Gone are the days when retail operations could be governed by the instinct and intuition of shopkeepers who knew customersí names, buying behaviours, seasonal trends, product preferences and likely future purchases. The complexity of todayís global, multichannel retail environment makes it impossible to glean that kind of knowledge.

Retailers have turned to a variety of technologies in their quest to improve revenues, customer service and operational efficiencies. However, customer transaction and market data collected from different channels often reside in disparate databases and systems, leaving no practical and consistent way to analyse the information for personalised customer insights. As a result, important decisions about merchandise selection, pricing, promotion, positioning, allocation, inventory replenishment, staffing and other aspects of retail operations are made and executed based on incomplete or inconsistent information, leading to suboptimal actions and even costly mistakes.



To survive and prosper in competitive markets, retailers need more. They need the ability to readily access and analyse data to gain comprehensive, accurate and forward-looking retail intelligence whenever it is needed. That kind of insight isnít generated by the operational systems that capture day-to-day transactions; they werenít designed for that. Nor is it generated by the spreadsheets and online analytical processing (OLAP) systems often called analytic systems. Those technologies usually offer rigid and simple views of data. They can tally, track, sort and filter, but they donít synthesise data into the best information or provide a window into the future, a window necessary for proactive decision making. They canít distinguish meaningful trends from noise, clarify why events occurred or identify the significant factors that would lead to repeatable successes or accurately predict future outcomes. In short, they donít deliver strategic analytic insight.

Delivering such advanced insight requires advanced capabilities based on true analytics, the in-depth mathematical investigation of relationships among many variables. While the definition may be intimidating, two key circumstances have opened up new opportunities for retailers to exploit analytics like never before:

􀂔 Retail automation systems yield more data than ever: The burgeoning popularity of loyalty cards and credit cards, the growth of the Internet as an alternative sales channel, the proliferation of operational automation systems and RFID (radio frequency identification) systems are creating a wealth of data that retailers are beginning to apply to better understand and optimise their businesses.

􀂔 All that data can be transformed into meaningful intelligence more readily than ever: Advances in data management and computer processing have made it feasible to quickly distill forward-looking intelligence from huge volumes of disparate operational, transactional and external data. Advancements in user interfaces and packaged applications enable business users to quickly conduct in-depth analysis, without relying on IT specialists or statisticians.

These factors, combined with the spiralling cost and competition pressures all too familiar to retailers are driving adoption of a new level of information technology based on retail-specific analytics. Analytics can be applied to optimise many areas of retail business, such as customer relationships, merchandising, operations and overall performance. Analytic intelligence gives retailers robust ways to understand what is happening and what could happen with quantified accuracy within stores, among stores and across chains.

Quality information
In the old millennium, product and service attributes were prime competitive differentiators. Excellence in products and service are still essential, of course, but they tend to be differentiators only for short windows of time before competitors catch up. The only enduring way to stand apart is to have better information which is the critical ingredient that enables a company to outmaneuver the competition through a continuing flow of renewal and innovation.

That means an enterprise's information management strategy can either be its most compelling asset or its most limiting deficit. Naturally, organisations vary in the degree to which they capitalise on information assets. SAS views different levels of maturity through the lens of the Information Evolution Model, a framework for describing the status of an organisationís evolution toward becoming an intelligent enterprise.

This model describes five fundamental stages that organisations pass through as they advance in their use of business intelligence for competitive differentiation:

􀂔 Level 1: Operate: At this most basic level are the companies rife with information mavericks: the people in isolated offices hammering away on desktop spreadsheets. If they go, the knowledge goes with them. There are no processes, and each request becomes an ad hoc data rebuild, resulting in multiple versions of the truth.

􀂔 Level 2: Consolidate: At this stage, a company has pulled together its data at the departmental level. Here, a question gets the same answer every time, at least within the department. However, departmental interests and interdepartmental competition can skew the integrity of the output and result in multiple versions of the truth.

􀂔 Level 3: Integrate: At this point in the evolution, a company bases its decisions on this more complete enterprise information. This company is beginning to have a true awareness of additional opportunities for the use of BI to improve processes and profits.

􀂔 Level 4: Optimise: At this stage, the retailerís knowledge workers are focused on incremental process improvements and refining the value-creation process. Everyone understands and uses analysis, trending, pattern analysis and predictive results to increase efficiency and effectiveness. The extended value chain becomes increasingly critical to the organisation, including the customers, suppliers and partners who constitute inter-company communities.

􀂔 Level 5: Innovate: This level represents a major, quantum break with the past. It exploits the understanding of the value-creation process acquired in the Optimise stage and replicates that efficiency with new products in new markets. Retail organisations operating at this level understand what they do well and apply this expertise to new areas of opportunity, thus multiplying the number of revenue streams flowing into the enterprise. Armed with information and business process knowledge, organisations approaching this level will introduce truly innovative products and services that reflect their unique understanding of the market, their internal strengths and an unfailing flow of ideas from continuously engaged employees.

We are finding that most large retailers have reached or are approaching the Integrate stage, with many making great strides toward the Optimise and Innovate levels. There is an enormous opportunity for the evolution to continue within every retail organisation.

But real competitive value coming from Level 4 or Level 5 is found beyond the limitations of operational and transactional software; it requires the ability to transform operational and other data into meaningful, accurate, enterprise intelligence and predictive insights.

Leading retailers around the globe have begun using analytic business intelligence to make an array of strategic decisions. Where to place retail outlets, how many of each size or color of an item to put in each store, how much square footage to allocate to a category, when and how much to discountóthe effects of better decisions in these areas can generate millions of dollars for retailers.



What can retailers gain with analytic intelligence?
Many retailers today rely solely upon OLAP capabilities from vendors claiming to be analytic business intelligence experts. Although OLAP stands for online analytical processing, thatís really a misnomer because it actually contains little analytic substance. Most OLAP technologies merely draw on simple descriptive measures and additive capabilities: summaries, weighted summaries, averages, percentages, minimum and maximum values.

OLAP provides a structured way to view and query data, and it may provide some insight into past trends and performance, but it is difficult to determine the significance of trends using OLAP tools. In volatile economies, with so many dependent factors at play, past history (taken at face value) can be a very poor predictor of future events. A retailer could surf the data for days and not find anything significant. Even if he did find something interesting, OLAP contains no mechanism to determine if the pattern, event or anomaly is actually significant.

In contrast, analytics can span not only the past and present to distinguish significance from happenstance, it can also predict specific future outcomes. Analytic processes quantify known attributes, examine complex relationships among many interdependent variables and detect patterns using techniques from a variety of mathematical disciplines, such as statistics, econometrics, time-series forecasting, data mining and operations research.

From huge volumes of raw data comes useful, forward-looking intelligence, presented in meaningful context. Users can drill into results to view detail, discern useful patterns from mere statistical noise, apply models and scenarios repeatedly to different data, select the visualisations that best clarify patterns and actions and change conditions and assumptions to ask ëwhat if?í Letís take a look at how analytic capabilities can enhance success in four key areas of retail business management:

􀂔 Marketing and customer relationship management: Targeting the right customers with the right messages at the right time to maximise the value of each customer relationship.

􀂔 Merchandising: Optimising the selection, placement and promotion of merchandise among geographies, store locations and store displays.

􀂔 Operations: Optimising the behind-the-scenes aspects of retail business, such as real estate decisions, staffing levels and IT portfolio management.

􀂔 Performance management: Assessing performance from the individual store to the whole enterprise and understanding where changes will yield the greatest progress toward strategic goals.

Analytics in marketing and customer relationship management
An organisation spends tremendous amounts of money designing and delivering campaigns to reach specific target audiences. Does the management know how effective those campaigns are and what factors determine success? Is it delivering the best possible message to the right people through the right channel, at the right time? Are campaigns designed for shortterm, one-shot gains or for maximising long-term customer value? Are campaigns coordinated and integrated across channels?

Marketers can no longer view their customer audiences with a product-level perspective or as a snapshot in time. To maximise return from each campaign and customer relationship, retailers are recognising that itís time for a broader approach. It isessential to understand and appeal to customers as individuals with known preferences and buying habits. Analytics make this customer-centric vision possible. A host of analytic tools are available that enable marketers to fully understand their diverse audience segments, assess and maximise the lifetime value of each customer relationship, model what-if scenarios, predict behaviours and optimise marketing communications. For example:

􀂔 Customer profitability analysis projects the initial sales curve and lifetime value of a customer relationship, enabling more effective use of marketing, sales and service investments.

􀂔 Channel usage and profitability analysis assesses and predicts the most suitable and efficient channels for each contact activity and each customer.

􀂔 Product preference and profitability analysis assesses value and ROI on a product basis across customer groups and channels.

􀂔 Bundling/cross-selling/up-selling analysis identifies products that complement each other or will sell well together.

􀂔 Customer loyalty/churn analysis identifies which customers are loyal, which are likely to leave, when they are likely to leave and what factors influence their decisions to stay or go. All this information helps retailers devise better strategies to keep customers.

􀂔 Demand forecasting generates reliable estimates of short, medium and long-term demand so that services, products and distribution plans are always in place to meet customer expectations.

􀂔 Market-basket analysis assesses links and patterns in the mix of choices/responses that a customer makes with a view to improving cross-sell/up-sell opportunities, improving product introductions, maximising browseto- action conversions on websites, and using loyalty promotions to increase retention.

􀂔 Customer segmentation analysis divides the market into groups that share common characteristics to support manageable, accurate, time-based market response propensity models.

􀂔 Event-trigger analysis reveals correlations between events, such as demographic changes or holidays, and the implications of those events.

􀂔 Marketing optimisation incorporates information about customers, offers and channels; factors in business objectives and resource/channel constraints; and calculates the optimal mix of choices for a multichannel, multi-offer campaign or set of campaigns.

By understanding customers better, retailers can create better-defined targeted campaigns, reduce expenses (printing, paper, postage) while increasing response rates, revenues and gross margins. As retailers gain a better understanding of customersí buying behavior, analysis can then be used to create more effective merchandising plans for the next season.



Analytics in merchandising
Which items should be stocked, in what sizes and colors, at what quantities, in which stores? How, where and when should products be displayed, priced, promoted? Traditionally, such decisions have been based on intuition and historical information from simple planning applications, both of which are less-than-ideal tools for combating intense competition and shrinking profit margins.

Because merchandising processes in retail are cyclical, it is vital that the output from one process be used in another. For example, the customer data used to create merchandise and assortment plans may also be used in allocation and space plans.

Analytic solutions for merchandising apply a rigorous, objective methodology to this cyclical process to help manage variability in supply and demand and to support optimal decisions about assortments, allocation and space planning. For example, by augmenting existing systems with analytic capabilities such as forecasting, optimisation and data mining, retailers can:

􀂔 Determine how to meet sales, revenue or profitability goals under anticipated conditions that are based on storesí past, present and future demand; time; and merchandise hierarchy.

􀂔 Analyse store-specific needs and quickly respond to emerging business trends in order to maximise inventory investments, allocation and replenishments while reducing liabilities.

􀂔 Build the ideal breadth, depth and visual appeal of product assortments to match customer needs, ensure a consistent shopping experience, make best use of available space and meet financial goals for individual stores, clusters of stores and the company as a whole.

􀂔 Effectively predict the success of planned promotions and their impact on demand of featured products, as well as the impact of promotions on other products and categories.

When a retailer has in-depth analysis of past performance combined with plans and forecasts of future customer demand, he can more accurately allocate and restock merchandise across channels and stores. Truly understanding customer demand patterns not just what was purchased, but what those patterns reveal about future potential enables the retailer to send the correct assortments, size and case-pack distributions to the correct stores.

Daily price, promotion and markdown optimisation ensure that items are priced for optimal profitability, both preseason and in-season. Space automation and optimisation ensure that departmental sales and profit per square foot are maximised, and that products are given the correct inventory and space on the shelf. Optimised fulfillment ensures that products are allocated or replenished according to demand. Accurate analysis also results in a more efficient use of manpower in picking, packing and shipping the first wave of product while minimising additional expenses.

Analytics in operational optimisation
In-store and customer-facing activities rely on a multitude of support functions behind the scenes all of which must also be optimised. For example, now that analytics have given the retailer an accurate forecast of demand by hour, by day, by location, by promotion and by price change, this knowledge must guide decisions for inventory replenishment, as well as for staffing on all store floors, catalogue call centres and fleet crews delivering orders from distribution center to stores.

That's just one example. In each operational area, retailers need to answer complex questions: How do Ialign resources with corporate strategy? Which locations will provide the most profitable return on real-estate investment? How can I leverage IT investments for maximum value? Operations intelligence solutions based on analytics enable retailers to answer those questions more effectively and profitably. For example, by using analytic capabilities to delve into operational data, retailers can do the following:

􀂔 Deliver predictive insight into supply chain costing, financial planning and activity-based costing.

􀂔 Plan more effective staffing strategies for all areas of the organisation.

􀂔 Enable the organisation to realise the full potential of each IT resource through proactive planning.

􀂔 Establish the most effective supplier strategies, based on a multitude of interdependent factors.

Without analytics, a typical operations report might tell the retailer how many units of a given product were sold through each outlet or inventory levels for a specific product at various locations over a given time period. Such information provides a useful rear view into operational performance, but not a road map on which the retailer can confidently guide the business forward.

By bringing analytics into the picture, the same foundation data could reveal why the products sold better at Region I locations than in Region II, what pricing modifications would produce the best combination of customer loyalty and business profitability, the anticipated impact of a specified promotion or merchandising strategy, and what would happen if the retailer adjusted any factors, from number of drivers and vehicles in the delivery fleet to product placement on store shelves.

Analytics in performance management
With retail outlets each responsible for their share of organisational success, thereís always the danger that strategies serving the local good could undermine higher-level goals. Or that a promotion that boosts sales of one product could cannibalise sales of another. Or that strategies designed to increase short-term profits could undermine long-term profitability. With performance management analytics, a retailer can align day-to-day decisions with goals and initiatives across the entire value chain and for the entire organisation.

Performance management analysis uses balanced scorecard methodologies to align diverse business processes toward shared goals, communicate those goals across the enterprise and measure progress toward achieving targets. This type of analysis quickly identifies areas where one marketing activity might be eroding others, or where product-level successes do not contribute to overall company success.

A consistent performance management process enables the organisation to fully understand how business processes are performing and where trouble is brewing. Retailers can then better align investments, people, infrastructure and capital with overall business strategy in ways that deliver expected results and meet overall objectives.

Analytic business intelligence
The portfolio of available analytic processes targeted for retail organisations is extensive, but the real intelligence story is much more than a shopping list of discrete point solutions. True business insight is about more than making smart investments in individual technologies. Itís about what happens when those individual technology areas come together into a synergistic system. Retail intelligence success stems from an integrated suite of applications and technologies working together from a common data foundation to create a unified perspective, generating consistent advantage within a climate of constant change.

The retailers getting the most significant returns on their investments are those that take a purposeful, pragmatic approach, establishing an intelligence platform on which they base all other enterprise business intelligence solutions. A single, reliable demand forecast, for instance, can also be used in merchandising, marketing, logistics, store operations or call center staffing for operational benefit.

Business intelligence that remains segmented by functional area can provide some value, but retailers gain much more value from the same IT investment when those functional areas operate from a shared, cohesive foundation. The requisite foundation is the bedrock of solid data management capabilities designed to ensure that analysis starts with the best quality data. In the ideal IT framework, a unified, integrated data repository stores and manages all relevant data for the interdependent arena of retail activities including data from disparate databases (such as merchandising, inventory management and marketing), proprietary tools and external sources (such as purchased demographic or market data).

Sophisticated data management processes transform operational data into cleansed, consistent, structured data in a form suitable for detailed analysis. This data management process is more than simply integrating data from disparate sources; it applies embedded rules that ensure data quality, so users can have faith in the accuracy of plans, reports and analyses based on that data. Common metadata (the information about how data values are derived and used) enables the system to readily use data from across functional areas and adapt easily to business changes historical and ongoing.

The right solution must be able to integrate with any other system or platform and take full advantage of existing IT infrastructure investments. For example, if a retailer wants to use customer behaviour data to make better merchandising or marketing decisions, the retail intelligence solution must interface with sales transaction systems, loyalty systems, in-house credit systems, coupon redemption systems, catalogue and internet customer data systems, regardless of operating system or hardware.

This integration must be a two-way street. There should be a closed-loop, continuously improving process between the operational systems that transact day-to-day business and the business intelligence systems that help guide that day-to-day business to maximum efficiency and profitability.

The future retail landscape will be defined by the retailers that know how to maximise customer satisfaction and profitability with the right combination of quality products, friendly and efficient service, unique value, differentiated shopping experience and a business model that truly serves local and global communities.

How will this be accomplished? It starts with understanding the customer and then linking that insight into every decision thereafter, from merchandising and marketing to distribution, store operations and finance, so retailers can predict how best to serve their customersí ever changing needs.

Courtesy - SAS