Artificial intelligence (AI) has made the move to mainstream. What began as proof-of-concept now underpins familiar functions such as predictive text and TV show recommendations from your favorite streaming service. What started as a scientific curiosity now forms the foundation of business chatbots and powerful tools like ChatGPT.
Not surprisingly, AI is also pushing into the supply chain. As noted by Supply Chain Brain, 90% of enterprises have now experimented with AI in their supply chain, and 29% consider this an area of “heavy investment” over the next few years.
The challenge? Turning the promise of AI into actionable output. Much like the cloud before it, many AI solutions are heavy on fanfare and light on functions. Here’s a look at how AI could practically support pricing, demand forecasting, and inventory planning for global brands.
Pinpointing product prices is no simple task. Multiple factors impact optimal pricing, from current consumer demand and available stock levels to competitor costs and the price differential of similar products in the same category.
As a result, pricing is typically a best-guess scenario. Using what they know about current market conditions and historical demand, companies attempt to select price points that both generate revenue and keep customers coming back.
Set prices too high, and customers will opt for competitors. Set them too low, and businesses must find balance with other product pricing to keep cash flowing.
AI tools can deliver price points that change in response to market forces. By collecting and analyzing data from multiple sources, AI tools can discover trends and patterns that help brands fine-tune product prices.
For example, companies could use AI to track social media mentions of popular products and the number of products sold over a set period. If both values increase, AI can dynamically recommend new, higher pricing to account for changing demand.
According to data from global consulting firm Boston Consulting Group (BCG), AI-enabled pricing offers practical benefits: Brands that have implemented this approach have seen gross profit margin increases of 5% to 10%.
While AI-driven pricing offers benefits for businesses, it also comes with concerns for consumers. Two issues are potentially problematic: Price gouging and inaccurate outputs.
Consider the recent announcement — and backtrack — of dynamic pricing by fast-food chain Wendy’s. As a Quartz piece reports, in February 2024 the company announced plans to install digital menu boards that would reflect food prices based on demand.
Put simply: Burgers and fries would cost more when the restaurant was busy and demand was high. Not surprisingly, customers weren’t pleased and the plan was quickly scrapped. While Wendy’s claimed price gouging wasn’t the plan, buyers weren’t buying it.
As noted by the MIT Technology Review, AI doesn’t always return accurate outputs. Although AI is bound to use the rules laid out by machine learning (ML) algorithms that underpin its operations, these tools can only analyze the data they’re given. If this data is incomplete, biased, or just plain wrong, prices could be wildly off-base.
Demand forecasting attempts to predict what people will purchase when they’ll choose to buy, and how many customers will buy the same thing at the same time. The more accurate companies are with demand forecasts, the higher their potential revenue.
AI makes it possible to improve demand forecasting accuracy by considering two key factors: Demand volatility and product life cycle.
According to the MIT Sloan Management Review, high-volatility products see large spikes in demand over short periods, while demand for low-volatility products is stable. Products with long life cycles remain on the market for months or years, while short life cycle items are sold quickly.
As the MIT Sloan piece notes, however, demand forecasting isn’t an AI-only job — humans also play a role.
For example, in the case of products with low demand volatility and long life cycles, the ideal human/AI split Is 60/40. When it comes to high volatility products with short life cycles, meanwhile, such as fast fashion or beauty products, businesses are better served with a 20/80 split of human expertise to AI frameworks.
Effective inventory management means hitting the sweet spot. Too much inventory can lead to overstocking — research and advisory firm IHL reports that in 2023, overstocking cost companies more than $562 billion.
Not enough inventory, meanwhile, sets the stage for stock-outs. As noted by McKinsey & Company, stock-out potential remains high as logistics costs continue to rise and critical shipping routes — such as the Panama Canal — experience weather-related delays.
AI has the potential to improve inventory planning and help companies reduce the risk of costly overstock and customer churn tied to consistent out-of-stock. By analyzing historic customer purchasing patterns and incorporating data current market data, AI tools can help predict how much stock companies will need to satisfy customer demand without keeping more than they need on hand.
Consider a business that sells beach umbrellas. Historical data provides a baseline for customer purchase patterns, while the seasonal nature of the product means more stock is needed for the summer months. AI, meanwhile, can take a deeper look.
For example, AI tools might use historical and current weather data to predict a hotter-than-average summer, in turn increasing the number of umbrellas that companies need in stock. AI-enabled solutions could also help track social media mentions — if well-known influences are declaring it #umbrellasummer, orders could rise sharply.
By using AI to augment existing inventory planning tools, businesses are better prepared to navigate changing conditions.
AI has reached an inflection point. Ecommerce brands can’t afford to ignore the potential of AI, especially in areas of pricing, demand forecasting, and inventory planning.
But with improved intelligence comes the need for greater oversight. While AI solutions make it easier to turn data into action, human expertise plays a critical role in ensuring fair pricing, integrating demand forecasting into long-term success strategies, and validating inventory planning predictions.