First-mover ASPs have come unstuck by failing to understand what
customers are willing to pay for
As 2001 progresses, the much-hyped application service provider
(ASP) business model appears tarnished. Europe's prominent
pure-play ASPs continue to shed staff, even dropping the ASP tag
from operations. Has the ASP model been damned? The short answer
is, "No".
The ASP phenomenon is not unlike the growth of PC usage in the
1980s when hardware became a commodity and software grew in
importance. ASPs represent the growing significance of the network
as software migrates on to the network - in effect becoming a
service.
The concept of delivering a technological capability or set of
business functions as a service is not new. Outsourcing, a service
that provides some or all of an organisation's IT, is a large and
active market. ASPs will follow. Analyst company IDC has estimated
the value of the ASP market in western European will be $5.8bn
(£4bn) by 2005.
Today's struggling ASPs are first-movers that lack money. With
avenues of funding closed off they need to prove that they can make
money. Mistakes have been made. Outsourcing a software service to a
third-party provider such as an ASP is attractive to the small and
medium enterprise (SME) sector, but most sales efforts have been
directed towards the blue-chip space.
Corporates prefer ownership of their data and applications and
require strong proof of adequate security from any potential bidder
to run them. Also, ASPs that do target cost-sensitive SMEs offer
applications not written for use on a one-to-many basis via
Internet protocol. As a result, ASPs are forced to offer separate
racks of servers in their datacentres for each customer, limiting
economies of scale - most ASPs are too expensive.
The "cheap" sell has failed and applications cannot be supplied
cheaply enough to attract the critical number of users through a
one-to-many model. So ASPs need to emphasise the ease of
implementation that is starting to attract custom.
This situation could reverse - the possibility of recessionary
pressure will push cost-cutting to the fore. But waiting or hoping
for a full-blown recession points towards uncertain business
strategies. It is more relevant to talk about the ASP effect. ASP
definitions centre on one-to-many applications, which work well for
commodity applications but in the UK one-to-many application deals
are rare.
You would expect that as the industry progressed the term ASP would
become clearly understood but ASP has changed into a number of
propositions for software suppliers: for some, it is about
delivery, for others monthly pricing, location or ownership.
The ASP effect, therefore, applies to the infrastructure used to
host the application. The one-to-many condition is moving from
applications to the middleware, security and other infrastructure
requirements all ASP operations must have. Customised one-to-one
application environments are found on top.
Pioneers will always capture "mind share" but second movers learn
from their mistakes. The large software companies are actively
devising ASP strategies: Microsoft has endorsed the model with its
.net initiative, launched in June 2000 and targeted at the
developer community producing applications residing on the
network.
The transition towards ASP business models will be difficult, but
the larger players like Microsoft, SAP and PeopleSoft can fund it.
Most software companies are geared towards the distribution model
where one-time revenue recognition and quarterly results lead to
equity valuations. Changing to an ASP model, based on annuity,
represents a huge commitment.
Clarification is on its way as second movers resolve these issues
and convince users of the benefits of utility computing. This shift
toward a utility model for delivering access to technology and
information has been anticipated for decades. The IT, comms
infrastructure, supplier needs and buyer demand are lining up to
make it a more pervasive reality. ASP represents its first
tentative steps.
Euan Davis is senior analyst of European services at IDC