ZDNet discovers the changing nature of the business model, the role of the team players, the applications, the customer view-point, the lackadaisical attitude of ISVs and the future of ASPs.

An Indian IT services company, which has done extensive research into the global ASP roadmap is Corpserve. Headed by finance veteran Rajeev Mehrotra, MD and CEO and Prem Bajaj, COO, Corpserve was formed in April 2000. The business vision--to deliver ASP services first to the Indian markets and then scale globally. Both Rajeev and Prem have had extensive experience in the outsourcing business for the financial community segments prior to forming Corpserve, and both are well versed with the pitfalls of outsourcing and demanding levels of customer management.

For them ASP is a technology extension of their core competence. While Corpserve was created to become an ASP player, Rajeev has not rushed hastily into this business. The company has spent considerable time studying the strategies of the global players and the ups and downs of the ASP market. They have also used consulting firms to sound out their gameplan. Rajeev is therefore is well endowed to speak on the dynamics of the ASP business.

In an extensive discussion with Arun Shankar, Executive Editor, Jasubhai Digital Media, they cover the changing nature of the business model, the role of the team players, the applications, the customer view-point, the lackadaisical attitude of ISVs and the future of this very dynamic IT service.

Some years back, a very popular buzzword was outsourcing and was coupled with downsizing of the IT department in enterprises. How far away from this approach is the ASP model?

The business remains outsourcing. The mindset is whether outsourced or not. ASP is just the technology vehicle. There is an overlap between the two models of business process outsourcing and ASP. We are basically talking of a BPO model in an ASP environment.

The ASP business model has remained a hot subject even through the dotcom meltdown over the last twelve to eighteen months. But we still don't see any big players even on the global scenes. What is the reason for this?

Twelve months ago when we started thinking and working on our ASP business there was not a single article and not a single company talking about it--that was the dotcom era. Now the survival kit for the big boys is ASP. Everyone is an ASP [laughs].

"The concept is one-to-many--it is so beautiful that the model has to work. There is no way this business model cannot work." Looking back into this span of time, there are very few success stories. Out of thousands of ASPs today, only a handful are actually doing okay. What has happened is the whole thought process has been evolving and everyone is learning. Not only ASPs, any new business model takes a decade to stabilise. In the first phase ASP companies with limited resources chose to acquire first datacenter, then networking companies followed by system integration companies. They also invested in a few applications as well as the application code for customising it for niche segments and standardising it for markets in general. These early ASP companies invested huge amounts, which is actually not required in the service industry.

Contents

	The ASP business model
	The evolving ASP models
	Customer/vendor relationship

The ASP business model

Assuming that early ASPs acquired companies with different technology expertise, today any vendor running an application off the web is termed as an ASP. Is this also likely to change? How do you see the nature of the ASP vendor changing? Is there a change in the acquisition model?

ASP is a consortium business. Take the example of an IDC. It is a full-fledged managed facility for the purpose of colocation of servers. Now the margins are shrinking and the IDC doesn't get enough boxes to put there. So the IDC says "why don't I tie up with application software and networking companies and become an ASP". Or take the case of the network service provider (NSP). The NSPs have smaller IDCs for their own captive tools. But they say, "why don't I go up the value chain and become an ASP. After all I am the guy who is carrying the application."

Similarly for the ISVs. They say, "it is my application that is going to be running, I have the domain knowledge and expertise, so lets tie up with an IDC and become an ASP." So everybody wants to become an ASP and everyone wants to be called an ASP. In reality as the thought process goes, technology companies, box companies, ISVs, IDCs, NSPs, ISPs, all put together form a good consortium and can be called an ASP. This is the model that has evolved over the last two-three years.

Now as part of the learning curve and in the second phase, companies realised that you do not need to acquire--partnering is better. So ASP companies approached IDCs and NSPs and said, "You are my partner. These are the services that I require and this is your core competence. My core competence is ASP delivery and customer acquisition. You do your job and I will do my job and lets share the revenue." It started like that and the entire chain started forming. But some of the things like hardware were still missing and the learning curve continued.

Some ASP companies then said why should we invest in hardware. "Let's have a leasing model because redundancy and obsolescence are the biggest issues in hardware." So they approached Sun, IBM, Intel and others and said, it is your responsibility to keep replacing the hardware in my server farm and we can share the revenues. So the model drifted from buying, acquiring and building to partnering and revenue sharing.

One important thing that happened when customer orders started coming in was putting all these pieces together--or systems integration. And this was not a joke. Providing the application, network, datacenter, NOC in a timely and managed manner through an SLA (service level agreement) took a longer time than expected. For the ASP front end you need to have trained and skilled people which was missing in the other parts of the world. There was a huge scarcity of right people in the other parts of the world. So left, right and center, the global ASP companies--I will not name them--all of them acquired companies with system integration skills.

What you are saying is that multiple technology players need to team up in an ASP delivery model. But the basic issue is whether the ASP business and revenue model is sound? Is the premise of running an application off a remote server with users paying on a rental basis commercially viable in the current technology environment?

We studied the US, European and Asia Pacific markets to confirm our thought processes. What we found is that the ASP business model is very new--it is just two and half to three years old. While outsourcing is dog years old, the ASP model is new and has many pitfalls and requires long thought out processes. Now when you acquire a company, you pay a price and you cannot filter that price into your bottom line in a three-quarter time period.

So called ASPs, have been in situations where they have acquired up to eight companies in a quarter. Those acquisitions will take four, five, six years to show results, not in the first two years. Although ASP companies have not fared well, the opinion built up by media and analysts is that the model is not working. See, the concept is one-to-many--it is so beautiful that the model has to work. There is no way this business model cannot work. To make a judgmental remark on a business model at a two-three year point of time is really, very early.

Now the ASP business model is not very complicated or highly technical. But the way it has been perceived and the way it has been structured is different from what is required. In a particular business model you can see the results in an optimum manner if it is structured well. That was the flaw with the ASP model. It will take some time for this industry to mature and get the right kind of backend infrastructure and products for high customer satisfaction.

The evolving ASP models

How would you summarise a successful ASP business model?

ASPs need to adopt the pure-play model, which is the only single robust model. This says all partners will remain in their area of core competence, work in a seamless manner, enter into a revenue sharing or any other commercially viable arrangement and own the customer.

What would you say is the biggest advantage of the ASP business over other IT service opportunities?

The answer is scalability. By becoming an ASP I have no geographical boundaries and I can scale up like anything. This is a big restriction with outsourcing services today.

The process of deploying applications on a remote server and making it available for enterprise users has a number of benefits. Which are the major market segments that can be early adopters of this technology?

ASP is a beautiful model when you look at small and medium enterprises (SME) and is working well in other parts of the world. The question is why is it suitable for SMEs? Now SMEs have not been able to afford or have not taken the initiative to afford or invest in a large application.

If you consider any of the large applications like SAP, JD Edwards, PeopleSoft, Ramco Marshal or any enterprise application in the world, you have to spend a few million dollars on full implementation which takes twelve to twenty four months. For the SME segment this is not a good proposition and they have been investing in applications of smaller size and value. Now even mid-size and large companies are adopting the ASP model. That's a big boost to the movement.

What about specific market segments and which applications are more suited for ASP delivery?

The private sector banks, financial institutions, IT-communications-entertainment, transportation, hospitality among others. In the Indian sub-continent these segments are lagging behind in IT enablement. There is a huge potential here. These segments are service oriented. Their mind set is outsourcing because someone is outsourcing to them. In all growth companies the mindset is tech savvy, employee focussed and multi-locational. If you provide them a service based on technology solutions they are ready to pay a premium. The applications are payroll, financial accounting, HR and related ones.

During the last few years we have seen and heard about technology vendors like Intel, SAP, Compaq and others taking the lead as ASP vendors. So, there has been considerable initiative and investment to make this service model work with the end user. Yet there have not been too many takers. What are the reasons for this?

Now in the entire learning process over the last three years, the major issue that was missing was the customer--nobody thought of the customer. The reason was the ASP initiative was taken up not by service companies but by vendors. The power of vendors is selling products, they have not been selling solutions. Now ASP is purely a service industry and you need to first understand the customer requirements. Then customers need to be trained on how you are doing it. That part was missing.

The first offerings in any ASP portfolio are office suites and other desktop applications followed by more critical enterprise applications. Do ASPs tend to deliver applications for a specific vertical segment or do they offer a much broader horizontal application segmentation?

Now desktop applications cannot be offered in an ASP environment. You have to write the applications again. I don't say that it is technically not possible--some companies in the US are doing it--but commercially a whole lot of work has to be done. Office suites in an ASP environment, sounds very good, but the application architecture is built for the desktop.

There are multiple ASP solution models evolving around the world. There are routine models with say a hundred applications available, which are not specific to any vertical segment. The other model is to offer applications for specific segments where you have domain knowledge and expertise.

Is web enabling an application also an attribute for ASP applications?

More than 95% applications are not written for the web. But there are many other technical requirements than just making the application web enabled. It has to have an inherent strength to work in an ASP environment--one to many. A single application used by many users, simultaneously and in an isolated and secure manner.

We have heard Citrix claiming that all leading ASPs use their application management products. But ASPs tend not to talk about Citrix as a key enabler of their services. Rather they say that Citrix is not essential. In the absence of Metaframe is there any other way that applications can be managed on the server?

Citrix is an ASP enabler. Citrix has lovely products, which help an ASP in a heterogeneous application environment to seamlessly talk to each other at the back-end. Band-width optimisation is one of the major advantages and the thin client-ICA (independent client architecture) environment is very good. It is a very effective alternative. But there are alternative ways of managing applications without Citrix.

The most critical and important factor for delivery of an ASP application is the architecture. It is very important. The reason why Citrix played such a big role in the ASP business was that 95% applications were not written in web language. And with Citrix they straight-away become web enabled--it's like magic.

If the application architecture is strong and robust, supports cluster and load balancing then it already has features that Citrix as a separate software package would provide. Unfortunately or fortunately, the earlier applications were not written in this form or capable of taking care of these needs that have arisen today. Future applications will definitely have these features built in and will not require Citrix.

ASP vendor/customer relationship

In a nutshell what is your opinion on the commercial feasibility of enterprise wide applications through an ASP delivery model?

Mission critical, sensitive applications, which truly have a market and make sense commercially for an ASP cannot be delivered right away.

Let's look at the infrastructure required for an ASP delivery model. A key component is the data center and a key issue is the lack of response time or latency, due to last mile congestion. Will deployment of the ASP application backend on multiple data centres take care of this requirement for some time?

Latency is something than can be solved technologically very easily. Using geographically distributed data centres to provide faster access for the customer is one way of looking at the entire solution. But you can still have one data centre serving the entire world. The customer will not even know where your data centre is provided it has a pipe and uptime suitable for that application. What is important are the pipes behind the ASP solution, the technology and the security mechanisms. The customer wants his SLA to be very clean. When he clicks on an icon he gets the application seamlessly and in real time. He doesn't care whether it's a Sun or NT box sitting in multiple data centres. Secondly the last mile is still left out.

Assuming that the last mile problem, which brings in unreliability of response time is going to continue, how do you extend an SLA to the customer covering access to the ASP application portfolio?

Till we take care of the bandwidth situation and the networks are well settled, the internet is a most unreliable cloud. I have said the internet is not a secure medium. An SLA for a customer coming from the internet is the most difficult task. He brings a lot of dirt into the system. Even a small payroll application is very sensitive to the company.

If I have to give an SLA to my customer I need to have an SLA behind my back. Each partner has to give me an SLA, which I have to put in my SLA for the customer. I own the customer. The catch is how do I own the product without actually owning it.

At the end of the road, after all the dust has settled, what is the relationship that an ASP vendor will have with the customer?

In a mature ASP life cycle, an enterprise will not have many service providers. They will have one or two service providers. They will have one pipe and one service provider and their expectations from that service provider will go up everyday. It is for the service provider to keep abreast with the latest technologies and requirements of the customer. Otherwise they will lose the customer.

An individual or a company do not use multiple banks. They are very happy with one bank. The same thing is going to happen with ASPs. The service provider will be one. In a real mature cycle--each of the employees is also my customer. So I have to meet all the needs of the individual employees otherwise I will be out.

Copyright © 2009 CBS Interactive, a CBS Company. All Rights Reserved.
ZDNET is a registered service mark of CBS Interactive. ZDNET Logo is a service mark of CBS Interactive.