When Retail is a matter of software and “platforms”
Omnichannels, big data, back office, e-commerce, unified commerce, digital transformation…
Does the world of retail now speak in software and web languages so everything becomes over complicated?
Never fear: here are the solutions you need to have a look at the big picture while improving customer experience.
Interviews with Evo Pricing, Software AG and Cegid
Italy, October, 2017
In this current digitized world, the retailer evolution becomes an indispensable factor in achieving a unified management of all channels and providing consumers with a pleasant shopping experience.
This means the most advanced marketing policies and attitudes of distributive marketing such as Client First, Clienteling, or Replenishment have to be implemented, not only as entries in the glossary of retailers, but also in the omnichannel and multichannel logics. Help for this comes from technology, starting with predictive analytics-based applications through to draft correct sales estimates and then having a comprehensive view and optimum stocks for the point of sales.
This gives us the ability to improve inventory management through an IT platform that connects all the touchpoints used to purchase products, providing a vertical and real-time view of each channel, thus allowing us to calculate margins and monitor customer satisfaction, to use it for improving business performances.
In the following pages, we offer an overview of interviews with companies that can offer useful software platforms able to integrate all the elements needed for omnichannel management logics.
Optimized management with predictive analytics.
Evo Pricing was born in 2014 from a PhD research project concluded in 2008 to develop a software for maximizing airline sales.
The focus then shifted to the retail, which had no real reference model.
Today, Evo Pricing is a consulting company that helps retailers worldwide to improve profits in a wide range of sectors, from electronics to fashion, online and offline, with applications for even wineries, taxis and souvenir sellers and, more recently, the banking and insurance sectors.
The first client on Evo pricing’s books was from the fashion industry, the Miroglio group in Alba.
“Our task was, and is, to apply personalized solutions of predictive analytics to customer activities,” says Fabrizio Fantini, the company CEO.
“From a predictive analytical point of view, retail is a homogeneous business and, with regard to quantities management, it only has the distinction between two macro categories of perishable and non-perishable goods. As a result, the activity takes into account pricing (prices, promotion, sale, discounts) and the merchandise handling (sales forecasting, items introduction in the stores, logistics, replenishment), which are rather technical issues because they deal with much of work in retail.”
Based on these assumptions, Evo Pricing, with offices in London and Turin, employs 21 people and uses a delivery area (factory) in Turin, where data scientists and mathematical statistical engineers work in collaboration with universities and other academic institutions in the recruiting process and development of algorithms.
But how do these softwares work?
“All of our solutions,” continues Fantini, “have a common predictive analytical matrix. In order to make analyses, we collect internal business data (transactions, customer data, products data) and external (weather, demographics) to understand the sales phenomenon – a latent gold mine – and transform them into a product on sale, that is, a rough product that we need to further process data for and then define the movement of goods and the likelihood of sales to calibrate prices and promotions.”
An apparently complicated process but, in reality, the shopkeeper’s problems are quickly identified and resolved. In practice, the fashion retailer, for example, faces enormous difficulties regarding replenishment because the product life cycle is very short, and it is difficult to estimate sales at a single size level.
To solve these problems in the fashion chain stores, a system has been developed based on the logic of the ‘stock exchange’ that can improve the warehouse. As Fantini explains, “We use artificial intelligence systems for making a proposal. Each shop then has the option of buying or selling articles, thanks to this system, changing, in fact, the proposal. In this way, local sales probabilities are best estimated while improving inventory and exchanges between sales outlets.”
“We have also shown that in fashion sector, the retailer intuition helps the more advanced artificial intelligence systems to make predictions while moving goods between stores is profitable despite the logistics, because the cost-benefit ratio is good. In addition, decentralization of the process simplifies the flow of retail decisions, which is often handled manually by headquarters through phone calls with area managers. Our system only takes an hour’s work on Monday morning to give benefits to the whole process.”
Evo Pricing’s challenge is to evolve the customer-retailer that uses standardized methods or still relies on the statistics of previous years to draw up sales estimates. The model of service offered is innovative and requires the delivery of a ‘supported software’ for which customer-retailers provide feedback and receive weekly suggestions.
In conclusion, Fantini highlights his platform’s ease of use and affordability. “Some software on the market boasts many features, but they are difficult to use and very expensive. Our model has no initial costs but starts from a risk-free diagnostic project because our goal is to provide continuous support to customers in price estimation, replenishment (stock management in this shared case), and promotions. Ultimately, we help people in the decisional process to become more responsive, quick and precise.”
A successful case in electronics accessories is Mobo, a family business that has created a real retail empire in Mexico for five years, with over 200 well-managed retail outlets. Growth was fast, and it was therefore necessary to structure the processes.
“After analyzing price sensitivity at a geographic level, we realized that there was no relationship to the level of prosperity (income per capita) by deducing that the willingness to buy depended on local availability (distance) and product placement (value for money). From an analysis carried out in Mexico, in the first phase we have identified four different bands of price sensitivity. Next, we helped retailers to introduce a price differentiation cycle between stores with price variations every three weeks: we do not talk about dynamic pricing but handling, variations in price lists based on demand and supply logic.
“In the third step, we helped retailers to learn how to systematically measure the impact of promotional initiatives, a difficult task for the average retailer, with volatile data and many factors that affect the results. Thanks to the synergy of the three phases, after having set up a clear analysis, a systematic price-change process and an automatic and stable metering method, the chain has exceeded USD 20 million in impact.”
This is a practical case where part of the work is about the elimination of the so-called Forrester effect, which indicates an increase in demand variability as we walk away from the final market while going back to the supply chain.
Continuous exchange of stock information between the headquarter and stores eliminates this problem that can usually occur, for example, in a stock rupture of both excess and defect. Practically, in addition to what is described by Evo Pricing as a way to overcome these drawbacks, the actions are based on the concept of Continuous Replenishment, a management model implemented in the distribution system of Walmart, where individual stores in this chain transmit data on “Reorder Points” (in terms of stock), from the cash register to their headquarters several times a day.
With this information, demand is used for queue shipments from the Walmart distribution center to the store and from the supplier. The result is a perfect visibility of customer demand and warehouse movements throughout the supply chain.
Better information leads to better inventory position and lower costs throughout the supply chain.
On this subject, you can consult Hau Lee, Corey Billington, Managing Supply Chain inventory: Pitfalls and Opportunities, SMR, 1992.