Conclusions


In a business-technology environment of constant and ongoing transformation, not only do business systems need to change and evolve, but decision-making perspectives do, as well. Where once the main decision for executives seeking an e-business solution was whether to build or buy, the critical issue today is finding the fastest path to fluid integration of key business processes and enterprise business systems.

In the next phase of e-business, customers want one vendor to provide all the pieces that make automated buying and selling of direct goods seamless, linking transactions to order fulfillment, manufacturing supply chains, inventory replenishment, and transportation. Customers don’t want to deal with the hassle of integrating all the disparate software pieces, costing them millions of dollars and years of work. No longer are the choices for enterprise e-business solutions limited to (1) buying more than is needed and living with a “closed” system in order to minimize surprises, time to market, and the lack of reliable support, and (2) building a system from scratch in order to achieve a custom solution, while surrendering to the variables of time and budget—with no guarantee of ultimate functionality, scalability, interoperability, or support.

Today, it’s possible to find an e-business solution that offers the best of both worlds. The best “buy” provides all the functionality that’s needed to be competitive today without requiring a business to buy more than necessary. The right system for e-business now comprises:

The right system for e-business, enabling best practices, rapid integration through fewer “moving parts” or variables, and 24-hour, 7-day customer service; this system is modular, distributed, and absolutely reliable.

The right company for now and the long run, focused on solving business problems, with a proven record of engineering excellence, with a proven, sizable customer base and the ability to guarantee comprehensive future customization to fit unique and changing business conditions[1].

Build versus Buy

As IT systems age, the Internet matures, and behemoth computing companies are left in the dust, one problem remains constant: how to find an IT solution that directly contributes to the larger mission of the enterprise—and fast. This eternal quest has been framed in different languages over the decades, but none so persistent as “build or buy?” The classic build-or-buy struggle has been ongoing for 29 years and is now expressed through three approaches:

  • Development suites

  • Point solutions

  • Packaged solutions[1]

Development Suites

Development suites allow IT departments to build whatever they need without requiring that they buy more than they need. Key challenges lie in the time and money required to build, test, and troubleshoot new systems while ensuring interoperability, scalability, and security. Furthermore, the true and total cost of the application may be difficult to calculate accurately.

Point Solutions

Point solutions focus on one specific problem each. So, in order to address larger business problems for the enterprise, additional functionality must often be added by stitching together multiple point solutions and/or development suites. Meanwhile, IT departments can find themselves left alone without support for custom integrations between changing versions of software. Support from vendors is a vital consideration when mission-critical operations are on the line.

Packaged Solutions

Packaged solutions seek to meet business challenges through software that addresses complete business problems. These end-to-end systems facilitate integration with existing mission-critical system investments and business-process modeling. These solutions are fully tested in real-world settings, undergo constant improvement, and are backed by 24-hour, 7-day support. Though all three of these approaches are very different and have their own advantages and disadvantages, at the center of all three are the following issues:

  • Openness

  • Best-of-breed

  • Scalability

  • Time to market[1]

This part of the chapter discusses the preceding decision points and also the fundamental importance of something even more critical to e-business success: ease of integration. In other words, it’s important to “Buy for the life of the site”—not its birth. That means bringing in tested functionality demanded by e-business customers today, while ensuring that the supporting vendor and the next generation of that system will be available (and work with other key enterprise systems) tomorrow.

The fact is that “buy” no longer means “one size fits all.” A genuine solution can and should provide a number of key advantages for the organization seeking to increase its agility through e-business:

  • Enterprise-ready technology

  • Enterprise-worthy functionality and support

  • Stable, partnership-oriented vendor[1]

Enterprise-Ready Technology

Without the right foundation, no e-business system is stable. The success of a system is dependent on its ability to solve a business problem while simultaneously ensuring unrivaled scalability and performance, foolproof security, and open enterprise standards to facilitate content exchange and integration with existing business-critical applications.

Enterprise-Worthy Functionality and Support

E-business waits for no one. The right system is an enterprise-grade, modular, stable, distributed, end-to-end solution that can be immediately rolled out onto the Web platform. It offers the high-quality experience customers seek, facilitates best practices, has been stringently tested, simplifies enterprise application integration, and enables facile evolution as enterprise needs change.

Stable, Partnership-Oriented Vendor

Real solutions solve problems now and later. This is made possible by working with a committed vendor with a proven record of working with its customers to meet their goals. No solution is complete without full support from knowledgeable representatives that is available 24 hours a day and 7 days a week.

Open Systems: Pathway to Freedom?

The open-systems issue is often the first to be addressed in the evaluation of potential e-business systems. Open standards do offer the possibility to more easily extend individual systems and combine disparate systems. However, when an open standards-based system leaves many business problems unsolved, its very openness can appear more like open air. Some leave room for a variety of options that must be carefully evaluated as to safety, security, and straightforward integration, but provide no up-front solutions.

It’s also important to note than many otherwise open solutions are coupled with a specific application server. This can neutralize the advantage of being free from proprietary architectures, as any additional point solutions required will need to operate with that same application server.

Development suites promise any IT group the ability to build tailored systems for each situation, thereby leaving the door open to freedom of choice in the future. However, it’s noteworthy that the organization is vulnerable to development time and expense factors that can be prohibitive when attempting to bring a finished product to market.

End-to-end systems based on proven enterprise architecture are the fruit of hundreds of hours of testing, tuning, and perfecting. The practical meaning of openness for the enterprise is the ease with which a system can be extended by modifying existing objects or by adding new objects and components and integrated with external systems. Beyond simplistic issues of language and application-server choice, five key levels of extensibility exist:

  1. Look and feel, ease of navigation

  2. Business logic

  3. Business components

  4. Enterprise Application Integration (EAI)

  5. Content exchange[1]

In the heat of implementation, these are the measures by which the practical openness of an approach must be judged. At the business level, key considerations in the purchase of any solution should be:

Whether it reliably solves a business problem today

Whether it is architected to readily integrate with existing systems

Whether the vendor will ensure its efficacy at solving that business problem tomorrow[1]

The Reality of Plug and Play

The promise of plug and play for the enterprise has been that, through an alliance with a specific application server, the enterprise will have the freedom to choose best-of-breed applications from expert vendors, with the assurance that each application will be easily installed and work cooperatively with other plug-and-play software. This ideal has not yet been realized. One of the major reasons for this underperformance is that most complex aspects of an enterprise e-business application don’t reside at the level of the application server, but rather one layer below. This lower layer includes:

  • Data repository

  • Data models, such as the definition of a user profile, purchase order, or product

  • Business logic, such as the workflow of a shopping cart for an order-entry application[1]

Whether they’re related to enterprise resource management or to higher-level business functionality, these e-business objects are critical and complex, often 10–100 times more complex than the underlying application server. And, the standard for the technology used to program that layer has never been a matter of agreement.

In reality, the application server or “plumbing” of an e-business system is the smallest system cost, yet it is often the subject of intense, unmerited focus. The real challenges of an e-business solution are:

  • Ensuring that all databases are consistent

  • Mapping the meanings of all data models across these objects

  • Encoding a precise translation from rules and operations in real business operations to underlying business objects[1]

Meeting these challenges is absolutely vital when starting from scratch or choosing parts of solutions from vendors who use different databases and differing business logic. An end-to-end e-business application ensures a consistent underlying data repository, a common data model, and streamlined business logic.

Scalability and Performance: Practical Definitions

The terms performance and scalability are often used interchangeably, when they are actually two different concepts. Both are of utmost significance for an enterprise e-business system, and the two are related. Here are their definitions:

Scalability: The ability of a system with multiple available processors to call as many of those processors into service as necessary when system load increases, as well as the ability of that system to be expanded

Performance: The ability to effectively increase throughput as needed on a single CPU in response to increased system load[1]

Very often, those making IT buying decisions do so with a heavy emphasis on scalability and performance, and rightly so. Many times in technology-evaluation situations, the word scalability is used when performance is the real issue. It’s important to note that performance benchmarks on individual applications or on the underlying plumbing of a system are irrelevant if the overall system doesn’t scale. Furthermore, linear scalability and geometric scalability are two different things:

Linear scalability: The ability to increase system resources by adding CPUs, with each CPU adding a linear increase in capacity

Geometric scalability: The ability of a system to increase system resources by calling upon a complex array of additional resources—typically less efficient than linear scalability

Linear scalability is essential because, without it, the cost of hardware required to ensure scalability becomes prohibitive. Linear scalability scales in four dimensions, across:

  1. Multiple threads within the same process

  2. Multiple processes/CPUs within the same machine

  3. Multiple machines (load balancing)

  4. Multiple tiers of machines[1]

Building scalability into end-to-end solutions is much more straightforward than when dealing with systems made with a patchwork of different applications or those built from scratch.

Easing Integration and Speeding Time to Market

Stated simply, integration can make or break an implementation. Backend and legacy systems in most enterprises not only represent considerable investments, but are responsible for mission-critical aspects of daily business.

The difficulty of troubleshooting an integration is multiplied when numerous solutions and vendors are involved, not all of whom provide 24-hour-a-day, 7-day-a-week support. All too often, the focus turns to solving technical problems and moves away from a tightly aligned answer to the original business problem. By approaching an e-business system as a time-critical, feature-rich solution that must be flexible enough to change in the future, a vendor offers customers the real prize: faster time to market and freedom to evolve.

Considering the Total Cost of Application

Anyone considering an e-business system is likely to be concerned with budgetary implications for obvious reasons. Unfortunately, it’s easy to factor in the up-front costs of tools, applications, and systems without considering the total cost of implementing an application. The total cost of an application consists of four basic elements:

  1. “Plumbing” or application-server costs: typically 5% of the total

  2. Scoped-out functionality: about 35% of the total cost

  3. Unplanned functionality: up to 50% of the total cost

  4. Integration: about 10% of the cost and a high percentage of the time[1]

All too often, as the unplanned functionality of a system increases, its performance has a correlating decrease. To address this shortcoming, some vendors suggest eliminating functionality. Although this may have been an acceptable alternative in the early days of e-business, it is no longer viable, given concerns with quality of service, customer satisfaction, and reduction of churn.

Note

The combination of increasing costs to capture and accumulate new Internet subscribers and intense competition in the marketplace has made customer retention and churn minimization critical factors in the survival of service providers. Occurring when new subscribers sign up for service while others are discontinuing their use, churn poses serious challenges to a provider’s ability to turn a profit. Effective churn management stems from a provider’s ability to determine the reasons for this customer behavior. This means that information is a provider’s best defense against churn. Root cause analysis of churn also removes ambiguity in the business planning stage, and allows service providers to create products, services, and e-business practices that make their e-business more efficient and profitable.

The problem is that customers don’t have robust, interoperable software suites for end-to-end e-commerce from a single vendor, and they have to cobble together solutions from different vendors, which is a costly and slow process. In other words, technology changes too rapidly and the market is too uncertain to plan and never fully execute. A Web site is a work in progress; you are always and changing and improving it. Buying a solution gives you a steady foundation from which to grow and easily adapt to changing customer demands and changing business models.

Once functionality has been purged from an e-business system, it is likely to be gone forever. Therefore, the significance of making performance and functionality prime components of an e-business system from the beginning is clear. And, that’s eminently possible with an end-to-end system designed to offer those fundamentals from the start.

When an e-business solution doesn’t address these principles organically, the development process and performance tuning are unending. Additional hardware resources, development dollars, and precious time must be added to the project, affecting the organization’s ability to reach the market in a timely fashion. When all is said and done, such delays mean business must be put on hold.

Finally, in the rush to establish e-business leadership, it’s critically important to focus on the real issues of end-to-end functionality, integration, and support. No longer are the choices limited to toolkits, mix-and-match systems, or “closed” packaged applications. Today’s enterprise needs an end-to-end solution fast, and it can’t accept the risks of uncertain performance, integration, or functionality in the rush to market.

[1]“Beyond ‘Build vs. Buy’: Winning at E-Business through Reliable End-to-End Integration,” 2000 BroadVision, Inc. All rights reserved. BroadVision, Inc. 585 Broadway, Redwood City, California 94063.




Electronic Commerce (Networking Serie 2003)
Electronic Commerce (Charles River Media Networking/Security)
ISBN: 1584500646
EAN: 2147483647
Year: 2004
Pages: 260
Authors: Pete Loshin

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net