|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Older, Windows-based systems may allow the Internet access by using “thick or
thin clients” installed on users’ computers or 3rd party tools such as Citrix.
This approach inherits most features of the older technology. It either shares
just a database or affords too much overhead and still lacks conveniences of
the Internet-native solution.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Of a number of considerations that come into play when choosing a deployment model,
cost, functionality, integration, and customization, not surprisingly, lead the list.
If a smaller business needs to automate simple import export functions, a vendor-hosted
solution can be very attractive. Though they’ll have to sacrifice some of the things
that ownership can bring — functionality, control, ad-hoc reports and queries — they
need fewer IT resources and hardware to get things underway.
Many enterprises, of course, don’t want to cede the control and functionality that
comes with running their software in-house. They rightly feel that deploying fully
automated and integrated solutions allows them to better differentiate themselves
from their competitors. Experts say that if a company requires a robust application
and has integration and customization needs, they will be better served by an
in-house deployment. Complex security and regulatory requirements in some vertical
industries, such as global trade and supply chain logistics, are other factors
that can make the in-house hosting to be the only viable option.
The cost-of-ownership question is a little trickier. Not surprisingly, the nod in
terms of costs initially goes to on-demand solutions: customers don’t have to
invest in hardware or factor IT personnel into their cost models. However, while
the cost of an in-house, licensed model drops over time and eventually comes down
primarily to maintenance fees, customers continue to pay the same transaction fees
for vendor-hosted offerings. Esteban Kolsky, Research Director at Gartner Group,
says that, all things being equal, the cost differential between the two models
begins to equalize somewhere in the third year.
|
|
|
|
Factor
|
Outsource
|
In-House
|
|
Cost
|
Lower start-up cost. Pay per transaction. Inability to
capitalize deployment costs.
|
Requires up-front investment. Becomes cost effective during
the 3rd year, faster with the higher volume of transactions.
|
|
Deployment
|
Standard implementation requires setup and training.
|
Standard implementation requires installation and training.
|
|
Ongoing Support
|
Less maintenance required in-house. Reliance on the vendor
for mission-critical system support.
|
Requires ongoing maintenance in-house, along with vendor
support via a Software Maintenance Agreement.
|
|
Flexibility
|
Reporting is limited. Functionality is restricted. No
direct access to transaction data. Avoidance of transaction fees may
result in uneconomical business practices.
|
Transaction data resides in-house. Complete functionality.
Maximum flexibility for integration, customizations, and reporting.
Improved business practices.
|
|
Security
|
Sensitive data stored by the vendor, normally with a disclaimer.
|
Sensitive data stored on-site.
|
|
Overall
|
Easier to manage
|
More control and functionality
|
|
|
|
|
|
|
|
|
|
|
|
|
Go Back