There's been a lot of talk about the Software
as a Service (SaaS) model. That debate has helped define what SaaS is
now and what it should be in the future. But the chatter also asserts a
number of erroneous SaaS definitions out in the IT marketplace.
So let's start by clarifying exactly what SaaS is and then later what it isn't.
- SaaS is a software delivery model in which the user accesses
software over the Internet from anywhere at any time. Users need only a
web browser to use the software. Companies don't need to invest in and
upgrade servers, operating systems, databases, backup equipment and
complex programming environments. The software vendor provides all of
that.
- All customers run off of the same code base. This ensures
that the software the customer uses is always up-to-date and saves the
vendor from having to support different databases, operating systems
and versions of its application. Customers can deploy the application
very rapidly, since they don't have the lead time and hassles
associated with configuring their local environments.
- Customers do not have to worry about upgrading their
software. For example, no one worries about what 'version' of Google
they are running.
- The software provider is contractually obligated to provide
acceptable availability and response time. Again, the customer doesn't
have to monitor disk and CPU usage and upgrade hardware.
The most mature instances of SaaS are multi-tenant,
vendor-hosted solutions include Salesforce.com, NetSuite and my
company, Plexus Systems, Inc. These characteristics offer distinct
advantages. For customers, they are the ease with which the software is
kept current and the cost savings associated with a shared
infrastructure. Some of the cost savings come from using a simple Web
browser to perform any transaction in the system -- not just those that
have been 'Web-enabled.' For our customers, SaaS is an enabler of
nimble, responsive organizations that embrace both innovation and
customer value.
Given these advantages over the basic Application Service
Providers (ASP) delivery model of 2000, it's not surprising that SaaS
has become the "hot" application delivery model. It seems every
software vendor claims to have SaaS applications, whether the model
meets all the criteria or not.
I have to admit, I take umbrage with the fact that the SaaS moniker has been hijacked by some in the IT industry.
If you are a customer seeking a SaaS solution that adds value
to your operation, make sure you get a true SaaS application; not a
pale imitation, which may really be just an old client/server
application with a customized front-end and a few new features on a
dedicated server in a third party data center.
Pretenders will offer a variation of the following:
- A third-party hosting company running a separate instance of
the solution for each customer -- This doesn't solve any of the
delivery problems. It just moves them from the customer to a third
party. Each individual customer must still go through the painful
upgrade process. With no shared infrastructure, no costs are taken out
of the equation. In fact, the costs go up because the third party
hosting company also needs to make a profit.
- A 'new' Web-based user interface with the same old database
structure and very old business logic still intact. Some call this
"putting lipstick on a pig" or "whipped cream on roadkill." If the
underlying business logic is still in RPG, Cobol, or other older
languages, the vendor has not kept up-to-date and the software will be
expensive and cumbersome to maintain.
- Not 100% Web Native -- unless the solution transmits just
HTML (web native), the software will consume a lot of bandwidth. Get
estimates as to bandwidth needs for a proposed application. You should
be able to support your entire organization with a fractional T1.
The SaaS model evolved to drive out costs and complexity in
the relationship between the software vendor and the customer. It
addresses many problems that have arisen over the years.
To get a real SaaS solution that will yield all the benefits
versus a hosted legacy application that will just add cost and
complexity to the customer-vendor relationship, software buyers should
do the following.
- Be sure to see the actual software, not just Powerpoints.
- See the software in use at a similar company.
- Ask the software vendor to demonstrate the entire system
from one of your computers without loading any software on it beforehand
- Ask what the bandwidth requirements are.
- See where it is hosted.
- Truly understand how many users will be needed to get the
full value from the system. SaaS is a subscription-based service, with
fees usually (but not always) based on the number of users at a
particular company.
There is no doubt SaaS delivers value to the customer in the
form of rapid time-to-market and low maintenance. Customers want new
features faster than ever before, and SaaS effectively enables
responsiveness.
And if responsiveness doesn't sell you on SaaS, ask any CFO
or VP of Finance if they are happy with spending a high portion of
their budget keeping outdated software running, or with hiring IT
people to babysit servers and applications, and they usually start
fuming. They want an IT team that are business process consultants, and
who focus on building value with technology, rather than simply
babysitting technology. Or ask an IT guy how many of his applications
are updated to the current version, and he'll likely laugh at you.
It's my belief that the SaaS model will be the future of most applications, as the advantages are just too great to match.
Mark Symonds is president and CEO of Plexus Systems, Inc.
Symonds has his MBA in finance and accounting from Cornell
Universityand a bachelor's degree from the University of Rochester. He
is a Certified Public Accountant and is also certified in production
and inventory management (CPIM) by the American Production and
Inventory Control Society. He holds a variety of industry association
memberships, including the Precision Metalforming Association (PMA),
where he is a member of its national board of directors.