouchPoints special report<\/em>.<\/em><\/p>\n<\/p><\/div>\nUnlocking the Power of Unstructured Data<\/strong><\/h2>\n
True Advancement Requires Humans who Understand AI\u2019s Potential<\/strong><\/h2>\n