many articles today discuss the capabilities of integration services and their
ability to tie together information from multiple disparate systems. In the
marketplace there are stand-alone products, provided by companies like Tibco
and Vitria that perform these services, and it seems that all
of the major ERP vendors now have an integration story of their own. The focus
on integration stems from the fact that businesses have business processes that
they are attempting to manage. These processes extend beyond the natural boundaries
of the applications historically available on the market.
To further complicate matters, as organizations set out to automate additional processes within sales, marketing, and service, or extended their reach using the Internet, they often deployed point solutions. This solved immediate problems, but also created additional silos of information and redundancy in the core tables, or master files (e.g. customer, vendor, pricing). Now the challenge is not only to connect the flow (e.g. web order-to-cash), but also to keep the master files consistent, and minimize the redundant data entry requirements. The need for business intelligence creates another set of problems, so companies create data warehouses, further adding to the complexity of the solution.
The focus of this article will take a slightly different approach to solving the problem. Our answer is found in the "Power of One." This means one primary business application, using a single data model to store all relevant information.
Note: This article, written by a staff professional from a
well-known vendor, was selected for publication to open a debate on the proverbial
and controversial dilemma — the one-stop-shop (i.e, the power of one) versus
best-of-breed concept. Vendors, consultants, users, etc. are more than welcome
to contribute their thoughts, experiences, and opinions, preferably as similar
articles. The article itself does not represent an endorsement by TEC of the
idea espoused. To do so, contact firstname.lastname@example.org
The Driving Factor
If we explore the root need for integration services, we will find that it is really driven by three primary business needs: integrated flow, redundant master files, and cross application business intelligence. We also see that this collection of disparate applications was not created because companies wanted multiple systems, but because the evolution of their businesses occurred faster then the evolution of the application software market.
That being said, an assumption could be made that, if a company could run their entire business from a single application (and had the resources to implement it) that would be the preferred solution. Realizing of course that that is a utopian state and that there are years of development and process improvement that must be considered, let's simply agree that the constant charge towards a single enterprise application is a worthy pursuit. It is my humble opinion that all IT executives seek the "Power of One."
an all-in-one solution would have a negative effect on many software vendors,
as they would experience what is known as consolidation. In fact, this is something
that is already happening and began shortly after Y2K. Even in the most talked
about CRM space, we saw companies like Siebel acquire eleven
point solution companies to become a "CRM Suite." Although Tom Siebel won't
admit the value of an all-in-one solution, he clearly saw the value in an all-in-one
CRM solution. It is interesting that Siebel now offers the Universal Application
Network (UAN), which "meets the key objectives of enabling organizations to
deploy end-to-end, industry specific business processes." In other words, although
Siebel does not offer an end-to-end solution, they do realize this is what the
market needs and will fight the best-of-breed battle, while providing integration
Obviously the independent business integration software companies would argue that integrating multiple systems is the right way to go. Tibco's solutions claim to "empower our customers to dramatically improve their business performance by enabling interoperability between diverse computer systems, helping them streamline activities that span their extended enterprise, and giving them the power to more quickly identify and respond to opportunities and problems." So at the root of it all these companies recognize the issue and have a solution to the problem. Unlike the "Power of One," Tibco claims to have "The Power of Now."
The Mid Market
The other assumption within this article is contrary to the assumption of most analysts. They will tell you that the top 100 companies lead the way, and the mid-market follows. I recently attended an analyst briefing where the analyst made it clear that the Global 500 would never get to a single application, therefore it will never happen. In my twenty years of experience with medium sized companies, I have observed that these analysts are wrong, and that the needs of a medium sized company are very different from the needs of a large enterprise. More importantly the opportunity for change is drastically different.
large global conglomerate has many corporate entities with many disparate systems
within each of those organizations. For example, GE is made
up of a broad range of primary business units, each with their own number of
divisions. These business units include manufacturing divisions (e.g. GE Aircraft
Engines and GE Plastics), financial services divisions, and broadcasting (e.g.
NBC), just to name a few. At this level, an organization must
accept the fact they could never stop everything and implement a single solution
across all business divisions. They still need to get to the "Power of One,"
so they will spend millions of dollars integrating hundreds of systems, and
building multiple data warehouses. No one can argue with the success of a company
like GE, so this approach must not be totally ruled out.
However, if we look at a typical mid-market company that has roughly $350 million in annual revenue, we will find that they don't have hundreds of legacy systems and millions of dollars to spend each year on IT staff. Many companies of this size have somewhere between two and ten full time employees in their IT department. They have two or three primary business systems, a handful of standalone tools, many spreadsheets, and multiple Access databases. In short, they have not fully automated their business processes, and would gladly give up their legacy systems if they could make it all work from one server, and give the executives the information they need in the process.
companies like Epicor have built their entire businesses around
providing integrated enterprise software solutions for mid-market companies
around the world. Epicor claims to have over 15,000 customers and the ability
to deliver end-to-end, industry-specific solutions.
as the Y2K boom started to settle, several ERP companies began exploring the
possibility of building CRM packages of their own. Around this time, SAP
took the position that no company could possibly provide that large of a footprint
and opted to partner with Siebel. Oracle, on the hand, made
a decision to invest in such an endeavor and in the spring of 2000 launched
the first version of its E-Business Suite 11i.
Oracle E-Business Suite is a comprehensive set of web-based,
enterprise-wide business applications. Oracle was the first major vendor to
combine ERP and CRM, and to launch a 100 percent web-based product. Today they
remain committed to the all-in-one concept. The footprint continues to grow
with additions such as advanced supply chain planning and daily business intelligence.
The product also continues to mature with over 2,500 customers live on 11i.
Oracle pushed the envelope and was mocked by the analysts for it, yet now all
of the major ERP vendors have followed suit and are attempting to build all-in-one
has spent the years since its statement that no single vendor could do it all,
building or buying the pieces of an end-to-end solution. However, SAP maintains
six separate data models for each major functional module: R/3, APO, CRM, Business
Intelligence, Enterprise Buyer and Portal. The duplication of common elements
prevents SAP from being able to perform real time analysis ("daily business
intelligence"), and causes information to be obsolete from one SAP application
module to another.
SAP continues to rely on different technology stacks for each application, making
real time analysis impossible and causing information to be inaccurate since
data needs to be replicated. SAP talks about the benefit of a global single
instance but only a few customers have successfully deployed SAP globally. This
is because SAP's complexity makes it difficult to reconcile global needs in
a single system, since it must often be customized and does not really have
an Internet-ready client. Until true multi-tier web architecture is implemented,
mySAP.com functionality will be hampered in the areas of extensibility,
integration, and performance.
Susan Kitchens' Forbes article "Too Little, Too Late," (7/23/01) SAP's CEO and
cofounder Hasso Plattner and co-CEO Henning Kagermann admit their mistakes.
"We didn't develop our Internet strategy as promptly as we should have," says
Kagermann. Plattner concedes that SAP should have gone into development of the
front-office side of software sooner: "Okay, forgive us," he says. "We are a
little late." Based on this information it is clear that SAP now sees the value
of the "Power of One" message and continues to work toward that goal.
known primarily for its HRMS solutions, started developing software in the mid
1980s, PeopleSoft has since developed an all-in-one enterprise application.
The current version, PeopleSoft 8, claims to be 100 percent
Internet-based and contains "no code on the client." However, according to the
PeopleTools 8.4 Hardware and Software Requirements document, the client machine
for a PeopleTools power user must include almost 650 MB of data, (e.g., Microsoft
Office, financial budgeting, supply chain planning applications, Visio [used
for workflow], Cognos, ESSBASE, and NetExpress Cobol Version 3). Additionally,
Peoplesoft's architecture is inefficient and not integrated. PeopleSoft has
four different data models: one for HR, one for financials and manufacturing,
one for CRM, and one for reporting. They are also moving toward the "Power of
One" model; however, they have a way to go.
Edwards is another example of an ERP vendor that has chosen to develop
an all-in-one solution. The latest evolution of its products and services, J.D.
Edwards 5, was released on May 21, 2002 as a "New Family of Products
To Make Customers Stronger." The current enterprise solution, J.D. Edwards 5,
is the fifth generation of J.D. Edwards offering and is a combination of internally
developed, acquired and integrated, and pure third party products. For example
J.D. Edwards acquired YouCentric to begin to provide a CRM
solution and Numetrix to enhance its supply chain offering.
There also continues to be a dependence on many third party or "licensed products."
I am not sure what this looks like under the hood, but clearly J.D. Edwards
is attempting to provide the total solution. Note: J.D. Edwards is now a wholly
owned subsidiary of PeopleSoft.
are several others vendors including Epicor, IFS,
and Microsoft that have also spent the last couple of years
rounding out their versions of an all-in-one solution. No need to go into the
details here, but it should be clear that the market (especially the mid-market)
is receptive to an all-in-one solution and many vendors are attempting to deliver.
the market desires an end-to-end solution and even those opposed to the idea
of an all-in-one solution are marketing to the need of an end-to-end solution.
There are valid reasons why a company would choose integration of many systems
over a single integrated system and for these reasons there will always be room
in the market for integration software and point solutions. However for any
enterprise just beginning its business system journey and for those considering
a major ERP implementation, you must seriously consider the all-in-one vendors.
For those Global 500 companies that realize they will never get to one global
enterprise system, reducing the complexity is still an admirable goal and continuing
to look for opportunities to consolidate is a worthy pastime.
Schweers is a senior sales consultant at Oracle Corporation, www.oracle.com.
He has an extensive background in the manufacturing and distribution industry,
with a focus on medium sized companies. Prior to joining Oracle he has worked
for various other ERP providers in both pre-sales and post-sales capacities.
He has also held management positions for various manufacturing companies. In
his 20+ years of experience, he has worked directly with greater than fifteen
unique enterprise business solutions, and assisted over one hundred companies
transform their business through automation.
can be reached at email@example.com