For Eldorado, Real-Time Data Is a Competitive Edge

Eldorado, a large Russian retailer, reduces shortages and improves profitability by accelerating its response time to changing market conditions.

Success in retail is often accompanied by complicated side effects —  like keeping inventory levels in line with demand following a marketing campaign. Eldorado, the largest retailer of consumer electronics and household appliances in Russia, recently took action to keep its stores from running out of goods when response to an advertised promotion is unexpectedly high.

“Out of stock is an issue all over the world,” says Fabricio Granja, Eldorado’s deputy vice president for IT. “The empty shelves affect the image of your entire company.”

Indeed, a new Deloitte survey of 5,018 U.S. consumers found that 45 percent of respondents would switch store chains or Websites when unable to find a desired item in a retailer’s store (see Figure 1, “Promotions, Out-of-Stocks and Customer Behavior”). Correcting this problem required tighter coordination between Eldorado’s marketing, sales and warehouse operations. However, getting timely insight about sales to synchronize those areas of the business was difficult, given the sprawling nature of the company.

Eldorado Retail Story Figure1

Many Moving Parts

Founded in 1994, Eldorado has stores in more than 450 cities across Russia. It sells more than 20,000 items in 110 product groups — everything from refrigerators to headphones. “When you operate across seven time zones, you are challenged on how to plan deliveries to stores, handle stock replenishments and take care of the customers in each region,” Granja says.

Eldorado also has a history of expansion and innovation. Although the retailer has long been a leader in mass-market sales, it has increased its presence in the last two years both through Internet sales and by moving into premium goods.

Adding to the retailer’s complexity, Eldorado introduced a concept that was new to the Russian market: Purchasing online, and picking up the goods at small retail outlets. That concept added more than 30 Internet units and 120 order and pick-up points to the company’s omnichannel portfolio. Eldorado’s marketing is just as multifaceted. The company continuously runs promotions for both its online and its brick-and-mortar stores. On its Website, a visitor might see offers ranging from “product of the day” to “30 percent discounts on Real-Time Analytics Case Studies Bloomberg Businessweek Research Services 2 30 popular items.” Tracking all these efforts — and making sure they do not cannibalize each other — Is another part of Eldorado’s big data puzzle.

The question was how Eldorado could obtain real-time insight into sales to adjust its overlapping promotions and inventory across its many regions and channels on the fly.

Speeding Information Flow

The solution was in-memory computing, which can analyze huge amounts of data quickly. Eldorado uses in-memory computing to power its warehouse and merchandise/assortment planning applica-tions, which support its four warehouses and retail stores. Those applications connect to the enterprise resource planning systems and point-of-sale data-management mobile solution.

This setup has dramatically improved data accuracy and Eldorado’s reaction speed. The retailer can now produce its 10 most-requested reports 90 percent faster than before. A report that would typically take five minutes to run can now be done in less than 30 seconds (see Figure 2, “In-Memory Impact”).

This improvement has important ramifications. Because information can be produced so quickly, employees rarely have to anticipate which data they will need, as it is at their fingertips, enabling them to react immediately to changing market conditions. For instance, marketing can instantly see how a promotion is affecting sales and then offer discounts or shift inventory to maximize revenues. The result is a significant reduction in shortages and cost of capital, which improves profitability.

Eldorado Retail Story Figure2Speeding the flow of information has enhanced many other elements of Eldorado’s business. Complex assortment planning sped up by two-thirds, reducing planning time from 35 hours to 12 hours. Granja says product gross-margin calculations can also be planned more precisely because they are based on real revenue. “It has enhanced my relationship with the bank that provides our credit line,” he says, “because having the financial information allowed us to do better financial planning.”

Real-time information enables executives to stay abreast of current strategies. “We reduced cost allocation from four days to four hours, which reduced financial closing to three days,” Granja says.
“I can now send reports to shareholders much faster.”

Because of the data compression features of the in-memory computing system, Eldorado has lowered its database storage from 2.33 terabytes to 535 gigabytes — a 77 percent reduction. The in-memory processing architecture is also easier to manage, because data warehouse administrators no longer need to spend time on the technical details of defining and managing data aggregates. “My staff can spend more time on analyzing the data and planning, rather than collecting the data,” Granja says.

The next stage for Eldorado will be to apply in-memory computing to customer-centric processes such as customer segmentation. This would enable personalized promotions based on information such as customer birthdays or previously purchased items. Given that the retailer’s loyalty program includes 10 million members, plenty of data exists for doing just that.

“Instead of taking days to plan a promotion, we’ll be able to do it in two or three hours,” Granja says. “Given how fast the world of retailing is becoming, we need to react more rapidly over the entire cycle of the customer relationship.”

See how Eldorado, STM and others profit from SAP solutions for retail in 2014:

Joe Mullich is a freelance business and technology writer based in Sherman Oaks, CA.
Source: Bloomberg Businessweek
Top image: video screenshot