Enterprise Applications

Enterprise apps have been talked of for quite some time now, but still remain an enigma for many. So, instead of the standard approach we have in other articles, we look at what they are and how they work

The term ERP is back in the reckoning and is fast descending on to Indian medium and small organizations. The world beyond the ERP, typically called Extended Enterprise Applications- notably SCM and CRM is also ripe for adoption. The article provides a snapshot view into the world of integrated business applications and places them in perspective. 

ERP (Enterprise Resource Planning) 



The simplest definition of an ERP: a software solution that takes the process view of an enterprise to meet the organization’s goals while integrating and interlinking the various functional areas of the business. As a result, it performs the role of an enterprise-wide information system that tracks all business transactions, internal or external, and is therefore considered the transaction backbone of the business. 

The software that ERP is, be it SAP, PeopleSoft, or JD Edwards to name a few, helps mimic the best business process for the particular organization because it is highly configurable. Designing the business process is a separate exercise, but is the starting point of an ERP implementation and is typically handled by process consultants. Technically, the ERP software comes in the form of various modules, some of which are technical components, while most of them are function-specific modules in areas of material requirement planning, bill of materials, master scheduling, shop-floor control, accounts payable/ accounts receivable, sales, distribution, financials and the like. 

The ERP sits atop a database, usually a relational database that is scalable as per requirement. At a technical level, the ERP software integrates the relational database, the base ERP module, the various functional modules opted for, other applications that the organization might have, and tools and interfaces between the applications and the database- in such a manner that the business process that was designed is satisfied. Typically, the design is done from an ‘events-based’ philosophy. That is, “ if such-such action happens, such-such is the action that has to succeed”. 

Inherently, therefore it helps tackle day-to-day business problems like excess of inventory, shortage of inventory, gap between planned output and realized output, unutilized cash beyond certain limit and potential working capital crunch. These are all identified and suitable actions can be taken. One reason the benefits of an ERP is considered to be real is that the benefits are for the whole enterprise rather than for a part of the enterprise, assuming that the ERP has been deployed across all the sites of the organization and across all the functions. Looking at it the other way, the entire enterprise gets aligned to handling processes in the same manner and the performance of the business gets reported from one place. 

There are two choices that companies could make: build an ERP in-house or go for a packaged product. The benefits derived from a packaged ERP application are far more than one that has been built in-house because the packaged ERP has most of the business logic culled from successful organizations already built-in. Besides the quality of documentation, standardization on technology, room for growth and adding more functionality, and availability of talent are the other plus points. Globally ERP vendors have been adding additional functionality and capability to their products in the areas of supply chain management, CRM, data mining, business intelligence, e-business enabling, enterprise portal, and some very industry-specific templates. Newer generation ERP that enables collaborative commerce is called ERP II. 

Amongst the various vendors of ERP in the market, there are ones that are suited to large organizations and others for small and medium companies. ERP packages also differ in their ability to address different industries and industry sub-segments. But the essence of all these products remain the same - to provide and integrated suite of business applications. 

Typical Steps to an ERP Implementation 

• Project planning
• Business and operations analysis
• Business process re-engineering and design
• Project team training
• Business requirements mapping to software
• Configuration of modules
• System modification and build ing interfaces
• Data conversion 
• Custom documentation
• End-user documentation and training
• Pilot
• Acceptance testing
• Production and rollout
• Post-implementation support 

SCM (Supply Chain Management) 



The supply chain in a business starts with the procurement of raw materials on one end and ends with the delivery of finished products to customers on the other. Depending on the nature of the business, managing the supply chain can get quite challenging and complex. Managing the supply chain involves making the right type of decisions along the supply chain. SCM includes business processes, information flow, information processing, and closing the loop between raw materials and finished goods. 

SCM software comes in four types: supply-chain planning software, supply-chain execution software, specialized sourcing and procurement software, and a combination of these which vendors like i2 Technologies call end-to-end supply chain solutions. While supply-chain planning software uses advanced mathematical modeling and forecasting methods for planning the activities across the five parts of the supply chain, the supply chain execution software automates the different steps along the supply chain. The most powerful part of an SCM solution is ‘demand planning’ which determines how much product one needs to make to satisfy customer demand, under all possible external constraints/uncertainties. 

Any supply-chain software depends heavily on external inputs like customer orders, sales data, stock positions at warehouses and the like and internal inputs that come in from various functional systems like sales, manufacturing, and finance. While this is possible with a legacy system, it is better to have all the information coming in from an ERP application. That is, while ERP is strictly speaking not a pre-requisite for SCM, it is better to have one before you embark on an SCM project. 

Supply chains extend beyond the boundaries of the organization to suppliers and distribution partners. SCM software functionality therefore has extended to cover the entire spectrum of this extended enterprise. All of the SCM solutions available are Web-enabled. Mature organizations have been able to connect their global supply chains together using a single SCM system. 

Choosing an SCM product is a very challenging task. Before choosing an SCM product, the company has to build the entire supply-chain process diagram, after suitably re-designing key business processes. In the next step involving data flow diagramming, input and output to and from the various processes in the supply chain, including the feedback loop are decided.Key areas like demand planning, scheduling, distribution planning, inventory management, fulfillment planning, outbound logistics, partner integration, warehouse management, transportation management are studied and are graded according to relative importance. In fact, organizations usually have a pre-decided supply chain strategy based on historical experience. The diagnosis about where the supply chain gets stuck could be done at this stage. Once the key functionalities are determined, various products are studied, compared, and piloted for their respective strengths in these areas and the final one selected based on the most suitable fit. An SCM project requires more functional experts and external consultants than technical professionals to see it through successfully. 

When SCM started as a practice in 1990, vendors began to leverage technologies such as in-memory modeling and advanced optimization algorithms to provide customers with tools that went beyond transaction support and sub-optimal planning solutions. These solutions largely fell into two areas: supply-chain planning and supply-chain execution. Although many of these initial solutions have matured to the point of commoditization, new applications and enhancements to current solutions are maturing to provide more cost savings and to enable more-agile supply-chain structures. Many of these solutions are just beginning to evolve, and should currently be treated as immature. 

CRM (Customer Relationship Management)



Customer-relationship management, per se, is not about managing relationships with the customer. It is about managing the various bits and pieces of information about customer behavior and buying patterns, market trends, ‘what worked’ and ‘what did not’ in marketing campaigns and the like, to improve revenues, increase customer base, leverage existing customer base and to have satisfied customers. 

Some of the inputs in CRM are account information, Web-registration data, data from surveys, payment patterns, defaulting patterns, response to campaigns, demographic data, service and support records, complaint records and other myriad information or clues that a customer leaves with the vendor with at various points. 

Consequently, one of the toughest challenges in a CRM implementation is to aggregate and cleanse data from various systems into a single repository or data warehouse. This may include operational and back-end databases, spreadsheets and applications, third party databases, and different channels like phone, retail outlets, Web, ATM and e-mail responses. 

Principally, there are three types of CRM or rather three types of CRM vendors: transactional CRM which covers the area of customer transactions and managing data, analytic CRM which deals with the area of sophisticated analysis and modeling based on customer data, and collaborative CRM which covers call-center interfaces, website and other interfaces. Looking at it in another way, there are three kinds of CRM specializations: marketing automation, sales force automation, and customer-support automation. Marketing-automation CRMs optimizes an enterprise’s marketing process, with the objective of allocating resources to target markets with the highest potential value. These applications, which succeed the earlier data mining and database marketing systems, assist in the planning and execution of marketing campaigns (by managing customer and market profiles, identifying target markets with high revenue and profitability potential, generating leads, selecting appropriate contact media) and tracking initial customer contact efforts across these channels.

Sales-automation CRM manages and automates the organization’s sales cycle, increasing its productivity by accelerating the contracting process, and by improving the speed to revenue. They also link the enterprise’s sales force with the corporate office and enterprise database and facilitate improved communication between sales force, management and customers.

Customer-service CRM applications were originally developed as call center and automated help-desk systems and automates functions such as order tracking and accounts-status checks. These applications integrate with front-office applications and are capable of receiving and tracking customer requests for new product enquiries, status enquiries, and support calls from a variety of channels. Large CRM-applications suite vendors like Siebel combine all the three functions into one product and also possess some analytic and collaborative capabilities. Further, some of the CRM vendors like SAS and Teradata have industry-specific analytical tools, for example, for churn management in telecom services and fraud management in credit cards. 

EAI (Enterprise Application Integration) 



SCM and CRM are two types of Extended-enterprise applications. Despite the value they bring to the table, their full potential cannot be realized without the technological glue that integrates them to the core of the enterprise. The integration challenge, often referred to as ‘stovepipes’, ‘islands of information’ and ‘vertical application silos’ is tackled by a diverse set of tools and technologies that have evolved out of the general category of middleware and have now been grouped under EAI tools. 

Whereas traditional middleware facilitates the integration of individual applications and discrete transactions between them, EAI enables an enterprise to manage relationships among multiple applications and the surrounding network of transactions that constitute a business process. It is essential to keep in mind that it is the chaining together of discrete transactions, in the form of a business process, from one application to the next that constitutes EAI. The amount of custom coding required to effectuate EAI with traditional middleware far surpasses that required through the utilization of special EAI tools like the ones from Tibco and IBM. 

So, how is integration between applications achieved? These EAI tools operate in three layers. 

• The business process layer that automates or replicates a business process 
• The business rules layer that dic tates the rules according to which a business process is to be executed
• The data-transformation layer, which interprets and transforms a large number of data into a usable format (often
  
  XML) that can be communicated between applications. Finally, the transportation layer, which is a message-oriented middleware handles routing of messages and guarantees delivery of messages between applications. 

Currently, in distributed application architecture, these layers form an independent layer of computing called the integration layer. This is where applications ‘plug and play’. 

Why is this layer important? Because this is the layer where an SCM can talk to a CRM, where an analytical CRM tool can talk to a custom-developed marketing application, where a small ERP like Navision used by the Indian operations of an MNC can talk to an SAP used by the worldwide HQ. 

Note that at this level, new islands of information are not created, something that typically happens when point-to-point integration is achieved by authoring individual API’s between the two programs. There is more to this layer. This is called as event-driven processing. A ‘business event’ is a piece of information generated by an application (such as an order entry by a customer) that must be shared with other applications (such as a requisition system with a supplier). Middleware has facilitated the transportation of this information on a point-to-point basis for a long time with the assistance of considerable custom coding. This layer allows business analysts, not custom developers, to configure applications so that ‘business events’, once published into the separate integration infrastructure, become available to all of the other applications within the enterprise that have a need to be aware of it. Hence, the phrase event-driven processing. 

New generation EAI tools automate business processes by taking advantage of the separate tier of computing for the integration infrastructure. Simply put, rather than having business processes hard-coded into applications, the business processes are configured by business analysts into the tool itself. In other words, the business process, like the integration itself, is “decoupled” from the native application. The advantage to this approach is that new processes can be configured into the integration infrastructure on the fly without having to rewrite any of the distributed apps. Come to think of it, the new acronyms ruling the IT industry (as recent as January 2003), BPM (Business-performance management) or BAM (Business-activity monitoring) belong to this very layer. 

By Easwar S Nair 



Associate Editor, Dataquest