“Shoeby, or not Shoeby…that is the question”

Picture walking into your favorite clothing store, excited to buy the perfect pair of shoes…only to find they're sold out of your size.

Frustrating, right?

That's exactly what customers were experiencing at Shoeby, a fashion chain with stores all across the country. Like many retailers, they were caught in a classic catch-22: some stores had racks overflowing with clothes nobody was buying, while others couldn't keep popular items in stock. Not only were they losing money on unsold inventory, but they were also disappointing shoppers who walked away empty-handed.

It's the kind of headache that keeps retail managers up at night.

The Problem

Shoeby’s inventory management relied heavily on manual tracking systems, such as spreadsheets and staff estimations. These methods couldn’t keep up with the complexities of managing stock across 240 stores, each with its own unique customer preferences and sales patterns. Staff often struggled to accurately predict which items would sell quickly in a particular location, leading to imbalances. For instance, trendy items might fly off the shelves in urban stores but linger unsold in smaller towns.

This mismatch caused several (not uncommon) issues:

• Overstocking: Some stores were left with excess items that didn’t sell, tying up money and storage space.

• Understocking: Meanwhile, other locations ran out of popular products, frustrating customers and losing potential sales.

• Season-End Waste: Leftover inventory piled up at the end of sales cycles, forcing the company to either discount heavily or write off unsold goods, further cutting into profits.

Without accurate, real-time data and predictive insights, Shoeby was essentially making educated guesses—a method that fell short in an increasingly competitive retail landscape.

The Solution

Enter…the AI Replenisher! (Which somehow sounds vaguely threatening to me. A kind of retail Terminator…)

The AI Replenisher, a smart inventory management system from a company called WAIR, T predicts SKU-level sales for each shop floor up to 14 days in advance, automating inventory adjustments to align stock levels with actual demand.

Here’s how it works:

1. Data Analysis: The AI Replenisher pulls in historical sales data from all Shoeby stores. This includes information like which items sold best at which times and locations, customer buying trends, and seasonal patterns.

2. Demand Forecasting: Using machine learning algorithms, the system predicts future demand for every product at every store. For example, if a certain jacket is popular in urban stores during autumn, the tool predicts how many will likely sell next season in each location.

3. Automated Replenishment: Instead of relying on staff to decide stock levels, the system automatically determines how much inventory to send to each store. It also adapts in real-time to unexpected trends, such as a sudden surge in demand for specific products.

The AI Replenisher removes the guesswork from inventory management for Shoeby’s.

The Results

Shoeby saw significant improvements almost immediately:

• Faster Inventory Turnover: The AI ensured that stores stocked the right products in the right amounts, leading to a 4% increase in turnover. In practical terms, items spent less time sitting on shelves and sold more quickly, which is crucial in fashion where trends change rapidly.

• Less Leftover Stock: With better predictions and distribution, Shoeby reduced unsold inventory by 2%. This not only cut costs associated with holding excess stock but also decreased waste, a key win for both profits and sustainability.

• Revenue Growth: By improving efficiency and ensuring popular items were always available, Shoeby boosted overall revenue by 3%.

Bringing in AI to crunch the numbers and predict what should go where freed up Shoeby’s team to do what humans do best: helping customers find the thing they are looking for.

Case Study Available Here.

Random Fun Fact of the Week

Slime molds, despite being single-celled organisms with no brain, can solve complex navigational problems.

In experiments, they've successfully mapped efficient routes between food sources that mirror human-designed transportation networks. For example, when researchers placed oat flakes in a pattern matching the locations of Tokyo's major cities, the slime mold created a network almost identical to Tokyo's actual rail system - but the slime mold did it in under 24 hours, while the rail network took engineers years to design.

Contact us at NorthLightAI.com to learn how we can help you build a stronger data foundation for your AI future.