Finn Flare (iT Pro: BI.Qube)

2018: Major fashion retailer optimises range

How to distribute goods so that they are sold quickly and without a residue - the headache of the management of any retail chains is treated by algorithms working with Big Data. We will talk about how you can optimize inventory in stores using the example of the implemented case of the Russian fashion network.

Big Data is no longer the preserve of IT companies and large corporations, business trusts recommendations based on analysis of big data more and more. Moving from intuition-based decision-making to data-based management is enabled by analytics built on dedicated technology platforms. Retail, with its endless matrices of articles that cannot be manually optimized, is especially interesting for its dreary approaches.

Valuable volume

To manage the assortment in retail, accounting system tools are traditionally used, the most common of them is 1C. But they simply cannot cope with the analysis of constantly growing amounts of data. Small and medium-sized businesses can still do with simple approaches and manual labor of commodity scientists. And if the volume of analyzed data is calculated by millions of lines, if they come from many poorly connected systems, if you need to transform accounting into managerial accounting, and add specific analysis algorithms from above? Here we already need complex technological solutions and updated capabilities of database management systems.

But, as a rule, such tools are either built into expensive software products of foreign production, which are not affordable for domestic medium-sized businesses, or are introduced for a very long time and not always efficiently.

The analytical solution BI.Qube from the Russian developer IT Pro will cost ten times cheaper and will be implemented in a matter of months.

Leftovers are sweet

FiNN FLARE, which introduced the Avtomezhmag functional block as part of BI.Qube analytics, understood how these technologies are applicable in retail.

The system helps to optimize inventory in stores and stores of the network in such a way as to maintain the completeness of the assortment in retail outlets and correspond to the volume of demand, which ultimately gives traffic and revenue growth. After calculating the speed of sales, Avtomezhmag predicts the risks of excess, calculates the balance deficit, and determines how much goods and how to move to bring the system to an optimal state - to meet the needs at one point due to the surplus of another. In manual mode, even the most advanced commodity expert cannot physically operate with billions of combinations. The amount of data that several people would calculate for two weeks is processed in a few seconds.

Moreover, if there is a lot of surplus, and there are few needs, then for each need the product will be withdrawn from the point where, according to estimates, the prospects for the sale of this product are lower. If there is little surplus and there is no way to meet all the needs, then first of all those who, in terms of sales speed, will bring more money will be satisfied. BI.Qube also helps to solve the specific problem of fashion retail - to get rid of the "tails," moving the unclaimed extreme sizes of the sold-out collection to where the full color model matrix is assembled and there will still be demand.

To evaluate the performance of the recommendation system, we tested the optimization of the entire range of FiNN FLARE between all stores in Moscow. With the ratio "all with all," a matrix was formed with a dimension of 173 thousand units of surplus for 40 thousand needs, or 7 billion combinations of movement.

Distribution of the entire assortment, taking into account fine tuning, perhaps literally, in three clicks. Users can pick up multiple stores in real time and get recommendations for moving balances between them. And commodity experts have more time left to work directly with the assortment of the store. More accurate movement of goods helps to improve the distribution in stores and better meet customer demand by constantly filling the size grid. The seller manages to avoid situations when the buyer chose a thing and wants to buy it, but only the wrong size remained on the shelf. Declining positions stop lingering on the trading floor, accelerated renewal of the assortment increases attendance and demand. An accurate forecast of the speed of sales reduces the need for transportation and, over time, reduces the cost of moving goods. "Avtomezhmag" is a designer, combining the tools of which you can optimize the business process with any accuracy and depth - operate not by sales speed, but by markup, optimize not by maintaining color-sized matrices, but by maximizing profits, taking into account logistics costs, predict demand to automate the order of new collections. According to FiNN FLARE estimates, the effect of the implementation of the analytical solution has already yielded revenue growth of 3-4% across the network throughout Russia.

Cloud of expertise

According to customers of the BI.Qube product, multi-criteria real-time inventory optimization has no analogues in Russia - at least at an affordable price for medium-sized businesses and the speed of implementation. Accurate and prompt analysis is necessary not only in retail. Neftegas, logistics, industrial production, mining and processing of minerals, finance and insurance - optimization methods are chosen based on the peculiarities of the assortment and sales strategy, flexible tools are configured for any commercial tasks.

Analytical tools BI.Qube are not sold "box" - they are selected from the established practice and adapted individually for each client. The more complex the structure of the business, suppliers, customers, product catalog, the more data, the more volatile the external environment, the greater the need for analytical systems. Manage inventory, sales, margins, forecast revenue for each day and for each business unit - the product and development team take into account the business logic and requirements of a particular customer.

The system runs on the Microsoft platform, the analytical solution can be deployed both on the customer's own equipment and in the Microsoft Azure cloud environment - this is the fastest, safest and most reliable enterprise cloud platform. The capabilities of the cloud completely redraw the implementation economy - there is no need to deploy and maintain additional IT infrastructure, in case of peak loads, the necessary capacities are connected on demand. Data from the cloud is available anywhere and anytime, the end business user opens a convenient visual interface on his device - and gets a complete and visual picture of all indicators.

The speed of reaction to changes in consumer preferences was and remains a critical development parameter. Timely identify trends, predict sales and automate purchases, reduce risks in real time - the nuances of enterprise data management hide the huge potential for growing business competitiveness.