Cloud-enabled retailing

Key design decisions for the retail industry

In this article we explore using the cloud to address some key operational challenges in the retail industry. Arguably, the cloud may not be the answer for all challenges, but we explain how many of them may eventually become strategic advantages. The views expressed belong to the authors and are not necessarily that of IBM.

Share:

Rahul Chenny, Senior IT Architect, IBM

Photograph of  Rahul ChennyRahul Chenny, is an IBM Certified IT Architect with over 19 years of experience. He draws experience from several industries and in the recent past has led Smarter Commerce engagements in retail, travel and transportation, and telecommunications. He's a member of the IBM Academy of Technology and heads the TEC-India forum. He's also the founder of Synergy Toastmasters and a special interest group in Performance Engineering (PE SIG). He currently is an industry lead architect for retail industry solutions.



Mayank Jain, Senior Managing Consultant, IBM

Photograph of Mayank JainMayank Jain, is a IBM Certified Senior Managing Consultant with over 13 years of consulting and management experience. His primary focus has been the retail industry. He has lead engagements for Operations Transformation, Smarter Commerce, and Technology Strategy. He currently is a consulting industry lead for retail.



Harinath Parameshwaran, IT Architect, IBM

photograph of  Harinath ParameshwaranHarinath, is a IT architect with over 14 years of experience in IT industry with focus on application architecture. His experience in the retail industry is with store operations and point of sales as specialization. He has hands on experience in customization and implementation of NextGen point of sales for large retailers.



Shashank Gopalakrishna, Industry Architect, IBM

Photograph of  Shashank GopalakrishnaShashank Gopalakrishna, is an Industry Architect with over 10 years of experience. He is experienced in Retail, Travel & Transport, Smarter City and Public Safety solutions. He has expertise in cross brand IBM Software products and specializes in solutions involving IBM software products.



Rambabu Parvatina, Principal Consulting Architect, IBM

Photograph of  Rambabu ParvatinaRambabu, is an IBM Certified IT Architect with 11 years of experience in IBM. Rambabu is a principal consulting architect for private and service provider cloud platforms in GTS Cloud SSA. He is a member of IBM Academy of technology.



17 October 2012

Overview

Retail, though one of the oldest industries, continues to experience costly, complicated, and cumbersome processes related to store IT operations and expansion. They impact launching new stores, the cost of store IT infrastructure, and integration of distributed information for effective analysis. Furthermore, despite the maturity of the industry, there has been limited innovation in key solution components such as point of sale (POS), in-store server monitoring and management, power management, and availability of key information to provide real time data insights.

An apt solution will enable organizations to realize simplified store infrastructure and quick turnaround time for new store openings. Additionally, the solution need to organize data through simpler integration mechanisms to derive information insights that can affect customer churn, enable smarter business decisions, and enhance collaboration between supply chains.

In this article we will address the aforementioned scenarios by leveraging increased reliability of Internet connectivity to that of a utility like expectation. We'll discuss the following business and technology benefits of cloud-enabled retailing:

  • Simplification of store IT
  • Expedited store launch
  • Retail solution as a service
  • Faster and better business decisions
  • Improved visibility through a hybrid cloud

We conclude this article with several options for a phased approach for transformation onto cloud and a solution approach using IBM SmartCloud Offerings.


Challenges faced in the industry

About the retail industry

Retail comes from the old French word “tailor”, which means to cut off, clip, pare, or divide. Later it evolved to mean sell in small quantities. Retail could happen through department stores, discount stores, specialty stores, and even seasonal retailers. Product range varies from grocery to automobile, apparel to home furnishings. Each of these might have their own unique characteristics, however, for most of this article we are referring to retailing in general.

In the last couple of decades, there have been sweeping changes in the general retailing business. For example, what was once strictly a made-to-order market for clothing has changed to a ready-to-wear market. It was a common practice to flip through a catalogue, pick the color, size and type of clothing a person wanted to purchase and then wait for it to be sewn and shipped. At the turn of the century some retailers ventured to an e-Storefront where people could browse, select and purchase. Meanwhile, new pieces were being sewn or customized in the back rooms.

These changes, along with fierce competition, have fueled the need for continuous improvement, flexible operations, and availability of information on demand.

Operational complexities

Operational complexities is best illustrated by a recent consumer product company launch of a global confectionary range in India. Due to operational complexities, it took them nine months from product conceptualization to launch. This duration was more than enough time for a nimble local competitor to launch a similar product a few months earlier and the rest is history. The global major is constantly in a catch-up mode now. This is a classic example where asynchronous collaboration between the parent company and the subsidiary incurred a huge opportunity loss. The intra-company collaboration breakpoint was mostly data dependent.

The retail industry is primarily end-consumer facing and when you consider how many consumers are out there the resulting number of transactions leads to unprecedented volumes of data. Leveraging this data to streamline operations is key for success. However, getting that data on demand is challenging.

While the number of transactions and associated data generated is huge, there is a greater need to access it real time, and to make smarter business decisions.

Broad challenges

Customer preferences are changing very fast. Lack of understanding of customer behavior and preferences is like jumping in water without knowing the depth. Data analysis gives insights into how customer behavior translates into shopping habits and what products and services the customer is likely to buy. Therefore, the first challenge is to continually develop customer insights by analyzing real time data. Subsequently you plan and expedite product and service development to satisfy customer demands in a timely, convenient and cost effective manner. These challenges could be the opportunistic differentiating factors for competitive advantage, if the organizations are able to:

  • Continually reduce cost of operations for higher margins and invest back in customers.
  • Increase their capability to manage high volume of data to draw closer references and hence faster decisions
  • Automate processes for more efficient operations and predictable outcomes
  • Build flexibility in operating model to leverage opportunities to service clients with innovative products and services
  • Provide access to information anytime, anywhere for stronger control on the operations, faster decision making and reduced set-up time

Cloud, the buzzword in the town

Given the operation complexities and broad challenges, there has been a constant endeavor to evolve further. Evolution in the computing paradigm mandated unprecedented demands on the infrastructure for computing resources. The natural response was to expand the infrastructure footprint in the data centers to address the need for growing computing power ranging from increase in transaction volume to deployment of new IT services. But was this effective and optimal utilization and were there alternatives to the ensuing approach particularly when the peak demands were of limited duration? Could the infrastructure be more agile or dynamic?

Cloud computing, provides a solution to realize an optimized approach and address a variety of computing infrastructure needs related to IT services. One of the defining attributes of cloud computing is elasticity in that it offers additional computing power to handle growing needs. Cloud computing benefits are accentuated by the fact that systems resources can be provisioned on demand, and favorably affect the capital expenditure towards pay as you use (operating expeditures).

Cloud service models

Definitions

  • Private cloud - A cloud in a private network, for example a company owned cloud.
  • Public cloud - A cloud on the internet with services for public consumption.
  • Hybrid cloud - A cloud that combines the private and public cloud and typically enables collaboration.

To address a variety of needs, cloud computing supports a number of service models based on the computing ask (see Figure 1). In infrastructure as a service (IaaS), the core need or service is for computing infrastructure and typically entails provisioning hardware and operating system. In the platform as a service (PaaS), the service in context builds on IaaS and additionally entails middleware, runtimes, development, and tooling. Software as a service (SaaS) opens up the possibility of a new business model as it typically provides multi tenancy of services to consumers similar to an apartment complex providing amenities to its residents. Some common examples of SaaS services include applications, business processes, and utility services. SaaS, the most sophisticated among these three delivery models, can be configured for a cost to end consumers based on flat fee, transaction value-based fee, or data volume. Cloud also offers business process as a Service (BPaaS), which delivers business processes as a service.

Figure 1. Cloud service models
Cloud Solution Approach

It is important to note that the cloud service models encompass both functional and non-functional requirements.


Cloud for retail can be a win-win proposition

Retail reaction

Retailing has matured into unimaginable realms that vary from increased customer intimacy to social retailing, and has grown increasingly dependent on information technology to maximize profitability.

Retailers are adapting to the trends and evolving their IT to address this need by purchasing new IT solutions, developing complex custom solutions, and leveraging data centers to host and mange the operational needs. They are also sizing their hardware to address peak transaction loads during peak business events. In the broader schemes of things retailers are also dealing with the inevitable outcome of investing in technology, obsolescence.

Undoubtedly retailers are reacting to close the needs gap. A quick examination of the reaction will reveal the dichotomy in focus, that is the business of retail and business of IT. The latter is not a core competency, but is necessary to evolve and presents opportunities for optimization.

Motivation for convergence

An apt solution will enable retailers to realize simplified store infrastructure and quick time to market (IT enablement) for a new store from several months to a few days. Additionally, it should organize data through simpler integration mechanisms to derive information insights that can affect customer churn, enable smarter business decisions, and enhance collaboration between supply chains.

We believe that there is an opportunity to address the aforementioned scenarios by leveraging cloud computing and IBM SmartCloud offerings. The following are some business and technology benefits of the suggested approach:

  • Simplification of store IT

    Cloud computing can breakaway from traditional and expensive, thick POS client applications and create POS on cloud thereby liberating from complex and expensive upgrades. Additionally, deploying remote monitoring may reduced capital and operating expenses.

  • Expedited store launch

    Cloud technologies and service models can centralize store operations and instantiate new stores leading to savings of months of effort.

  • Technology as a service

    Small retailers can significantly benefit using cloud by applying a pure operational expense-based model and leveraging best of breed solutions.

  • Faster and better business decisions

    Moving data aggregation across a wide area network to that within the cloud enables faster analysis of information related to sales, transaction and POS logs, and more.

  • Improved visibility

    Improved collaboration with supply chain partners can improve demand visibility and customer service. For example, real time inventory information can help avoid running out of stock and improve overall customer satisfaction.


Solution: What, how, and where

Solution description

Given the background of opportunities and challenges, cloud presents an evolutionary alternative. Extending an enterprise on to the cloud is undoubtedly a significant decision for any retail organization. You have a variety of options to consider, such as:

  • Deploy your own private cloud for internal IT optimization. (Large retailers may take this route by leveraging IBM SmartCloud Foundation offerings.)
  • Leverage cloud services provided by organizations such as Amazon, and IBM SmartCloud services and solutions. (This option may be lucrative for small and medium businesses.)
  • Build a private cloud and leverage or consume cloud services from the vendors on other clouds. This approach is also known as a hybrid cloud.

Among the other decisions related to cloud, a retailer needs to identify, or assess, what you can, or should, deploy on the cloud. This assessment may be performed by the retailer's IT team or by a cloud consulting organization that provides cloud strategy and migration of services. It's best to understand the associated workload or characteristics of the existing or proposed systems. You could approach this from the bottom up (traditionally, more accurate) or from the top down (the approach of this article). Some common workload characteristics include peak loads (transaction), computation intensive (driven by analytics), integration needs (currency of data), and data volatility (static versus fast changing). Knowing the workload helps you identify a rationale to select a specific service model.

For a precise outcome, cloud consulting vendors engage with clients to identify or recommend a service model. They also employ a mature and often proprietary methodology to drive data collection and analysis. However, for this article, we reviewed the workloads characteristics for specific business areas and recommend the potential service models to overcome operational complexities, and as shown in Table 1.

Table 1. Map of business function workload to potential cloud service model
Business functionBusiness componentsWorkload characteristicsPotential service model
Demand forecasting (across SC)Demand forecasting, POS execution, replenishment, purchasing / sourcing, returns, and reclamationCompute intensive(trend analysis), periodic workload, large data sets, collaboration and integration needsPaaS/SaaS
AnalyticsCustomer analytics, merchandise analytics, supply chain analytics, corporate audit, sales auditCompute intensive(trend analysis), periodic workload, large aata sets, internal, and few integration needsPaaS/SaaS
Order captureMultichannel, price, product, catalog, promotion, member, loyalty, payment, marketingVarying/elastic workloads, external and high integration needs, automated provisioningSaaS
Store operationsPOS systems, back office operations, cash managementBusiness critical, data intensive, varying/elastic workloads, desktop managed, high users growthSaaS

High-level architecture diagram

Having identified the workload characteristics of the IT systems in the enterprise, one gets a snapshot of what is as-is recommendation of a cloud service model, which can be further influenced by other factors. For example, a factor could include the cloud service provider’s operating cost assessment and recommendations for additional optimizations.

Having identified the workload characteristics of the IT systems in the enterprise, one gets a basic understanding of what cloud service model to adopt. This outcome can be further influenced by a cloud service provider’s assessment related to operating cost and risks and recommendations to realize optimizations.

Figure 2 shows a high-level solution approach of how business functions can be realized on cloud.

Figure 2. Cloud solution approach
Cloud Solution Approach

(View a larger version of Figure 2.)

Figure 2 presents a limited depiction of a cloud environment for a retailer with the various solution components drawn from IBM Cloud offerings.

Offerings in SaaS and Paas delivery model:

  • DemandTec Lifecycle Price Management: A comprehensive solution for managing pricing across the entire product lifecycle through cloud-based applications. A retailer can strategically price items (new items, regular retails, promoted items, and clearance items) at any stage in the lifecycle.
  • DemandTec Assortment Optimization: DemandTec Assortment Optimization not only leverages the power of shopper insights as the platform for advanced analysis, but also seamlessly integrates other key assortment factors, such as market data, space, and profitability, to ensure actionable and relevant assortment recommendations by store.
  • DemandTec Shopper Insights: DemandTec Shopper Insights is delivered on the cloud-based network. Retailers can provide their trading partners with immediate access to these insights. While the retailer remains in control regarding the scope of access, the manufacturer can leverage the same information to develop better trade plans, marketing plans, and product offerings.
  • IBM Sterling B2B Collaboration Network: IBM Sterling B2B Collaboration Network is a cloud-based, B2B integration as-a-service solution providing secure connectivity and collaboration with customers and business partners. It delivers unprecedented visibility and control over the business processes shared with outside companies.
  • IBM Commerce on Cloud: This SaaS option opens up opportunities for companies of all sizes, from small and medium sized businesses to large enterprises, to quickly set up and maintain an online storefront while delivering a superior customer shopping experience similar to the largest retail brands in the world. It addresses extended e-commerce platform requirements by bringing together product functionality from across the IBM Smarter Commerce portfolio.
  • IBM Coremetrics Web Analytics: This cloud-based digital analytics product provides real-time key performance indicators, benchmarks, and intelligence to inform about average peer and competitor performance data. Its marketing attribution capabilities help improve marketing spend allocation. An associated campaign management product, the Unica suite, offers a solution to optimize email communications from creation, deployment, social sharing, and landing pages (IBM Unica Email Optimization). Unica also includes solutions to empower marketers with tactical campaign tracking, accountability, strategic program collaboration, and asset management across the company (IBM Unica Marketing Operations OnDemand).
  • IBM Sterling Supplier Portal: This allows companies to create and exchange business documents electronically with trading partners, regardless of their size or technical expertise. IBM Sterling Supply Chain Visibility helps optimize the inbound supply and outbound fulfillment processes through real time, end-to-end visibility across supply chain networks.
  • IBM Sterling Order Management Service: This enables companies to manage the order lifecycle and orchestrate complex order fulfillment processes across a dynamic business network from a single-tenant SaaS architecture.
  • IBM Cognos Business Intelligence: As shown as a PaaS in Figure 2, this is a key component of the IBM Smart Analytics Cloud. It enables the delivery of business intelligence and analytics in a private cloud deployment and reduces the number of departmental solutions to a single business intelligence environment. It supports a self service approach to sharing business intelligence and analytics services that reduces the time, resources, and costs for delivering services to new divisions, departments, and users.

Offerings in IaaS model

  • IBM SmartCloud Enterprise: This delivers secure and scalable hosted IT infrastructure with on-demand access to virtual server and storage resources. It is a dynamically provisioned and scaled elastic environment that provides an environment to develop and test application code. It includes a web portal to configure and manage the cloud resources, software images of IBM products, and APIs that to control cloud resources and software programmatically.
  • IBM SmartCloud Enterprise+: This is a fully managed, security-rich and production-ready cloud environment designed to ensure enterprise-class performance and availability. It offers complete governance, administration, and management control along with service-level agreements (SLAs) to align your specific business and usage requirements. Multiple security and isolation options built into the virtual infrastructure and network keep this cloud separate from other cloud environments.
  • IBM SmartCloud Foundation: This family of technologies is designed to help organizations quickly adopt private and hybrid clouds. SmartCloud Foundation solutions enable an IT delivery model that helps you shift from the limited focus on virtualization's technology efficiency to providing the flexibility to support end user productivity, business agility and greater business value for IT.

Store operations on cloud

Store operations are of paramount importance to retailers and have a profound impact on the entire supply chain ecosystem and revenue inflow. Managing store operations requires extensive IT systems such as POS, stock replenishment, and back office operations, which all play critical roles and demand a high capital expenditure as they need to be installed across the entire retail chain.

Modern day retailing and evolving technology challenge retailers to re-think store related IT in innovative ways to manage store operations. Traditionally, the POS systems, are built as desktop based rich user interface systems with other solution components including core business logic related to sale. The latest shopping trends with variety of touch points such as self checkouts, mobile shopping, kiosks and multi-channel shopping requires the existing business functions to be re-used, which is very challenging.

The POS system is complex by itself with its own controller, server, registers, and sets of peripherals. It demands heavy infrastructure, and in the case of large retailers, large capital expense. In reality, not all the registers are used on day-to-day basis and when they are their use is dependent on the customer inflow in the store. The initial investment may be minimized by enabling POS in cloud and using less expensive touch points, such as mobile POS, which result in reduced infrastructure costs while providing greater flexibility and agility for store operations.

Because of the complex IT store infrastructure, time consuming set up process, coordination with various teams, and efforts to load the huge inventory data, retailers often find it challenging to launch a new store swiftly, which impacts the business growth, especially when it comes to opening new stores in strategic locations amidst the competition. The labor cost of a number of technicians in the fields also adds to the expense too.

Benefits of store operations on cloud

Delivering store operations on cloud can help overcome several challenges. Figure 3 illustrates a solution approach through the SaaS delivery model.

Figure 3. Cloud store solution
Store on cloud solution

(View a larger version of Figure 3.)

As shown in Figure 3, POS when delivered from cloud as SaaS eases enabling a variety of touch points and centralizes the backend business logic. Additionally, the efforts required for setting up, upgrades, monitoring and management of the POS software, hardware, server, and network setup is dramatically reduced due to the nature of centralized point of control. Setup of network and end point device systems is highly simplified, helping launch a new store quickly, with low manpower. Because the inventory data is centrally maintained in a cloud environment, retailers can perform data setup and inventory loading operations prior to store setup from a centralized management portal, which expedites the processes involved during store launch. This simplifies the store network and the systems to be installed in the store and allows the retailer to launch a new store swiftly using less manpower.

Apart from the reduced infrastructure costs and other benefits we previously mentioned, when POS is delivered through cloud, the transactional data is instantly fed to enterprise systems, which provides a wealth of information for critical decision making. An additional benefit is that services are easily extendable beyond traditional registers and can be reused in other business functions. ARTS-XML Retail transaction interface (RTI) standards provide a consistent way for the cloud provider to enable POS functions

Other areas of store operations, such as back office operations, can be monitored and managed effectively when deployed on the cloud because it provides greater visibility of the entire chain for a given business day and can be managed with minimal personnel.

Generally, stores are located across multiple geographies with distributed back office operations and have a high penetration of POS usage, having a centralized control would greatly benefit the retailers.


Transformation: Phased approach considerations

Table 2 shows several considerations to initiate a move to the cloud in a phased manner and are as shown below.

Table 2. Considerations to move to the cloud in phased manner
Service modelEvolving through the service models, IaaS, PaaS, SaaS. A retail organization may chose to start the cloud journey by exploring any one of the options described in the solution description section. The phased movement assumes a detailed assessment of the workloads.
GainsPrioritization based on the operating cost benefits, that is the phased movement may consider the savings to initiate the phased move to the cloud.
Cloud ready workloadSome workloads that are less critical are more cloud ready, such as:
  • Development and test workload
  • Elastic workloads - B2C and B2B
  • Numerical computation – Trend analysis, analytics
  • Collaboration – demand forecasting
Business function affinityCo-deployment of solution components that have dependencies and hence form a basis for consideration, for example order capture and order fulfillment.
Cloud service provider servicesCloud service provider provision of services that benefit the retailer. Some instances include PaaS services that enable development and test, for example Rational tooling. SaaS services to address business functions such as campaign management, web analytics, and order capture.

Before adopting any phased approach a retailer must work with the cloud service provider to understand the assumptions, risks, service level agreements, and standard operating procedures. This will help strengthen collaboration and set the platform for managing cloud-based expectations.

In addition to transformation the following considerations also affect a retailer’s decision to deploy a solution on the cloud:

  • Regulations stemming from concerns regarding data security and privacy

    Some government organization mandate that specific IT systems should be deployed within the geographic boundaries while others disallow storage of personal data such as credit card details in systems physically deployed outside the country. In the latter case, a way to ensure compliance may entail having a different hosting location for the application tier and the resource tier. It is best to measure this decision against other considerations, such as performance (network latency).

  • Security concerns

    Owing to collocation of data, a security compromise in a SaaS delivery model would have a higher impact then that of a PaaS or IaaS delivery model. In the former case data from multiple organizations though logically isolated is physically collocated.

  • Affinity to on premise enterprise information systems

    We are referring to those applications deployed on the cloud that have high throughput requirements, and depend on systems hosted in the data centers or client premise to complete the business transaction. Consider an inventory verification in a checkout flow, where the availability of a product needs to be verified at several steps such as add to Cart, and submit an order. If the inventory management systems are not optimally deployed such transactions could impact the overall throughput of the order management system.

  • On demand needs

    This relates to the need to drive faster change, flexibility to automate provisioning. Additional proximity needs for systems management activities such as restart a system on demand.

  • SLA support

    Validation of cloud SLAs to ensure they match with workload SLA requirements.

  • Software support on hypervisors

    Software solution components must be capable of being deployed in virtualized environments.

  • Non-supported OS platforms

    Workloads running on OS platforms that are not supported in a cloud environment. A retailer must verify whether the commercial software product can be deployed on virtualized environments.

  • High compute/memory capacity

    Workloads that have constant high compute/memory capacity needs, which are greater than that provided by a cloud solution.


Conclusion

In the foreseeable future, the customer will become more demanding. Retailers, in their endeavor to service customers better, will have to:

  • Enhance their capital efficiency to reprioritize more customer experience centric initiatives.
  • Improve their operations efficiency to free up resources invested elsewhere and reinforce customer centric resource investment.

To achieve this, organization will go far and evolve multiple initiatives, a prominent one being adoption of cloud for their technology platform. Cloud services will help make organizations more nimble, turn focus from capital investment to operation expenditures, and reduce time to market; contributing towards improving their capital efficiency. Bringing information on demand for faster decisions and building more collaborative ecosystem of suppliers, vendors, retailers, and customers will help improve operations efficiencies and thus allowing for better competition and focus on customer value.

The journey to explore the art of possible has just begun!

Resources

Learn

Get products and technologies

Discuss

Comments

developerWorks: Sign in

Required fields are indicated with an asterisk (*).


Need an IBM ID?
Forgot your IBM ID?


Forgot your password?
Change your password

By clicking Submit, you agree to the developerWorks terms of use.

 


The first time you sign into developerWorks, a profile is created for you. Information in your profile (your name, country/region, and company name) is displayed to the public and will accompany any content you post, unless you opt to hide your company name. You may update your IBM account at any time.

All information submitted is secure.

Choose your display name



The first time you sign in to developerWorks, a profile is created for you, so you need to choose a display name. Your display name accompanies the content you post on developerWorks.

Please choose a display name between 3-31 characters. Your display name must be unique in the developerWorks community and should not be your email address for privacy reasons.

Required fields are indicated with an asterisk (*).

(Must be between 3 – 31 characters.)

By clicking Submit, you agree to the developerWorks terms of use.

 


All information submitted is secure.

Dig deeper into Cloud computing on developerWorks


  • Bluemix Developers Community

    Get samples, articles, product docs, and community resources to help build, deploy, and manage your cloud apps.

  • developerWorks Labs

    Experiment with new directions in software development.

  • DevOps Services

    Software development in the cloud. Register today to create a project.

  • Try SoftLayer Cloud

    Deploy public cloud instances in as few as 5 minutes. Try the SoftLayer public cloud instance for one month.

static.content.url=http://www.ibm.com/developerworks/js/artrating/
SITE_ID=1
Zone=Cloud computing, Commerce
ArticleID=840249
ArticleTitle=Cloud-enabled retailing
publish-date=10172012