There was a time when it was not unheard
of to see CIOs spending a million dollars buying licenses of some
ERP and then another million or more to implement it in their
enterprise. Moreover, anecdotal evidence suggests that 40% or more
of these multi-million dollar implementations did not produce the
magic that was expected out of them. In response to the resulting
uproar emerged the ASPs – providing access to expensive
applications in a pay-per-use mode, apparently aimed at easing the
pressure on the need to make large capital investments.
Unfortunately, ASPs failed! The reasons were many but primarily
because although the basic idea was sound, the eco-system required
for the model to thrive did not exist.
The pay-per-use model is reemerging as SaaS – Software as a
Service. Some of the old myths (and some new ones too) about it
still continue to cloud common perceptions. However, advances in
technology – especially SOA - as well as the readiness and
ability of enterprises to adapt to the constraints of SaaS indicate
that, unlike the first time, this second coming is here to
stay.
Key Characteristics of SaaS
Pay-as-you-go, per-user monthly charges: You do not need to buy the
software. You only use it and pay a fee for the privilege of using
it. Typically SaaS providers price the service on a per-user basis
with additional charges for storage, bandwidth, SLAs, performance
guarantees, etc. While the SaaS revenue from a single customer
(especially if it is a small business) may be small, its assured,
predictable recurring business to the vendor.
Web-based access to the software: Anywhere, anytime access is one
of the core value propositions of SaaS, wherein the end user can be
anywhere in the world and get the same service over the
Internet.
Multi-tenant architecture: Application delivery typically uses a
one-to-many model, wherein a single instance of the software serves
multiple clients.
Centralized ownership of software, upgrades, versions, patches,
etc: Unlike the usual situation where the end user needs to set
aside funds, processes, people and time to purchase, install, host
and manage the software, the SaaS vendor takes complete
responsibility for everything, including ongoing evolution of the
software, sharing the costs across multiple users.
Myths and Truths
Myth 1: SaaS is a synonym
for CRM and Sale Force Automation applications
Without question the most celebrated example of SaaS in operation
has been Salesforce.com. After the disastrous record of the early
ASPs, skepticism about the commercial success of providing software
in a pay-per-use model was rife. Salesforce.com certainly
demonstrated that there was hope in the basic idea. However,
adoption of SaaS cuts across a much broader spectrum of business
areas. In fact research done by both Forrester and Gartner,
indicates that CRM/SFA is distinctly behind in adoption of SaaS
compared to areas like Content, Communication and Collaboration or
Human Capital Management.
What is more important is that the same surveys showed significant
increase in usage of applications in the SaaS mode for Order
Management, Supply chain Management, Enterprise Resource Planning,
etc. Recent moves by the dominant vendors of software packages in
areas other than just CRM seem to corroborate the findings of this
survey. Both SAP and Oracle have recently launched their ERPs in a
SaaS model, e.g. Business By Design offering from SAP and Oracle On
Demand. Other moves in the market such as Ariba’s acquisition
of Procuri, which provides SCM capabilities in a SaaS model,
further validate these findings. Procuri is widely used for supply
chain management by large enterprises like ConAgra, Domino’s,
and Eastman Kodak, to name a few. The interesting implication of
this is, these vendors are committing large investments because
they believe that SaaS will enter into the hallowed, core business
processes of enterprises – areas that are typically closely
guarded, and considered mission-critical and central to the
enterprises’ success.
Myth 2: Barriers to wider
SaaS adoption are fears about data security, lack of required
functionality, and constraints imposed by
standardization
The usual dogma is that wider adoption of SaaS is hampered by the
fears potential users have about keeping their data confidential.
Since the SaaS vendors control the functionality, behavior and
look-and-feel of the applications, potential users fear that they
will have to alter their business processes to match whatever the
vendors choose to implement. And that they will have to make do
with shortcomings in the functional capabilities and will have to
re-train their users to live with unfamiliar user interfaces.
Interestingly, none of the above figured as top concerns of
enterprises seriously evaluating SaaS in a recent business
technographics survey by Forrester. The top two barriers to SaaS
adoption that came out from the survey were: (a) Integration of
software bought as a service with the remaining in-house
application portfolio and (b) Operating costs. Fear about data
security was only third in order of priority. Lack of customization
was fourth, and the others figured even lower in the list
(shortcomings in functional capabilities did not even figure in the
list of top 10 concerns!)
The implications of this are important. Techno-functional barriers
can be difficult to overcome, requiring long time and large
investments in R&D. Despite that, they may remain as obstacles.
Overcoming commercial barriers like cost considerations are often a
matter of negotiation and market pressures and therefore more
readily solvable. So concerns about integration are important to
understand. In today’s world, organizations cannot afford to
have silos of applications that do not interact well with each
other. There is no such thing anymore as a stand-alone CRM
application: the business process that starts with the
identification of a prospect needs to work its way through a
multitude of applications seamlessly – from lead generation
to contract closure to order management to production planning to
procurement and supply chain management to shipping and service
management – and in the background, interact with financial
systems, human resource management systems and so on. Selecting one
or more of these business applications to be bought in a SaaS mode
creates a complex integration and change management project that
must be efficiently executed if SaaS is to succeed. That is the
biggest challenge in wider adoption of SaaS.
Myth 3: SaaS is for Small and
Medium Businesses
The source of this myth is in many ways linked to the perception
that SaaS is cheaper, that smaller businesses do not need
sophisticated functionality, and that they can live with lack of
customizations.
The facts are to the contrary: SaaS is not necessarily cheap (as
borne out by the Forrester survey mentioned above), and SaaS does
not necessarily imply inferior functionality.
Moreover, an examination of SaaS usage by large enterprises reveals
that as many as 25% of the large enterprises in North America and
Europe are already using the SaaS model somewhere in their
application portfolio. Merrill Lynch reportedly has 25,000 of its
employees using Salesforce.com, while Procuri as a SaaS SCM
solution is being used by the likes of ConAgra Foods and Toyota
Motor.
The next frontier of SaaS
The motivations for an enterprise to consider SaaS are obvious:
reduce large capital expenditure, reduce waiting time for large
development projects (which more often overrun than not) or large
ERP implementation projects to complete access to new
functionality, and make operating expenses variable and
proportionate to business growth. The biggest challenge in
large-scale adoption of SaaS is integration of functionality bought
as a service with applications owned and managed in-house as
products. A hidden challenge – and something that has not
received enough attention – is managing the change that is
inevitable due to SaaS adoption. Change management is not just a
technology issue; it will involve business process reengineering,
re-training and re-orienting users, overcoming inertia and
resistance to change within the enterprise, and so on. To determine
whether SaaS is cost-effective, therefore, is not simply a matter
of understanding the pricing models that the vendors are proposing.
It will require an assessment of integration and change management
costs, with the latter seldom being a simple exercise.
On the positive side, however, the evolution of technologies,
especially Service-Oriented Architecture (SOA) and the ecosystem
that has grown around SOA, has provided answers to many of the
problems that dogged the old ASPs – anywhere, anytime access,
quality of service, ease of integration, and aggregation of diverse
applications, data security, etc.
This new wine in old bottle may well be the toast of corporate
buyers!