|
Now that most project teams are justifiably celebrating the successful completion of their Y2K enterprise resource planning (ERP) projects, organizations can turn their attention to leveraging these investments to increase business competitiveness in the post-Y2K era. These business improvements focus on tapping the awesome power of the Internet to extend the enterprise from the back office (where ERP systems are primarily deployed) to the front office. Connecting customers with Internet-enabled customer relationship management (CRM) systems and suppliers with supply chain management (SCM) systems is the holy grail for the new “e-era.” Perhaps the most significant challenge in reaching this goal is the fact that many information environments include a variety of best-of-breed components, legacy systems, and new SCM and CRM systems that are not currently integrated with recently completed ERP systems. Such integration enables the creation of customized enterprise solutions — solutions built around the customer — that can drive improved business performance objectives.
In response to this challenge, a new category of software solutions and services known as enterprise application integration (EAI) has emerged. In this article, I’ll overview these EAI solutions and services and how EAI — when “done right” — is a potentially solid approach that offers real financial value and flexibility.
Can You Really Have it Your Way?
A few years ago, a well-known company in the fast-food industry used an advertising strategy to capitalize on a common human desire to order something just the way you want it. The consumer hype promoted the ability to “have it your way,” as well as quickly and at a good price. The same marketing ploy pervades the software industry today: Many vendors tout their abilities to automatically link disparate applications into one integrated, end-to-end business system. As IT organizations are deploying packaged applications, business intelligence tools, and e-commerce software at a frenetic pace, the ability to tie them all together is critical. The intense need for a comprehensive solution leaves most companies asking the same question: “Can we design systems ‘our way’ so they can communicate, which will let me address mission-critical business issues such as enterprisewide CRM?”
EAI is not as new as the buzzword implies; organizations have been looking for ways to integrate applications and data for years. For 1999 alone, International Data Corp. forecasted that IT organizations would spend an estimated $53.3 billion for worldwide systems integration services. Fortune 1,000 companies report that this figure will increase to nearly $100 billion annually by next year.
EAI products are designed to link disparate applications by making systems inside and outside an enterprise share information and logic in ways that were previously nearly impossible, or at least too expensive. No longer do companies require legions of IT professionals to custom-build communication links among systems. Working at the data, message, or process level, new EAI tools offer companies the promise of integrating independent information systems into a single, flexible framework. Increasing competition, supplier demands for transparent business processes, and customers who expect instant access have fueled the emerging EAI industry.
The huge growth of the EAI market does raise several questions, however. First, what is driving the market for these integrated solutions? Second, how does EAI actually work? Finally, what is EAI’s impact on the ERP solutions that were supposed to remedy all our needs for integrated data?
What’s Driving the EAI Market?
The market for EAI is being driven by five significant business trends:
Growth in specialized packaged applications. The largest factor driving the market is the proliferation of packaged applications. This trend began as companies bought packaged applications to address anticipated Y2K problems; eventually, IT organizations realized the benefits of integrated ERP data and the comprehensive functionality in ERP solutions from companies such as SAP, PeopleSoft, Baan, Oracle, and J.D. Edwards. Unfortunately, even the ERP buyers found varying functionality in financials, distribution, manufacturing and human resources, which dictated a need for solutions that no single ERP package could provide. Consequently — and contrary to popular belief — many organizations decided to sustain homegrown systems because they wanted to preserve specific data or functionality. But in a relatively short time, most companies discovered that the need to integrate applications intensified as they added functionality and addressed specific application needs with specialized packages.
Streamlined business processes. The Internet fosters a culture of connectivity and speed that translates into a need to develop systems faster than the competition and create business processes that cross functions and organizations. In order to engage in e-commerce with business partners, rapidly synthesize shared business processes, and continuously innovate the business for greater efficiency, organizations are making EAI a business imperative. EAI is becoming an excellent tool for aligning processes to the business using new technology. For example, the development of a common language of data exchange such as extensible markup language (XML) expands the possibilities for making all applications communicate.
Mergers and acquisitions. Mergers and acquisitions have typified the business environment in the last decade, and little slowdown in this activity is anticipated. The ability to incorporate and leverage resources gained through mergers and acquisitions may promise competitive advantage but commonly requires massive integration efforts. Companies seeking to integrate their applications face several challenges compounded by the current shortage of skilled IT labor and the maze of available technology solutions. By providing the communication infrastructure and packaged “solution templates,” EAI reduces the time as well as the cost of integration. Don’t mistake these integration tools for out-of-the-box solutions — they cannot provide an instant cure — but they do suggest a better way.
Supply-chain management. Supply-chain integration is a mission-critical goal as companies strive for better solutions to help distributors and suppliers work together to address real-time supply issues. Companies are realizing they have to extend their processes out to partners and customers by providing access to business data and process flows. Furthermore, the possibility of information exchange through the Web has pushed the need for integration even further. EAI offers platform independence that gives customers, suppliers, and trade partners direct self-service access to internal business processes.
Technology wealth. EAI flourishes in a business environment where technology is a driver for competitive advantage. Technology advancements now enable integration through message queuing, data transformation, business-process modeling, and connections to multiple applications. There is little doubt that tying a constantly changing array of applications together will continue to be a challenge, but these integration tools should make that goal less imposing. EAI combines the best of technology with a pragmatic business focus on how to help companies maximize the speed and flexibility of business processes.
How Does EAI Work?
As with almost any new technology, everyone is promoting their own unique approach to EAI. The basic concept is to use technology to integrate existing and new systems using minimal IT resources. Here are some of the most common EAI approaches:
•Data-level integration is the most mature integration method. The data-level approach works best when the objective is simple information sharing. Systems that need only share customer information or sales data are good candidates for data-level integration. Vendors such as Informix/Ardent, Oracle/ Carleton, Smart DB Corp., Data Junction Corp., and Constellar Corp. represent some of the data-level providers; their products extract and transform data and move information among databases. Generally speaking, data-level EAI is inexpensive, relatively low risk, and provides a useful solution when integration needs are light. When you need to share business processes or more complex information records, other solutions are required.
•Method-level integration includes concepts of brokering or message-oriented exchange. The method-level approach lets you share data as well as the business rules, restraints, and real-time transaction information. The message-oriented vendors supporting interapplication data exchange include Active Software Inc., Frontec Inc., New Era of Networks (NEON) Inc., Oberon Inc., and CrossWorlds Inc. This method is based on creating a host for shared methods (such as an application server) and enabling IT departments to connect and coordinate events among multiple applications via this shared “host.”
•Business-process integration solutions add capabilities such as business-process modeling and workflow simulation in order to let users define potential solutions, simulate solutions to analyze the flow of information, and optimize the business process prior to solution implementation. Using process modeling, companies can focus on process optimization and easily change the process (or implement a new one) with minimal programming or coding. This method, which is also called the process-automation approach, enables movement among application outputs through “intelligent” routing. EAI vendors that focus on business or process integration include Vitria Inc., Hewlett-Packard (with ChangEngine), Tibco/InConcert, and IBM (with FlowMart).
How Does EAI Fit?
With all these potential solutions and approaches, how does any company begin to choose one? Similar to choosing any other IT solution, selecting an EAI tool is a matter of assessing your present situation, analyzing business drivers for change, and looking at the architecture and people resources necessary to make a change. The decisions about integration approach and technical solution depend on the volume of information you need to move, the type of data you want to move (data, events, or processes), and the overall number of systems involved.
Traditionally, ERP packages have been a panacea for integrating and automating business processes; establishing one database and working around a common workflow model has been a tactical advantage of large, integrated ERP solutions. Unfortunately, demands for specialized data, variances in systems among operating partners, and the sheer diversity in available applications have made it virtually impossible for any company to manage all their business processes with one software package. Legacy systems and new Internet applications based on Java and XML further complicate the integration process.
Furthermore, as businesses automate operations with everything from ERP software to front-office suites, the need to share information and processes with partners, customers, and suppliers is increasing. ERP systems have been primarily deployed to enable companies to automate back-office business processes, but now, in the “extended enterprise,” these business processes transcend departmental and even company boundaries and push for integration that extends beyond the traditional company IT-ERP infrastructure. Although the dominant ERP vendors are rapidly moving their products into the extended enterprise, the sheer magnitude and variety of installed systems have necessitated the development of integration strategies for companies to compete effectively in the Internet era.
In a perfect world, emerging data interchange standards such as XML would create universal standards for database connectivity that would remove the need for translation tools and middleware vendors. Unfortunately, the variance in developed solutions and the software industry’s history of failing to set de facto standards will continue to foster a variety of approaches until customer experience and ongoing product demand differentiate the winners from the losers.
The complexity of these “easy” solutions can be illustrated well by ERP integration: The major ERP vendors all publish APIs designed to let external systems communicate with the ERP system, and although such APIs are useful for information exchange, they are insufficient for distributed processing across different applications. Oracle has been most aggressive in developing APIs for various portions of its ERP applications, but other parts of those apps are inaccessible. PeopleSoft, Baan, SAP, and the others share similar progress with the rudimentary interfaces and integration tools that have been developed to date. Writing multiple low-level APIs is a difficult and time-consuming task that multiplies exponentially as the number of systems grows. As a result, many organization have resorted to improvising interfaces, building customized data-level integration tools, and even scraping screens to get important data for e-commerce applications.
One of EAI’s interesting consequences in the ERP market is the opportunities it enables for best-of-breed applications, which were hurt by the large market share consumed by the major ERP vendors in the 1990s. Some of the financial, manufacturing, and HR functionality that many organizations may have sacrificed in the interest of promoting an integrated ERP solution can be assessed again without an overwhelming concern about integration capabilities. The recent decline of ERP sales at the end of the ’90s — and the emergence of several new EAI vendors intent on solving the integration problem regardless of the ERP vendors’ assistance — may have a profound impact on this market’s evolution.
One Step of a Long Journey
EAI is one of the most complex and yet significant issues that IT organizations will face in the next decade. As organizations commence any enterprise integration activity, they will need to fully understand the architecture, modules, and interfaces available within their current enterprise systems. For initial EAI projects — which will focus on sharing basic information among multiple systems — the best method to initiate integration is through standard data-level integration, but increasingly these basic steps will be followed by more complex ones.
The real value of EAI is to help companies improve the speed and flexibility of how they conduct business. As IT organizations evaluate their business processes and customer needs for information, the demands for integration will be clear. EAI offers formal, proven models built around new enabling technologies designed to optimize enterprise performance. The results can have a dramatic impact on transaction speed, costs of operations, and the flexibility to change and adapt over time.
EAI Checklist
Navigating the product-selection labyrinth
To evaluate your investment in EAI, consider the following factors:
•How much information do you need to move from one application to another?
If the volume is low and is moving between two databases, it is generally easier to use some form of data-level integration tool.
•How many applications are you integrating?
Again, if the number of applications involved is low (one or two in number), the need for sophisticated EAI tools is minimal. Data-level integration or method-level integration tools work well with lower application volumes.
•Does the EAI vendor provide publish-and-subscribe and other asynchronous messaging support?
This issue involves scalability and the compatibility of your data with other information sources. EAI vendors that do not provide publish-and-subscribe or asynchronous messaging support generally have some compatibility issues and lack full flexibility to export data to other applications.
•Does the vendor support at least one major RDBMS?
The answer to this question will reflect the product’s robustness as a fully functional EAI solution.
•Is there integrated security (including secure sockets, encryption, and authentication)?
Integrated security is an important issue to consider, unless you’re willing to address it with internal IT resources.
•Is workflow or process modeling included? Does your EAI solution need to address workflow at all?
The increased complexity of workflow-capable EAI tools increases the price and technical capabilities necessary to gain maximum value from them. EAI tools with process-modeling functionality can more flexibly adapt to changing business processes while providing the integrity of accurate data exchange.
•Does the product integrate well with third-party simulation tools? What types of data will you share among systems (customer information, methods, entire databases, and so on)?
The complexity of the information involved should largely dictate your choice of EAI tool. For example, financial data comprising customer numbers and currency is a much simpler data transfer task than customer information databases, which typically contain a mix of text and numerical data.
•How much is e-commerce driving the need for EAI?
E-commerce-driven EAI should focus on Web-enabled data transmission through XML or Java.
Dan Vander Hey (dan_vander_hey@hunter-group.com) is a practice lead for management consulting with The Hunter Group, a Renaissance Worldwide company. He has more than 16 years of experience in providing business solutions and project management assistance.
|
|
|
| |||||||||||||||||||||||||||||||
