Not even acronyms stand still in the
IT sector. ASP used to stand for Active Server Pages, but now it stands for Application
Service Provider too. Network service providers and software developers are proposing this
new technology model as a means of cutting costs and making technology easier to implement
for small and large businesses alike. As with many marketing opportunities, companies use
the term ASP very widely. Its used to capture customer mindshare, but in reality,
there is a clear dividing line between the real ASPs and the pretenders. Mark Roberts, ERP
alliances director for Cap Gemini, explains that in the ISV market for vertical market
applications, many people are calling themselves ASPs when theyre actually little
more than software hosting companies. They sell customers the license up front and offer
to host it for them, or they take customer software and run it on their own servers.
ASPs, on the other hand, develop their tools, build the infrastructure and then actively
go to the market and sell it to people. Their pricing models are a key defining part of
their existence as an ASP. They generally charge on a per-user, per-month basis, rather
than leasing or selling applications outright.
The three layers of ASPs
There are three layers of player in the ASP market, says Roberts. The first consists of
those service providers offering access to productivity tools such as email and word
processing, while the second is populated by ISVs that concentrate on line-of-business
applications like human resources and payroll. Finally, the third layer of ASPs focuses on
the value-added area, enhancing back-end infrastructure and providing links with their
customers other business partners.
This third layer is still relatively immature, but offers some exciting growth
opportunities, says Roberts. It will develop into an aggregated space incorporating
functions such as customer relationship management, business to business e-commerce, and
supply chain management, he says. Existing players in this area are those firms that you
wouldnt normally define as ASPs, such as Commercenet and Ariba, who provide large
trading networks for business customers.
Potential streams of revenue
Market research figures on the potential size of the ASP market vary, but all of the
vendors are flying in to feast on the potential revenues. Much will depend on the size of
the company that ASPs target, says Peter Bell, Internet business development manager at
Microsoft. Enterprise customers move a lot slower and their sales cycles will make their
take-up of the technology sluggish, he argues. The medium-sized companies will be the
target market for ASPs, because of their mix of a lack of professional staff and a need
"They want ASPs quickly because they want to continue their progress in the
market," he says, pointing out that many of these companies are on a fast track and
need to stay ahead of the competition. "As they become larger and more successful,
they want to continue to operate in a successful way." This will create markets for
applications such as customer relationship management software to improve customer
service, and the quick implementation of ASP software is vital here.
David Angwin, U.K. marketing manager for managed Internet applications at IT consultancy
ESOFT Global, agrees that a big driver for companies to move to the ASP model is the lack
of in-house skills and an unwillingness to invest in lots of technical resources,
especially in the SME market. These companies will find a price-per-user model much more
attractive than a large up-front capital expenditure, he says. When youre a small
business, cash flow is important, and anything that you can do to regulate costs will be
attractive.With their lack of technical expertise, SMEs may suffer their fair share of
system crashes and data losses, which is another reason for investing in an ASP model,
says Angwin. He argues that ASP systems have the potential to be more robust and secure
than many SMEs systems. "What happens when you get in one morning and the
system has fallen over?" he asks. "Often, you put up with this stuff because
thats just what IT is, but ASPs will offer more reliability."
The reliability of an ASP will depend on the models that it uses to bring the software to
the customer. Angwin says that ESOFT can ASP-enable most applications by going through a
technical evaluation process and making sure that it scales. The firm uses Citrix as the
enabling product here.
Unlike a pure play ASP, ECSoft wont host applications for thousands of users. It
simply acts as the facilitator to make an application ASP-ready, including making it
secure. It will also manage service level agreements for an application that is being
hosted by a large service provider, he says. This means that companies can concentrate on
providing the infrastructure services as a core competency. Making applications robust
represents only half of the technical support opportunity for ASPs, however. The service
provider can provide proactive value-added technical support by shadowing the customer,
argues Angwin. Customers can ask questions online, and software technicians can
theoretically take control of the mouse and show them how to resolve specific problems.
Cap Geminis Roberts sees another market dynamic that will drive the uptake of ASP
services. Business application vendors often come up with radically different upgrades,
and a lot of SMEs look at the competition when evaluating such upgrades because major
product shifts make it feasible for them to switch vendors. If an ASP service is on offer,
they may also take the opportunity to change the way that their applications are
Negative market factors
Nevertheless, not all of the market factors are positive. Some may inhibit the growth of
the market, according to Lance Travis, service director for enterprise architecture at
market research company AMR. One thing holding the market back is that the ASPs want to
host cookie-cutter applications, he warns, where the level of customisation is
minimal. "When you get into applications like ERP and back office software, most
customers want some amount of integration with legacy systems, and most companies
arent set up to do that," he says.
The ASPs that will succeed in the short term are therefore those providing applications
that dont require much modification, he argues. Those that can cope with the
extensive integration work will survive in the long term the switch in emphasis
between vanilla software and more customisable applications will happen in the next twelve
months or so, he says.
Customisation comes at a price, of course. Neal Quin, strategic business development
manager at IBMs recently-acquired NUMA-Q division, argues that many ISVs will
replicate the functionality of enterprise resource planning products under an ASP regime,
for example, but the problem is that ERP could make implementation expensive for smaller
businesses to implement. Smaller ISVs focusing on particular industries will do well here,
he says, because they will be able to change their product to suit different companies in
the same market.
Customisation will be more important to some than others, says Quin. Larger customers will
be more attracted to it, because they are more willing to make the investment to tailor
products to mirror their sophisticated business processes. Smaller customers will not be
as preoccupied with making applications fit their exact requirements, he argues. Angwin
agrees. No matter what kinds of application you have, there will need to be some level of
personalisation, but the ASP model will force ISVs to make their applications more
intuitive, requiring less bespoke work for each individual customer. Many SMEs may not
want to pay for 20 days of consultancy time to tailor an application to suit their exact
needs, he argues. There will be fewer IT skills in the customer site, and so the value
proposition for ASPs is to make the tools easy to use while maximising the business
advantage. Perhaps a bigger inhibitor for the much smaller businesses is the lack of
available bandwidth, argues Bell, which is a surprising indictment, coming from a vendor
that is trying to cash in on the ASP market. He argues that the lack of installation of
DSL technologies and most peoples reliance on modem or ISDN technology will restrict
market growth. ASP environments need fat pipes, he argues.
Angwin isnt so sure. "I can run a perfectly acceptable (although at times a bit
clunky) Word 2000 session over a 33Kbit/sec dial-up line," he says, arguing that any
bandwidth considerations will disappear over time anyway as the UK catches-on to high
bandwidth networking technologies such as DSL. As bandwidth becomes a commodity, then this
will also drive the ASP market as telcos have to add value to their services. ASP
operations will be a good way of differentiating themselves, he says. "The bandwidth
growth will create a vacuum that will pull services through it." Nevertheless,
performance is a tricky service level to guarantee, he says, especially when youre
accessing ASP services over the Internet. He explains that his firm is working with
Packeteer to meet this challenge. Packeteer sells technology that sits at the edge of the
network to prioritise data streams over relatively small, unreliable bandwidths. The
companys PacketShaper product is therefore useful for providing quality of service
for ASP implementations.
To make their SLA monitoring more effective, ASPs will need to introduce monitoring agents
that exist at the customer site, or at random sites across the Internet, says Travis.
Certainly, systems management and reporting software vendors such as Jyra and Micromuse
are starting to offer these services, putting agents out on the Internet that can sample
performance at key data exchange points. ASPs in the US are not providing performance
guarantees, says Angwin, but argues that it will be an important area as the market
develops. ESOFT offers availability and performance guarantees at the present time, he
says, along with security. The managed service guarantee also handles issues such as the
frequency of system backups.
Performance will probably separate into different service levels. People will offer high
service-level guarantees at a higher price, says Bell. Some people will also deploy
applications over private VPNs rather than over the Internet in the short term, to
maximise response times. None of this is likely to thrill the finance director unless it
is possible to get a reliable return on investment figure for an ASP project. You can
obtain an ROI for an ASP model using the same sorts of calculations that youd use
for the evaluation of an in-house technology system, says Travis, but deducting much of
your capital expenditure and recruitment costs.
Angwin, on the other hand, argues that return on investment models are not relevant at
present. Instead, the benefit of these services will be in taking advantage of
applications that you couldnt have used before. "The main thing is that it also
provides them with decent cost controls," says Quin. "They wont be trapped
into very long lease times in financing terms, but instead will be able to understand what
their cost per end-user will be."
Small to medium-sized businesses are therefore set to benefit from the ASP model,
especially if they restrict their initial activity in this area to applications that
dont require much customisation, such as email handling, for example. As the market
matures, businesses will become more adventurous, taking advantage of more functionality
from ASPs and customising more facilities to fit their business processes. They will hit
the real sweet spot when they begin taking advantage of ASP applications that would not
have been possible to implement on an in-house basis.
and the ASP market
Determined not to make the same mistake that it did in underestimating the importance of
the Internet in the early nineties, Microsoft is rushing to meet the ASP market head-on.
The company is piloting its Office 2000 productivity suite as an online service at
present, working in conjunction with network service providers such as BT and Equant.
Peter Bell, Microsofts Internet business development manager, says that the company
needs to learn about the changes in licensing structures that it will need to make under
an ASP model. Long-term, the critical success model for ASPs will be three-tier
applications, with a Web-based client, application layer and then a data layer at the
back. In the short term, very few applications are built this way, says Bell. "DNA
2000 is all about three-tier and Web-based architectures," he says. "In the
short term, as you will see with the BT deployment, were using Windows
The move to a three-tier system is important, says Bell, who argues that using Web-based
clients rather than terminal services enables you to access applications using non-browser
devices such as PDA and smart phones.
Charles Fitzgerald, director of business development at Microsofts developer group,
explains that existing applications can best be ASP-enabled by using terminal services,
whereas new applications would best be built in a three-tier model using DNA. He argues
that a mix of local functionality and Internet browser functionality is an ideal model for
modern applications. "As people are building business applications that are tailored
for this environment, you have to have an offline solution," he says. "We have a
client architecture for Windows DNA that lets you do straight HTML, or full-blown Win 32
clients, or you can take this hybrid approach," he says.
The Internet doesnt, at present, have much in the way of quality of service or
built-in security, and Microsoft is working with Cisco and other companies to provide
these features. It is trying to curry favour for these technologies with its ISVs so that
they can produce more attractive online applications, says Bell. The companies are backing
IPv6 as the means of bolstering Internet services.