Through near real-time analysis of electronic point of sale (EPOS) data from iconic British retailer Marks & Spencer, Irish convenience food manufacturer Greencore has delivered substantial business benefits: improved on-shelf availability for pre-prepared sandwiches, increased sales, reduced waste—and an ROI payback estimated at less than a year.
The business intelligence project, first tested in April at Greencore’s sandwich manufacturing facility in Northampton, U.K., used Qlik Technology’s QlikView BI tool and proved itself within three months, say Greencore management. Using a pre-prepared ‘wrap’-style sandwich line as a testbed, the facility’s commercial team set out to better forecast consumer demand, enabling the Greencore merchandisers responsible for stocking Marks & Spencer stores to respond with more agility than was previously possible with Microsoft Excel-based reports and pivot tables produced on a weekly basis.
“We were often slow to react if a particular product wasn’t selling well in a particular store,” says Jenny Thompson, a commercial executive at Greencore, a $2 billion company based in Dublin. “We had no real‑time information on which were the best‑selling lines, so stores in some areas ran out of stock, and in others they had to throw away food.”
Today, in contrast, QlikView has made the creation of daily reporting and analysis of wrap sales at Marks & Spencer faster, easier, and more flexible. The result: faster decisions, better decisions, and more profitable decisions—for both Greencore and its customer.
Accelerating Supplier’s Access to Customer Sales Data
The starting point: Marks & Spencer’s online supplier management information system, which Greencore employees such as Thompson access in order to run reports detailing the sales performance, by store, of particular products or product categories.
The data from these reports is first saved locally on the desktop in Excel format, says Thompson, and then saved again on the server, for subsequent analysis in QlikView.
“In an ideal world, we’d have a direct data transfer onto the server, but that’s a project for the future,” she says. Even so, one early problem has now eased, thanks to a move to Excel 2010: the 65,536-row limit of earlier versions of Excel, which sometimes meant breaking larger spreadsheets into narrow product categories in order to circumvent the limit.
And analyzing the EPOS data in QlikView, Thompson stresses, is much faster than attempting the same thing in Excel: analyses that would take over two hours in Excel takes just five to ten minutes in QlikView.
“Accurate decisions can be made quickly because all the data is available within a day in near real‑time from a single source,” says Thompson. “Each member of the commercial team can now easily access information on a self‑service basis—that wasn’t possible before.”
By July, three months after moving to QlikView-based analyzes, says Thompson, on‑shelf availability of wraps had risen by 20 percent, delivering incremental sales of $480,000 for Marks & Spencer, while waste levels reduced by 3 percent.
“Instead of 68 percent availability of wraps, we were recording daily averages of around 88 percent, which in turn means improved sales and lower waste for Marks & Spencer. The 20 percent improvement is due to QlikView and the ease of accessing the data,” says Thompson. “There will always be some waste with convenience food, but the reduction figure speaks for itself. None of this would be possible without the ability to access the real‑time data in an easy‑to‑use format.”
Roll the clock forward to today, and the impetus has continued. With the QlikView-derived analyses now extended beyond wraps into sandwiches more generally, sales have continued to grow, and waste levels reduce—although the exact levels of each are not for public disclosure.
“We’re also carrying out more investigative work: targeting particular stores where waste levels seem high, or targeting particular sandwiches where sales or waste levels seem out of line,” says Thompson.
A Popular and Perishable Product
Pre-prepared ‘ready to go’ sandwiches are highly popular in the U.K. market, with Marks & Spencer widely regarded as setting the standard. Yet it’s not a market for the faint-hearted, and good data is critical, says Hugh Williams, managing director at supply chain consultants Hughenden Consulting in High Wycombe, U.K.
For the name of the game is keeping inventory levels down to little more than a day’s stock, he explains: sandwiches have a limited shelf-life, and price discounting is prevalent as the shelf-life nears. And once that life has been exceeded, product must be discarded.
“The shelf-life of a sandwich is so short that there’s no margin for error,” says Williams. “It’s either waste, or it’s a sale: there’s no middle ground—and good data, and good analytics, can determine if it’s one or the other.”
With 28 production sites in the U.K., United States and Ireland, and 11,000 specialists making “ready to go” convenience foods, Greencore would seem to be at the forefront of the battle to make sure that every sandwich is sold, not discarded.
“We look at Greencore as being a good example of our supply chain capability,” says David Telford, Qlik Technology’s senior director of global marketing for manufacturing and hi-tech. “It’s all about taking EPOS data, and integrating it with demand planning data in order to deliver a better informed forecast, and improved availability. In other words, it’s about giving people within functional areas of the supply the ability to make better decisions.”
Back at Greencore, QlikView power user Thompson agrees, pointing to the ease with which it is now possible to drill down into phenomena such as falling sales or rising waste, and investigate potential causal factors.
“QlikView has changed my working life. If I want to know why a product isn’t selling, I can get the answer in seconds instead of waiting hours. Other team members are now also less reliant on me to produce reports because they can do it themselves.”
Freelance writer Malcolm Wheatley is old enough to remember analyzing punched card datasets in batch mode, using SPSS on mainframes. He lives in Devon, England, and can be reached at email@example.com.