Software On-Demand, Pricing by the Byte?

By now, the
technology world is well-acquainted with the movement among
major systems vendors such as IBM , HP and
Sun Microsystems to offer on-demand computing services.

The trend makes sense, technology veterans say. On-demand is the next
progression from the client-server era of computing to a more distributed
model of offering utility-like data center computing services to corporate
customers.

The positioning for on-demand computing services, utility-style, also
reflects the IT industry’s profit margin shift to higher-value, value-added
technology such as software and services.

But are enterprise application vendors prepared to price their products
by bits and bytes, or recognize revenue in dribs and drabs as customers
order up spurts of software-enabled seats across an enterprise?

For enterprise software providers such as Computer Associates and JD Edwards , as well as application service
providers such as Salesforce.com, the answer is yes.

Gone are the days when software providers would “back up a truck to the
company’s front door” and “dump as much software as possible,” said Louis
Blatt, chief technology strategist for enterprise software company Computer
Associates, and expect a huge up-front license fee from a customer.

Blatt said many of CA’s products are being “granularized,” or comprised of separate components, which also
reflects the company’s adoption of Web services
protocols for open standards that are also part of a technology shift to
provide on-demand computing.

CA recently unveiled its own foray into on-demand with its Unicenter line
of products, which it calls “Managing On-demand Computing.” (CA’s enterprise
management products, such as BrightStor for storage, help IT professionals
monitor their network’s varying applications that operate across a sprawling set of computing platforms.)

The latest Unicenter products consist of modules that are designed to
help customers “dial-in” as much network management resources as they need,
instead of deploying extensive overhauls of IT structures.

The idea is to offer integrated applications that help clients manage
heterogeneous infrastructure, provide self-management/self-healing
capabilities within the applications, and create service-oriented architecture
that radically simplifies operations for IT staff.

Perhaps the biggest change in CA’s on-demand offering is its flexible licensing model, which CA calls “FlexSelect Licensing.”

Blatt said the program is fundamental to delivering the company’s
products as a service. And it means that, as software companies increasingly
offer smaller software packages, revenue recognition models are changing as
well.

Offering software applications “by the drink” now means that, instead of
one big purchase up front of software licenses, revenue is being recognized
on a weighted basis, in scale to how much a customer is using each month.

Les Wyatt, chief marketing officer of enterprise supply chain software
company JD Edwards, said the company began offering up its various supply
chain management products in smaller packages, or modules, during the
mid-1990s.

“The difference is how we’ve packaged our software so we give the
customer choices,” whether that entails application service
provider models, or pricing software usage by the seat or users, said Wyatt.

Historically, for software companies, licensing their product code to
customers has been the lion’s share of their revenues, followed by maintenance or user fees for the life of the license.

The value of the product was in licensing huge chunks of code. Additional users were a smaller portion of total revenue. But amid the trend toward delivering computing services on-demand, that model has now flipped, with the user fees gaining in importance.

“Over the last few years, we’ve moved very strongly to a user-based
pricing model,” said Wyatt. In JD Edwards’ case, that might translate to a
customer that decides to implement the company’s invoice-processing
software, or sales force automation tools for its sales staff, and perhaps
order up some chunks of distribution management applications.

“That’s one way we’ve packaged our pricing so it scales. Once you buy a
user (license), that user can pretty much use any application” the company
is running.


To read what an IBM executive and Forrester analyst have to say about on-demand, please turn to page 2…

Mark Hanney, IBM’s vice president of independent software alliances, said partner companies such as JD Edwards are already moving
forward in building on-demand functionality into their enterprise
applications.

Their applications are increasingly infused with autonomic (or
self-diagnostic) functions, as well as virtualization capabilities. “There
are many ISVs offering this kind of capability,” Hanney told
internetnews.com.

They would be foolish not to, given the interest in intelligent agents
for enterprises, especially on the part of IBM. As part of its own
on-demand strategy that entails hardware, services and
software, IBM recently purchased Toronto’s Think Dynamics, which
makes software that dynamically addresses network peaks and system failures by
allocating resources where a network’s system is bottlenecked.

Hanney said pricing by the seat is one example of how on-demand computing
models bring in money. “A lot of the real benefits of on-demand are about
lowering total cost of ownership,” he said, which drives customers’ demand
for integration capabilities among different applications.

For that reason among many, enterprise software providers are realizing
that they to have to offer smaller software packages in bits and pieces in
order to run on top of an architecture built for such kinds of provisioning,
Blatt and other experts said.

Frank Gillett, infrastructure analyst for Forrester Research, said the
pricing issue really hasn’t been settled yet for software makers moving to
on-demand models, which Forrester calls Organic IT.

“Organic IT is an architecture, not a pricing scheme,” according to a
recent research report by Forrester that focused on how to make sense of
buying computing services on-demand.

The report cautioned CIOs and IT professionals mulling those decisions to
think of the on-demand movement as an “evolutionary, not revolutionary
investment,” which will unfold over a decade or more.

“Many buyers presume that organic technologies must be purchased on a
pay-as-you-go basis, rather than on a more traditional CapEx basis,” the
report said, referring to the shorthand term for capital expenditures. “Why?
Vague advertising from IBM – among others – has confused the issue.”

Forrester analyst Charles Rutstein, who wrote the report, also noted that
“organic infrastructures don’t require that firms buy their technology on a
by-the-drink basis. But they do open up some innovative new possibilities
for pricing. For example, IBM recently signed a deal with Canada Life
Financial in which the insurer will pay IBM for software application
services on a per-policy basis, firmly tying its IT costs to its level of
business activity.”

But generally speaking, Forrester’s Rutstin and Gillett note, companies
are still uncertain about how to purchase technology, including software,
“by the drink” when, traditionally, IT budgets are set a year in advance.

Forrester’s recommendation for those making the decision to buy computing
on-demand is to ask three questions: what’s the right architecture? whose
datacenter? and what’s the pricing scheme?

CA’s Blatt said the shifting pricing model in software on demand is just
as good for the software companies as they are increasingly becoming for
customers.

Enterprise “management on demand is a journey,” Blatt said. “Our
solutions need to match up to that journey.”

News Around the Web