Back in the days when electronic point of sale (EPoS) was still a novelty, retailers and analysts agreed that the most useful thing you could do with the copious amounts of data it produced was prop open the office door.
Thankfully, things have changed. Today's PC screens have easy-to-follow graphics that alert users to stock problems, possible fraud, profitable customers or whatever else the number-crunching tools are set up to discover. And business intelligence (BI) is the all-embracing name for a raft of IT applications described in the past as "decision support", "executive information" or just plain "analytics".
EPoS data is generally a starting point, but today it includes much more than the details of items sold. It provides data about when the item was sold and by whom; about which other products were bought at the same time; about the price - whether it was discounted or later refunded; and, if the customer can be identified by a loyalty card or by providing an email address or postcode, by whom it was bought.
All this information can be fed into a central system along with other data on product assortment, stock levels, supplier performance or staff scheduling, and regurgitated to provide insights that will help the business run more efficiently and profitably.
The central system may be a powerful data warehouse or, thanks to changes in technology, something much smaller. It may even be a mixture of platforms communicating via web-based technologies such as a services-oriented architecture (SOA).
Harrods has already taken the SOA route. It has restructured its core IT applications over the past four years to combine its surviving legacy applications, a SAP enterprise solution and assorted point of sale solutions with the help of a SOA system from SUN. The new system includes master data management architecture, which provides a single source of information about the business. Harrods now has a single view of its customers, regardless of where or which channel they choose to shop.
"We can now measure issues such as response rates very accurately because there is no duplication of data," says IT director David Llamas.
In the past, the company typically sent 89,000 seasonal discount offers to its store card holders by post. But by identifying customers who had spent £25 or more at the store in the previous period, the company cut the mailshot to 59,000, saving £12,000 in marketing costs and achieving a far higher response rate.
Establishing a single customer view was a vital step in setting up Harrods' loyalty scheme, launched last year. The company now has up-to-date data on the best customers and any changes in their shopping habits. It can profile these shoppers by demographic and lifestyle, to target similar groups to expand its customer base.
Shoppers are ranked as green if they spend more than £5,000 a year at the store, gold if they spend more than £15,000 and black if they splash out £50,000 or more each year. Spend can be in any channel - store, call centre or website - but thanks to the single customer view, it can be attributed accurately to each shopper. Those in the black group will be given personal one-to-one service and bespoke products, gold shoppers will receive exclusive invitations to store events, while greens receive specially targeted offers.
"Our aim has been to provide an operational CRM (customer relationship management) solution that enables us to align departmental strategies, which can communicate a consistent message and brand experience across all our customer touchpoints," says Llamas.
While Harrods has used SOA to stitch together IT systems, Spanish fashion retailer Cortefiel has built its business intelligence around a Teradata enterprise data warehouse. The system has been used for several years to manage product and performance information for its 1,200 stores in 25 countries.
Cortefiel has started using this data to drive a markdown optimisation system. Planning and distribution director Richard Gum says markdown periods account for about 33% of total sales. In many countries in which the group operates, Sale periods are defined by local by-laws, so stock must be cleared at optimum prices if the store is to avoid being left with racks of unsaleable product. "Clearance pricing really is key to profitability for us," says Gum.
The company's 'markdown adviser' system draws on existing merchandise and sales data, looking at sell-through rates, historic performance and stock levels. It then suggests the best markdown strategy for that particular market, given the relevant Sale regulations.
Work on the project began last spring and the first trial was held in November. Cortefiel applied the new markdown calculations to end-of-line women's knitwear. It used the application to set markdowns for its Portuguese stores and the usual manual merchandiser's calculations for the same ranges in its Spanish stores. The new system suggested what Gum calls "very aggressive markdowns" - far more drastic cuts than the merchandiser would have applied at the start of a Sale period.
The results were impressive. The Portuguese stores sold 52% more product than the Spanish outlets and achieved 23% more in gross margin than in Spain. This produced a three percentage point difference in gross margin which, applied across the entire range, Gum calculates would bring a 6% increase in profit for the year.
Cortefiel uses a traditional enterprise data warehouse, but smaller number crunchers - or "appliances" - have also been launched, putting this sort of capability well within reach of mid-sized companies. Debenhams, for example, went live with a Netezza appliance last June.
"These systems plug and go without the need for time-consuming set-ups," says Huw Ringer, principal consultant for Smart Associates, which was involved in the Debenhams roll-out. "It makes tasks such as pattern analysis for loss prevention easily achievable and it's possible to carry out near real-time fraud indication at point of sale. You can also drill down and look at customer behaviour to improve marketing."
Many systems aimed at the independent sector also have business intelligence capability. Young fashion chain Bank, which has expanded from 22 stores to 44 in the past two years, has used Futura Retail Solutions' tools to analyse trends and ensure goods are in stock in the right place. Using historical data, the system forecasts buying patterns and enables buyers to use "what if" functions to see the effects of buying one range over another.
When Bank first analysed its customers' size profiles, buyers found that traditional size ratios of 1,2,2,1 were not always appropriate for every style. Analysis showed that to meet this profile, they needed to buy 2,2,1,1. Bank estimates that this increased sell-through by 5%.
These case studies show that business intelligence can have a real, quantifiable effect on a retailer's front line. Be it via an enterprise data warehouse, retail management suite or clever new IT such as SOA, business intelligence is delivering sizeable benefits to fashion retailers big and small.
BUSINESS INTELLIGENCE AT THE RETAIL SOLUTIONS SHOW
Systems to improve business intelligence will be on show at this year's Retail Solutions event, which takes place at the Birmingham NEC on June 5-7.
- TXT will launch TXT Drive Analytics, which can be integrated into its existing demand forecasting and replenishment software. It includes a powerful analytical engine that can drill down through the data to provide more control of supply chain operations by offering increased visibility of demand from warehouse to store.
- VcsTimeless will demonstrate Colombus Business Intelligence, an end-to-end decision support and analytical software solution that comes pre-configured and includes data warehousing, data analysis and reporting.
- K3's BI tool is modular to speed up implementation and ensure rapid return on investment, while focusing on usability. Dedicated reporting packs include loss prevention and sales analysis, and customers can combine the packs to build a package to suit.
- Manthan will demonstrate ARC, an integrated BI suite with a target market that includes fashion and footwear retailers. It is built on an enterprise data warehouse and includes an extensive library of retail KPIs (Key Performance Indicators), retail reports and analysis.
- JDA will unveil the first results of its takeover of Manugistics last year with applications running on the JDA Enterprise Architecture. The platform provides one view of demand across all enterprise planning and optimisation activities, to improve decisions in everything from sourcing to shelf optimisation.
For more information on Retail Solutions, visit www.retailsolutions2007.com.