KIKO Milano

KIKO’s dramatic growth created greater complexity in a highly competitive and constantly evolving cosmetics market. SO99+ came to the rescue.


Improved forecast accuracy including seasonality of products and stores Precise management of retail promotions and product launches Greater planner productivity with automated exception management

KIKO Milano: A New Customer-Centric Era

An extremely wide and diversified assortment, colorful offers and safe and high-quality make-up and face and body treatments are distinctive features of Italian cosmetic brand KIKO Milano, founded in 1997. The dramatic growth of the company created additional complexity and the need to find new competitive momentum within a constantly evolving cosmetics market. With the arrival of a new CEO in July 2017, an operational renewal program began which initiated significant changes to KIKO’s supply chain forecasting and replenishment procedures.


KIKO had been a ToolsGroup customer since 2013. The partnership took a new direction as the replenishment model shifted to better respond to market demands and lay the foundation for future company growth. “We were able to increase product availability at the store level by improving our service level to compete with large-scale retailers,” explains Gianmarco Mangili, KIKO Milano Planning Director. “Aligning our supply chain planning processes with customer service and commercial goals helped us more dynamically respond to consumer preferences and improve the pace of promotions.”


Complexity of Planning for Retail Cosmetics

The new supply chain process had to tackle several challenges, starting with shop layout: slots for each item must be set up with exhibitors and must never be empty. “The product on display should not be considered safety stock: the supply chain task is to guarantee, at the point of sale, the availability of products for all categories while preserving the minimum,” explains Mangili. This task is made extremely complex by the large number of products—around 1,500 SKUs. These include continuous baseline products renewed at a fast pace, new seasonal collections, special events, and capsule collections with very short lifespans.


Even in the case of products with a high consumer loyalty rate, sales performance is strongly influenced by promotions introduced continually throughout the year. Each promotion has a varying impact on different stores, and is influenced by seasonality, weather, and fickle fashion trends. “In cosmetics there are many factors influencing customer behavior that are difficult to predict,” explains Mangili. “Our previous systems were not able to deal with these complexities; with SO99+ we are finally able to precisely manage promotions and product launches.”


Providing our customers with the best possible service is the driver of everything we do.” – Gianmarco Mangili, KIKO Milano Planning Director


Demand Forecasting with Machine Learning

KIKO utilizes ToolsGroup Service Optimizer 99+ (SO99+) both for the definition of the warehouse purchase plan and store replenishment. This supply chain planning automation software generates a reliable demand forecast by blending machine learning automation with probability forecasting. The ToolsGroup machine learning engine allows models to “learn” from existing data and accurately identify trends for future demand. It augments planners’ knowledge and skills, working as an intelligent assistant helping them perform their work more efficiently and profitably. The ToolsGroup solution generates clear logic that can adapt dynamically and easily align with company policies and procedures.

Rapid Response to Sales Trends

The re-engineering of replenishment procedures was complementary to the reorganization of supply chain logistics, now entrusted to a single partner, which manages the central warehouse and distribution to over 900 stores worldwide. The integrated planning of these activities requires ordering twice a week: on Monday to ensure the arrival of the goods in the store the following Thursday and on Thursday for arrival at the following Tuesday. How does KIKO accurately predict consumption trends in each store? Every Monday morning the sales analysis and replenishment planning process is performed, which generates store orders. This allows KIKO to quickly identify consumption trends. “Thanks to the tight integration between functions, we can ensure the availability of products in very short timeframes, increasing the chances of satisfying demand in time for the sales peaks of the following weekend.” says Mangili.


In cosmetics there are many factors influencing customer behavior that are difficult to predict,” explains Mangili. “Our previous systems were not able to deal with these complexities; with SO99+ we are finally able to precisely manage promotions and product launches.”


An Accurate SKU/Store Forecast

With the previous rudimentary replenishment model, forecasting ability was poor. Today, thanks to SO99+ KIKO’s forecast accuracy is much improved and pinpoints even down to daily sales of single items in each store, considering the seasonality not only of the product but also of the store. The replenishment—the quantities to be sent to the points of sale—is calculated based on the forecast, in-store stock and coverage level defined by KIKO for each store, based on the planned service level, with a rounding to account for the minimum product handling units in stock.


Working by Exception Saves Planner Time

KIKO’s new forecasting and replenishment process requires a very tight and synchronized turnaround: to ensure orders are delivered by Thursday, the replenishment plans calculated on Monday morning must be validated by noon the same day. How is this possible? KIKO’s planning team optimizes the order approval process by working by exception through an automated process that maps a total of eight KPIs and allows for quick evaluation and validation of the replenishment plan. The KPIs consider critical issues such as significant forecast changes, anomalous values of stock per store or the reporting of an anomalous forecast error. In the event of a “red light” on one of these, the plan enters a validation flow that involves manual verification by the team to understand what happened and what action to take. “The system enables true exception management of the replenishment process and provides an excellent percentage of correct pre-validations, positively impacting the organization and quality of the work of the dedicated team,” explains Mangili.


SO99+ has also successfully passed the difficult ERP change test with the move to SAP and is now ready to face new challenges. “We are working to refine and automate the forecast of new seasonal and very short-lived products, and improve the management of promotions which, for KIKO, are the key to further improve our service level.”


↓ Download PDF